Guidelines For Evaluation and Management of LBP

C L I N I C A L G U I D E L I N E F O R T H E E VA L U AT I O N A N D M A N A G E M E N T O F
Guideline
for
the
Evaluation and Management of Low Back Pain

Evidence Review
L O W B A C K PA I N
roger Chou, Md laurie hoyt huffman, MS
American Pain Society, Publisher Glenview, il
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain
TABLE OF CONTENTS Introduction
Page
1
Purpose of evidence review ......................................................................1 Background

1
Previous guidelines...................................................................................2 Scope of evidence review

4
Key questions ...........................................................................................4 Populations ...............................................................................................6 Interventions .............................................................................................6 Non-invasive interventions ....................................................................6 Invasive, non-surgical interventions ......................................................7 Surgical interventions ............................................................................8 Invasive diagnostic tests .......................................................................8 Outcomes..................................................................................................8 Conflict of interest ............................................................................................9 Methods
9
Literature search and strategy ..................................................................9 Inclusion and exclusion criteria .................................................................10 Data extraction and synthesis ...................................................................11 Systematic reviews................................................................................11 Individual trials on efficacy and safety of interventions..........................11 Observational studies of treatment efficacy...........................................13 Studies of invasive diagnostic tests.......................................................13 Dual review ...............................................................................................13 Assessing research applicability and clinical relevance ............................13 Data synthesis ..........................................................................................14 Rating of good quality............................................................................14 Rating of fair quality...............................................................................14 Rating of poor quality ............................................................................14 Assessing magnitude of benefits or harms ...............................................15
i
American Pain Society
EVIDENCE REVIEW
TABLE OF CONTENTS Results
Page
16
Size of literature reviewed.........................................................................16 Key Question 1a. How accurate are features of the history and physical exam for predicting presence of serious underlying conditions (red flags) or other conditions that may be responsive to specific therapies in patients with low back pain (such as nerve root compression or spinal stenosis)?.......................................................17 Accuracy of history and physical exam features for identifying specific diagnoses associated with low back pain ..................................18 Cancer ................................................................................................18 Infection ..............................................................................................19 Cauda equina syndrome.....................................................................19 Compression fracture .........................................................................19 Ankylosing spondylitis.........................................................................19 Herniated disk or radiculopathy ..........................................................20 Spinal stenosis ...................................................................................20 Key Question 1b. How accurate are features of the history and physical exam for predicting the development of persistent low back pain and associated disability (yellow flags)?.................................23 Accuracy of history and physical exam features for identifying patients more likely to develop chronic and disabling low back pain...............................................................................................23 Key Question 1 Summary Table: Systematic reviews of prognostic factors for identifying patients more likely to develop chronic and disabling low back pain ..................25 Key Question 1c. How effective is identification and treatment of yellow flags for improving clinical outcomes in patients with low back pain?.................................................................................................29 Efficacy of interventions for identifying and treating yellow flags...........30 Key Question 2a. How accurate are different diagnostic tests for identifying serious underlying conditions (e.g., tumor, infection, compression fracture)? .............................................................................34 Accuracy of imaging for diagnosing cancer ...........................................35 Accuracy of imaging for diagnosing vertebral infection .........................35
ii
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Accuracy of imaging for diagnosing vertebral compression fracture ..................................................................................................35 Accuracy of elevated erythrocyte sedimentation rate for diagnosing cancer .................................................................................36 Key Question 2b. How accurate are different diagnostic tests for identifying other conditions (e.g. nerve root compression, herniated disc, spinal stenosis) that may respond to specific therapies? .................................................................................................37 Accuracy of imaging for diagnosing nerve root compression, herniated disk, and spinal stenosis........................................................37 Accuracy of imaging for diagnosing ankylosing spondylitis ...................38 Diagnostic accuracy of other (non-imaging) tests..................................38 Key Question 2c. In patients with red flags, how effective are different diagnostic tests for improving patient outcomes? .......................40 Efficacy of diagnostic testing in patients with red flags ..........................40 Key Question 2d. In patients without red flags, how effective are different diagnostic tests or test strategies (including no testing) for improving patient outcomes? ...............................................................40 Efficacy of routine, early plain radiography versus usual care or imaging only if clinically necessary (or without improvement) ...............40 Efficacy of routine MRI versus MRI only if clinically necessary (or without improvement).......................................................................43 Efficacy of rapid MRI versus lumbar radiography in patients with low back pain referred for imaging .................................................43 Key Question 3. How effective are self-care advice, education, or other self-care interventions for improving patient outcomes? ..................46 Self-care advice or education ................................................................46 Advice to rest in bed ..........................................................................46 Advice to remain active......................................................................48 Self-care back education books.........................................................53 Self-care e-mail discussion groups ....................................................58 Self-care exercise videotape..............................................................59 Advice to restrict early morning flexion ..............................................60 Lay-facilitated groups for self-care.....................................................61
iii
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Self-help tools for back surgery decisions..........................................62 Other self-care interventions .................................................................64 Lumbar supports................................................................................64 Mattresses .........................................................................................65 Superficial heat or cold ......................................................................67 Key Question 3 Summary Tables: Systematic reviews on efficacy of self-care therapies for low back pain...............................................................................................70 Summary of evidence on self-care therapies for acute low back pain...............................................................................................72 Summary of evidence on self-care therapies for chronic or subacute low back pain .........................................................................73 Summary of evidence on self-care therapies for radiculopathy or sciatica ..............................................................................................75 Key Question 4. How effective are different non-invasive interventions for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? ..........................76 Medications ...........................................................................................76 Acetaminophen..................................................................................76 Non-steroidal anti-inflammatory drugs (NSAIDs) ...............................78 Non-selective NSAIDs....................................................................78 COX-2 selective NSAIDs................................................................81 Aspirin................................................................................................82 Other medications..............................................................................83 Antidepressants..............................................................................83 Benzodiazepines ............................................................................86 Antiepileptic drugs ..........................................................................87 Skeletal muscle relaxants...............................................................90 Opioid analgesics ...........................................................................92 Tramadol ........................................................................................98 Systemic corticosteroids.................................................................101 Topical lidocane .............................................................................103
iv
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Herbal therapies .............................................................................103 Key Question 4 Summary Tables: Systematic reviews on efficacy of medications and herbal therapies for low back pain....................................................................106 Summary of evidence on medications and herbal therapies for acute low back pain...............................................................................111 Summary of evidence on medications and herbal therapies for chronic or subacute low back pain ........................................................113 Summary of evidence on medications for sciatica or radicular low back pain ........................................................................................115 Acupuncture and related interventions ..................................................116 Acupuncture and dry needling ...........................................................116 Acupressure.......................................................................................121 Neuroreflexotherapy ..........................................................................122 Educational interventions ......................................................................124 Back schools......................................................................................124 Brief educational interventions...........................................................126 Exercise and related interventions ........................................................129 Exercise .............................................................................................129 Hydrotherapy .....................................................................................134 Yoga ..................................................................................................135 Interdisciplinary interventions ................................................................137 Interdisciplinary rehabilitation (multidisciplinary rehabilitation)...........137 Functional restoration (physical conditioning, work conditioning, or work hardening)........................................................140 Physical modalities................................................................................142 Interferential therapy ..........................................................................142 Low level laser therapy ......................................................................144 Shortwave diathermy .........................................................................147 Traction..............................................................................................149 Transcutaneous electrical nerve stimulation (TENS) .........................151
v
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Percutaneous electrical nerve stimulation (PENS) ............................153 Ultrasound .........................................................................................156 Other non-invasive interventions ...........................................................157 Psychological therapies .....................................................................157 Massage ............................................................................................159 Modified work.....................................................................................162 Spa therapy and balneotherapy.........................................................163 Spinal manipulation ...........................................................................165 Key Question 4 Summary Tables: Systematic reviews on efficacy of non-pharmacologic therapies for low back pain ...................................................................................172 Summary of evidence on non-pharmacologic therapies for acute low back pain...............................................................................185 Summary of evidence on non-pharmacologic therapies for chronic or subacute low back pain ........................................................187 Summary of evidence on non-pharmacologic therapies for radiculopathy or sciatica........................................................................190 Key Question 5. How effective are decision tools or other methods for predicting which patients are more likely to respond to specific therapies like spinal manipulation or different types of exercise therapy?......................................................................................191 Reliability and validity of manual spinal palpatory exam or clinical tests of the sacroiliac joint..........................................................191 Utility of clinical prediction rules for spinal manipulation........................191 Clinical prediction rules for exercise ......................................................193 Patient classification systems for individualizing physical therapy interventions .............................................................................193 Key Question 6. How effective is referral from primary care providers to back specialty providers for improving patient outcomes? What are the outcomes for patients who are managed by different types of care providers or by multidisciplinary or interdisciplinary clinics? ............................................................................195
vi
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Key Question 7. What is the diagnostic accuracy and what are the potential harms associated with invasive tests for identifying patients who may benefit from invasive procedures? How effective is prior use of these tests for selecting patients for invasive procedures in improving outcomes? ...........................................197 Provocative discography .......................................................................197 Diagnostic selective nerve root blocks ..................................................205 Diagnostic intra-articular facet joint blocks and medial branch blocks ........................................................................................206 Diagnostic sacroiliac joint blocks ...........................................................208 Key Question 8. How effective are injections (and different injection interventions) and other interventional therapies for nonradicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? ................................................209 Injections outside the spine ...................................................................209 Local injections ..................................................................................209 Botulinum toxin ..................................................................................212 Prolotherapy ......................................................................................213 Intraspinal steroid injection and chemonucleolysis................................215 Epidural steroid injection....................................................................215 Intradiscal steroid injection.................................................................234 Chemonucleolysis..............................................................................236 Radiofrequency denervation, intradiscal electothermal therapy (IDET), and related procedures .........................................................243 Radiofrequency denervation ..............................................................243 Intradiscal electothermal therapy (IDET) ...........................................249 Percutaneous intradiscal radiofrequency thermocoagulation (PIRFT) and Coblation nucleoplasty ...............................................252 Spinal cord stimulation.......................................................................254 Key Question 8 Summary Tables: Systematic reviews on efficacy of interventional therapies for low back pain...............................................................................................256
vii
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Summary of evidence on interventional therapies for low back pain...............................................................................................269 Key Question 9. How effective is surgery (and different surgical interventions) for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? .........................272 Surgery for non-radicular low back pain with common degenerative changes ...........................................................................272 Surgery for isthmic spondylolisthesis ....................................................279 Surgery for spinal stenosis with or without degenerative spondylolisthesis ...................................................................................282 Surgery for radiculopathy with herniated lumbar disc............................291 Key Question 9 Summary Tables: Systematic reviews on efficacy of surgery for low back pain .................299 Summary of evidence on surgery for low back pain ..............................311 Key Question 10. How effective are combinations of therapies for acute and chronic low back pain? .............................................................313 Combinations of medications ................................................................313 Self-care advice combined with other interventions ..............................315 Exercise combined with other interventions ..........................................317 Acupuncture combined with other non-invasive interventions ...............319 Spinal manipulation combined with other interventions.........................320 Massage combined with other interventions..........................................322 Psychological therapies combined with other interventions ..................322 Traction combined with other interventions ...........................................323 Combination therapy for spinal stenosis................................................324 Key Question 11. How effective are different treatment strategies for failed back surgery syndrome? ............................................................325 Adhesiolysis and forceful epidural injection ...........................................325 Intrathecal therapy.................................................................................331 Non-invasive interventions ....................................................................333 Spinal cord stimulation ..........................................................................334
viii
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Key Question 12. How effective are different methods of integrating or coordinating low back pain care? ........................................336 Key Question 13. How effective are interventions for secondary prevention of low back pain in patients who have had an episode of acute low back pain, or for prevention of flares of low back pain in patients with chronic low back pain? .....................................................338 Back schools .........................................................................................338 Exercise.................................................................................................339 Lumbar supports ...................................................................................341 Advice to stay active..............................................................................342 Early occupational medicine intervention ..............................................343 Psychological therapies, interdisciplinary rehabilitation, spinal manipulation, acupuncture, patient information or education ................344 Key Question 14. How effective are interventions for managing low back pain during pregnancy and post-partum?...................................344 Acupuncture during pregnancy..............................................................344 Physical therapy during pregnancy .......................................................346 Massage during pregnancy ...................................................................348 Supportive devices during pregnancy ...................................................349 Key Question 15. What is the cost-effectiveness associated with different interventions or management strategies (such as care provided by different types of providers) for managing low back pain? .........................................................................................................350 Summary and discussion
350
Research gaps..........................................................................................350
ix
EVIDENCE REVIEW
TABLE OF CONTENTS
Page
Bibliography .....................................................................................................352 APPENDICES Appendix 1: Systematic reviews search strategies .........................................401 Appendix 2: Primary studies search strategies...............................................402 Appendix 3: Systematic reviews quality rating system....................................419 Appendix 4: Randomized controlled trials quality rating system .....................421 Appendix 5: Quality ratings of included systematic reviews............................423 Appendix 6: List of excluded systematic reviews............................................436 Appendix 7: Quality ratings of trials ................................................................449 Appendix 8: Quality ratings of diagnostic accuracy trials ................................466 Appendix 9: Systematic reviews published too recently to be included in this evidence review................................................................468 Appendix 10: Glossary......................................................................................469
x
EVIDENCE REVIEW
INTRODUCTION
Purpose of evidence review
This evidence review focuses on evaluation and management of low back pain in adults. The American Pain Society (APS), which commissioned this report, used it to develop evidencebased clinical practice guidelines on evaluation and management of low back pain. The guidelines were developed in two stages. The first stage, published in October 2007, focused on initial (primary care) evaluation and management of low back pain, and was conducted in partnership with the American College of Physicians1. The second stage, published in May 2009, focused on use of interdisciplinary rehabilitation, interventional therapies, and surgery for low back pain2.
BACKGROUND
Low back pain is extremely common. Though estimates vary widely, studies in developed countries report point prevalences of 12% to 33%, one-year prevalences of 22% to 65%, and lifetime prevalences of 11% to 84%3. In the U.S., nonspecific mechanical low back pain is the fifth most common reason for all physician visits, and the second most common symptomatic reason, accounting for approximately 2.3% of all physician visits4, 5. About one-quarter of U.S. adults report low back pain lasting at least a whole day in the last three months5. 7.6% of U.S. adults randomly surveyed by telephone had at least one occurrence of severe acute low back pain during a one-year period, with 39% of those seeking medical care for the episode6. Low back pain is also very costly. In 1998, total health care expenditures incurred by individuals with back pain in the U.S. were $90.7 billion, with incremental costs attributed to back pain $26.3 billion7. Medical treatment for chronic low back pain is estimated to cost $9,000 to $19,000 per patient annually, and interventional treatments cost a minimum of $13 billion in 19908. Additional costs are associated with days lost from work due to low back pain. Low back pain is the most common cause for chronic or permanent impairment in U.S. adults under the age of 65, and the most common cause of activity limitations in persons under the age of 459. Between 2% and 8% of the U.S. work force is disabled or compensated for back injuries each year8, 9. Approximately 5% of people with back pain disability are thought to account for 75% of the costs associated with low back pain10. Many patients with acute episodes of low back pain do not seek care because symptoms are often brief and self-limited. Among those who do seek medical care, rapid improvements in pain (average improvement of 58% of initial score), disability (average improvement of 58%), and return to work (82% of those initially off work return to work) are seen in the first month11. Further improvement generally occurs through approximately three months, after which pain or disability levels and rates of return to work tend to remain relatively constant. Up to one-third of patients report persistent back pain of at least moderate intensity one year after an acute episode requiring care, and one in five report substantial activity limitations12. Recurrences of pain also are common, with 60% to three-quarters of patients experiencing at least one relapse
1
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain within 12 months11, 13. Factors associated with the development of chronic disability due to low back pain include pre-existing psychological conditions and distress, presence of other types of chronic pain, job dissatisfaction or stress, and disputes over compensation issues14. Many options are available for the evaluation and management of acute or chronic low back pain. However, there has been little consensus, either within or between specialties, on appropriate uses of diagnostic tests15 and interventions16. This is demonstrated by numerous studies showing unexplained variations in use of diagnostic tests and treatment. The rate of back surgery in the U.S., for example, is over five times higher than the rate in the U.K.17. Within Washington State, rates of back surgery vary up to 15-fold among different counties18. Despite wide variations in practice, several studies have shown that patients experience broadly similar outcomes, though costs of care can differ substantially both between and within specialties19, 20. In addition to unexplained practice variations, another historical feature of low back pain management has been the widespread uptake and use of unproven (and sometimes invasive and costly) interventions, some of which have later been shown to be ineffective, or even harmful21. Other interventions are widely used despite studies showing only marginal benefits22.
Previous guidelines
The Quebec Task Force on Spinal Disorders published one of the first evidence-based clinical practice guidelines for management of low back pain in 198723. This early attempt at using an explicit scientific basis for issuing management recommendations found insufficient evidence to support the use of most common diagnostic procedures and treatment modalities. In 1994, a multidisciplinary expert panel convened by the U.S. Agency for Health Care and Policy Research (AHCPR) issued its recommendations on management of acute low back pain24. The approach recommended by the AHCPR guidelines emphasizes history taking and physical examination to exclude red flag symptoms suggestive of serious underlying pathology; targeted physical examination focusing on neurologic screening; diagnostic triage into broad categories including nonspecific low back pain, radicular syndrome, or specific pathology (which were felt to be diagnosable in only a small minority of cases); judicious use of diagnostic testing; and consideration of psychosocial factors when there is no improvement. Despite an exhaustive literature search and review, none of the 40 recommendations made for clinical care were viewed as supported by strong research evidence, and only six were judged as supported by at least moderate quality evidence. At the time, the AHCPR guidelines were subject to intense criticism and scrutiny25. Nonetheless, nearly all multidisciplinary guidelines published since 1994 have recommended an approach similar to the AHCPR guidelines26. There are now at least 11 international guidelines for management of low back pain. Most of their diagnostic and therapeutic recommendations are similar26. However, there are some discrepancies, particularly with regard to recommendations for exercise therapy, spinal manipulation, use of muscle relaxants, and provision of patient information. These differences may in part reflect contextual differences between countries that can affect interpretations of the evidence and how the trade-offs between benefits, side effects, and costs are weighted27. In
2
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain addition, systematic reviews of back pain guidelines found several important areas in which the overall quality of guidelines could be improved, including better descriptions of how the evidence was identified, selected and summarized; more attention to patient preferences; increased consideration of how guidelines could be implemented; better use of external peer review; and more transparent descriptions of editorial oversight and potential conflicts of interests28, 29. Most low back pain guidelines have focused on management of acute low back pain, and do not provide specific guidance for management of chronic low back pain28. The effects of using evidence-based clinical practice guidelines on clinical outcomes in patients with low back pain are difficult to assess. However, several trials evaluating outcomes associated with the selective imaging approach recommended in nearly all guidelines are now available (see Results, Key Question 2d). In addition, an observational study from Australia found back care based on guidelines and provided in multidisciplinary clinics was associated with improved pain scores after 12 months, decreased use of imaging and opioid medications, greater patient satisfaction, and decreased health care costs compared to usual care provided in general practice clinics30. A challenge in interpretation of this study is that it is difficult to know how much of the benefit was related to following guidelines and how much to provision of care by multidisciplinary clinics. Another observational study found a mass-media campaign in the state of Victoria, Australia based on evidence-based guidelines (encouragement of normal activities, exercise, and continued work while providing positive messages about likelihood of recovery) and aimed at altering back pain beliefs was associated with a decline in the number of claims for back pain, rates of days compensated, and medical payments for low back pain claims compared to a neighboring state without such a campaign31. Changes in clinician beliefs about back pain and reported back pain management appeared to be sustained 4.5 years after the end of the media campaign32. A U.S. trial found randomization of communities to an educational intervention for low back pain based on national guidelines resulted in a decline in the rate of surgery by about 9% compared to usual care33. The American Pain Society initiated this project to systematically review the current state of evidence and develop updated recommendations for management of acute and chronic low back pain using an evidence-based, balanced, and multidisciplinary approach. Throughout this evidence report, we highlight previous recommendations and findings from the AHCPR guidelines24. We also summarize recommendations from a federally funded U.S. guideline issued by the Veterans Affairs/Department of Defense (VA/DoD) in 199934 and a guideline issued by the U.K. Royal College of General Practitioners (RCGP), which was initially released in 199635 and updated in 199936. The AHCPR, VA/DoD, and UK RCGP guidelines primarily focus on acute low back pain, though some recommendations for evaluation and treatment of persistent or chronic low back pain were included. We also summarize recommendations from a recent, multinational guideline from Europe issued in 2004 (the European COST B13 guidelines) addressed both acute and chronic low back pain, as well as prevention of back pain37-39. Methods used to grade strength of evidence by these guidelines are as follows:
3
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain AHCPR, VA/DoD, and European COST guidelines A = Strong research-based evidence (multiple relevant and high-quality scientific studies) B = Moderate research-based evidence (one relevant, high-quality scientific study or multiple adequate scientific studies) C = Limited research-based evidence (at least one adequate scientific study in patients with low back pain) D = Panel interpretation of information that did not meet inclusion criteria as researchbased evidence UK RCGP guidelines: *** Generally consistent finding in a majority of multiple acceptable studies ** Either based on a single acceptable study, or a weak or inconsistent finding in some of multiple acceptable studies * Limited scientific evidence, which does not meet all the criteria of acceptable studies Although the European COST guidelines use the AHCPR method for grading evidence, they do not explicitly grade strength of recommendations.
SCOPE OF EVIDENCE REVIEW

Key Questions
The Key Questions used to guide this evidence review were developed by a multidisciplinary expert panel convened by the American Pain Society. The Key Questions were viewed as critical questions that needed to be answered in order to develop clinical practice guidelines. 1a. How accurate are features of the history and physical exam for predicting presence of serious underlying conditions (red flags) or other conditions that may be responsive to specific therapies in patients with low back pain (such as nerve root compression or spinal stenosis)? 1b. How accurate are features of the history and physical exam for predicting the development of persistent low back pain and associated disability (yellow flags)? 1c. How effective is identification and treatment of yellow flags for improving clinical outcomes in patients with low back pain? 2a. How accurate are different diagnostic tests for identifying serious underlying conditions (e.g., tumor, infection, compression fracture)? 2b. How accurate are different diagnostic tests for identifying other conditions (e.g. nerve root compression, herniated disc, spinal stenosis) that may respond to specific therapies?
4
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain 2c. In patients with red flags, how effective are different diagnostic tests for improving patient outcomes? 2d. In patients without red flags, how effective are different diagnostic tests or test strategies (including no testing) for improving patient outcomes? 3. How effective is self-care advice, education, or other self-care interventions for improving patient outcomes? 4. How effective are different non-invasive interventions for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? 5. How effective are decision tools or other methods for predicting which patients are more likely to respond to specific therapies like spinal manipulation or different types of exercise therapy? 6. How effective is referral from primary care providers to back specialty providers for improving patient outcomes? What are the outcomes for patients who are managed by different types of care providers or by multidisciplinary or interdisciplinary clinics? 7. What is the diagnostic accuracy and what are the potential harms associated with invasive tests for identifying patients who may benefit from invasive procedures? How effective is prior use of these tests for selecting patients for invasive procedures in improving outcomes? 8. How effective are injection procedures (and different injection interventions) and other interventional therapies for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? 9. How effective is surgery (and different surgical interventions) for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? 10. How effective are combinations of therapies for acute and chronic low back pain? 11. How effective are different treatment strategies for failed back surgery syndrome? 12. How effective are different methods of integrating or coordinating low back pain care? 13. How effective are interventions for secondary prevention of low back pain in patients who have had an episode of acute low back pain, or for prevention of flares of low back pain in patients with chronic low back pain? 14. How effective are interventions for managing low back pain during pregnancy and postpartum? 15. What is the cost-effectiveness associated with different interventions or management strategies (such as care provided by different types of providers) for managing low back pain?
5
EVIDENCE REVIEW
Populations
Target populations for this review are: Adults (>18 years old) Pregnant women (not including management of back pain during labor) Persons with acute (less than 4 weeks), subacute (between 4 weeks and 3 months) or chronic (greater than 3 months) low back pain
Persons with non-radicular low back pain (including presumed discogenic pain, presumed
facet joint pain, spondylosis, degenerative disc disease, presumed sacroiliac joint pain, etc.), radicular low back pain (including symptomatic nerve root compression associated with lumbar disc prolapse), spinal stenosis, degenerative or isthmic spondylolisthesis, and failed back surgery syndrome Treatment of spinal infection, cauda equina syndrome, cancer, spondyloarthropathies, systemic inflammatory disease, fibromyalgia syndrome, and vertebral compression fracture was excluded from the scope of this review, though evaluation to rule out such conditions was considered within the scope. Evaluation and management of osteoporosis without clear fracture and acute major trauma was also outside the scope of this review. Evaluation and management of children and adolescents with low back pain was also excluded, because diagnostic and therapeutic considerations are substantially different than in adults40, 41. Low back pain presents as a continuum ranging from acute (often defined as less than 4 weeks in duration) to chronic (often defined as greater than three months in duration). Patients may present to providers at any stage on this continuum, have mixed presentation (e.g., chronic low back pain with an acute exacerbation), or unclear date of onset. In addition, many trials evaluate mixed populations of patients with different durations of symptoms. Therefore, we reviewed evidence on low back pain of any duration.
Interventions
Target interventions (see Glossary for how interventions were defined) for this review are:
Non-invasive interventions Medications

Acetaminophen Non-selective non-steroidal anti-inflammatory drugs (NSAIDs) Cyclo-oxygenase-2 selective NSAIDs Aspirin Skeletal muscle relaxants Antidepressants Opioid analgesics
6
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Tramadol Antiepileptic drugs Systemic corticosteroids Topical lidocaine Interventions involving injection of medications into the back (such as botulinum toxin; local, epidural or intradiscal steroid injections; and intrathecal administration of medications) are covered under invasive, non-surgical interventions (see below).
Other non-invasive interventions

Herbal therapies Brief educational interventions Back schools Exercise Hydrotherapy Spa therapy Acupuncture Acupressure Neuroreflexotherapy Spinal manipulation Massage Shortwave diathermy Interferential therapy Ultrasound Psychological therapies
Interdisciplinary (multidisciplinary) rehabilitation

Functional restoration/physical conditioning programs/work hardening Traction Low-level laser therapy Self-care interventions (including advice for bed rest or on remaining active and self-care books) Modified work
Invasive, non-surgical interventions

Epidural steroid injection Intradiscal steroid injection Chemonucleolysis Local anesthetic injections (including tender or trigger point injections)
7
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Facet (zygapophysial) joint injection Therapeutic medial branch block Prolotherapy (sclerosant injection) Botulinum toxin Adhesiolysis and forceful epidural injection Radiofrequency denervation Intradiscal electrothermal therapy (IDET) Percutaneous intradiscal radiofrequency thermocoagulation Intrathecal therapy Spinal cord stimulation (Percutaneous discectomy and related procedures were considered surgical interventions)
Surgical interventions
Fusion and vertebral disc replacement for non-specific low back pain and degenerative disc disease Surgery for degenerative spondylolisthesis Surgery for spinal stenosis and lumbar isthmic spondylolisthesis Discectomy for lumbar disc prolapse (including open discectomy, microdiscectomy, laser- or endoscopic-assisted discectomy, percutaneous automated discectomy with nucleotome, Coblation nucleoplasty, and disc Dekompressor)
Invasive diagnostic tests

Provocative discography Selective nerve root block Facet joint block and medial branch block Sacroiliac joint block
Outcomes
We selected target outcomes based on the five core domains for low back pain suggested in recent recommendations: back specific function, generic health status, pain, work disability, and patient satisfaction42, 43. The two most commonly used measures of back-specific function are the Roland Morris Disability Questionnaire (RDQ) and the Oswestry Disability Index (ODI)44. The RDQ is reported on a 0 to 24 scale and the ODI on a 0 to 100 scale. Improvements of 2-3 points on the RDQ and 10 points on the ODI have been proposed as minimal clinically important differences45.
8
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Studies usually evaluate generic health status with the Medical Outcomes Study Short Form-36 (SF-36) or other multi-question assessments. These questionnaires measure how well an individual functions physically, socially, cognitively, and psychologically. The SF-36 measures 8 dimensions, each on a 0 to 100 scale46. The individual dimensions can also be combined into several commonly reported subscales (such as the Physical Component Summary and Mental Component Summary). Most studies measure pain intensity using either visual analogue or categorical pain scales (using either numbers or a list of adjectives describing different levels of pain intensity)47. Visual analogue scales (VAS) usually consist of a line on a piece of paper labeled 0 at one end, indicating no pain, and a maximum number (commonly 10 or 100) at the other, indicating excruciating pain. Patients designate their current pain level on the line. Categorical pain scales, on the other hand, consist of several pain category options from which a patient must choose (e.g., no pain, mild, moderate, or severe for a verbal rating scale, 0 to 10 for a numerical rating scale such as the Brief Pain Inventory). Many studies also report the proportion of patients with significant improvement in pain, often defined as at least a 20-point (or 20%) improvement on a VAS48. The SF-36 bodily pain scale has been recommended as a preferred method for reporting pain outcomes because it measures both pain intensity and interference with activities42. Work status is often measured by employment status, days off work, or time before returning to work. Patient satisfaction is usually assessed using a generic global scale, though more formal methods have been developed. Some studies also report effects of interventions on mood or the preference for one medication over another. We also reviewed evidence on adverse events and safety as well as costs. We converted cost data using other currencies to U.S. dollars using conversion rates as of January 2007 (1 British pound=$1.96 U.S., 1 Euro=$1.30 U.S., kr 1 kroner=$0.143 U.S.)
CONFLICT OF INTEREST
The evidence review was conducted at the Oregon Evidence-based Practice Center with funding from APS. None of the investigators conducting this review (RC and LHH) have any known conflicts of interest to disclose.
METHODS
Literature search and strategy
We searched the topic of low back pain using multiple electronic databases. The searches were performed in stages. All searches were initially conducted from 1966 (the start date of MEDLINE) through July 2005 and updated through November 2006. Searches for Key Questions 7 (invasive diagnostic tests), 8 (interventional therapies), and 9 (surgery) were subsequently updated through July 2008. In addition to MEDLINE, we searched for systematic reviews using the Cochrane Database of Systematic Reviews and the NHA Health Technology Assessment Programme and for primary studies using the Cochrane Central Register of
9
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Controlled Trials, EMBASE, PsychINFO (mental health topics), CINAHL (nursing and allied health topics), and PEDro (physical therapy topics), as appropriate. Searches for primary studies initially targeted only those interventions for which we identified no relevant, recent, higher-quality systematic review. We later modified our approach so that searches for primary studies were conducted for all invasive diagnostic tests, interventional therapies, and surgery, regardless of availability of previously published systematic reviews. Detailed search strategies are shown in Appendix 1 (systematic reviews) and Appendix 2 (primary studies). Electronic searches were supplemented by reviews of reference lists and additional citations suggested by experts.
Inclusion and exclusion criteria

All identified citations were imported into an electronic database (EndNote 9.0) and considered for inclusion. Papers were selected for full review if they met all of the following criteria: 1. Were about low back pain and evaluated a target population 2. Were relevant to a Key Question 3. Evaluated prognostic factors for low back pain, at least one target diagnostic test, or at least one target low back pain intervention 4. Reported predictive values for prognostic factors, accuracy of diagnostic tests, or at least one target outcome (pain, function, generic health status, work disability, or patient satisfaction) associated with a low back pain intervention We included relevant controlled clinical trials and systematic reviews. We excluded outdated systematic reviews, which we defined as systematic reviews with a published update, or systematic reviews published before the year 2000. Because of the large scope and body of literature covered by this review, we included controlled observational studies only for surgical interventions and for assessment of adverse events. Other observational studies (such as uncontrolled case series and pre-post analyses) were excluded. Studies of cost were included if they were conducted alongside a randomized trial or were a full economic analysis (costeffectiveness, cost-minimization, or cost-utility study)49. For prognosis or diagnostic accuracy, we only included systematic reviews. The exception was for invasive diagnostic tests (discography, facet joint block, medial branch block, diagnostic selective nerve root block and diagnostic sacroiliac joint block), where we also included primary studies that focused on clinical outcomes. We only included non-English language trials if they were already included in English-language systematic reviews. Studies of non-human subjects and those without original data were excluded. We also excluded studies published only as conference abstracts.
10
EVIDENCE REVIEW
Data extraction and synthesis Systematic reviews

For each systematic review, we abstracted the following information: 1. Purpose of the review 2. Databases searched 3. Dates of the searches 4. Language restrictions, if any 5. Number of studies included 6. Criteria used to include studies 7. Limitations of the included studies 8. Methods for rating the quality of included studies 9. Methods for synthesizing the evidence 10. The interventions evaluated 11. Main efficacy outcomes (including number and quality of studies for each comparison and outcome) 12. Adverse events The reliability of systematic reviews depends on how well they are conducted. We used predefined criteria to assess the internal validity of included systematic reviews. We assessed the internal validity (quality) of systematic reviews using the methods developed by Oxman and Guyatt (Appendix 3)50. Each study was scored between 1 and 7 based on the following criteria: comprehensiveness of search strategy; application of pre-defined inclusion criteria to select studies; appropriate assessment of validity; and use of appropriate methods to synthesize the evidence. Using this system, systematic reviews with a score of four or less are considered to have potential major flaws and we classified these as lower quality. Systematic reviews with major flaws are more likely to produce positive conclusions about the effectiveness of interventions51-53. We considered systematic reviews with scores of five or more higher quality.
Individual trials on efficacy and safety of interventions

We independently abstracted all randomized trials of interventional therapies versus placebo or sham therapy, surgery versus non-surgical therapy, and artificial disc replacement versus fusion. We also abstracted recent, large (N > 250) trials of non-invasive therapies and activecontrolled trials of interventional therapies and surgeries that were not included in a previously published, higher-quality systematic review. We did not abstract randomized trials (placebo- or active-controlled) of non-invasive therapies or active controlled trials of interventional therapies or surgery if they were included in a higher-quality systematic review. Instead, we relied on the
11
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain systematic reviews to determine the number and quality of trials and estimate the magnitude of effects for each comparison and outcome of interest. Although methods for rating internal validity varied across systematic reviews, we considered studies that received more than half of the maximum possible quality score to be higher-quality for any quality rating system used54, 55. For systematic reviews that only assigned a categorical overall grade for quality, we considered studies higher-quality if they were rated good, high-quality, or the equivalent. For each clinical trial not included in a higher-quality systematic review, we abstracted the following information: 1. Study design 2. Purpose of study 3. Inclusion and exclusion criteria 4. Number of patients approached, eligible, and randomized 5. Demographics and baseline characteristics 6. Setting 7. Funding source 8. Interventions evaluated 9. Main efficacy results 10. Adverse events (including withdrawal due to adverse events) 11. Duration of follow-up 12. Loss to follow-up 13. Compliance to treatment. We assessed internal validity of randomized clinical trials using the eleven criteria proposed by the Cochrane Back Review Group (see Appendix 4 for details on how we operationalized the criteria)56. We rated the internal validity of each trial based on the methods used for randomization, allocation concealment, and blinding; the similarity of compared groups at baseline; the use of co-interventions; compliance to allocated therapy; adequate reporting of dropouts; loss to follow-up; non-differential timing of outcome assessment; and the use of intention-to-treat analysis. Trials were scored between zero and eleven, according to the number of criteria met. For interventions for which blinding was not feasible, we removed blinding of providers (acupuncture, acupressure, neuroreflexotherapy, spinal manipulation, massage, trials of surgery and some interventional therapies), blinding of patients and providers (brief educational interventions, back schools, coordination of care, exercise, hydrotherapy, spa therapy, psychological therapies, interdisciplinary rehabilitation, functional restoration, interventions involving different types of self-care advice, trials comparing an interventional therapy to non-interventional therapy, trials comparing surgery to non-surgical interventions), or blinding of patients and provider and use of co-interventions (trials of different imaging
12
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain strategies) as quality criteria, so the maximum score was ten, nine, or eight, respectively. We considered trials that received more than half of the total possible score to be higher-quality and those that received less than or equal to half lower-quality54, 55.
Observational studies of treatment efficacy

To assess the internal validity of observational studies, we evaluated whether they used nonbiased selection methods; whether rates of loss to follow-up were acceptable; whether predefined outcomes were specified; whether they used appropriate methods for ascertaining exposures, potential confounders, and outcomes; and whether they performed appropriate statistical analyses of potential confounders. Although many tools exist for quality assessment of nonrandomized trials, there is no consensus on optimal quality rating methods57. We therefore did not use a formal scoring system to rate the quality of the observational studies included in this review, but noted methodological deficiencies in any of the above areas when present.
Studies of invasive diagnostic tests

Studies of invasive diagnostic tests (provocative discography, diagnostic facet joint block, medial branch block, diagnostic selective nerve root block, and diagnostic sacroiliac joint block) differ from typical studies of diagnostic test accuracy because there is no clearly accepted reference standard for the conditions these tests are meant to identify. We assessed the quality of these studies using nine criteria adapted from methods developed by the U.S. Preventive Services Task Force58 and on empiric studies59, 60 of sources of variation and bias in studies of diagnostic tests. For each study, we determined if it: 1) evaluated a consecutive series of patients or a random subset, 2) evaluated patients prospectively, 3) evaluated patients with a broad spectrum of symptoms, 4) adequately described the diagnostic test technique, 5) used current diagnostic techniques, 6) adequately described criteria for a positive diagnostic test, 7) used an appropriate definition for a positive diagnostic test, 8) performed statistical analysis on potential predictors or confounders of positive diagnostic tests, and 9) performed testing blinded to patient symptoms and other clinical characteristics. Studies that met at least five of the nine criteria were considered higher-quality.
Dual review
Two reviewers independently rated the quality of each systematic review and primary study. Discrepancies were resolved via a consensus process.
Assessing research applicability and clinical relevance

To assess the applicability of trials, we evaluated whether the publication adequately described the study population and interventions, whether the setting or population was so different from typical U.S. settings that results might not be applicable, whether the differences were clinically (as well as statistically) significant, and whether the treatment received by the control group was reasonably representative of standard practice61. We also recorded funding sources and the roles of the sponsors.
13
EVIDENCE REVIEW
Data synthesis
We assessed the overall strength of evidence for the body of literature, addressing each comparison and outcome evaluated for the Key Questions, using methods adapted from the U.S. Preventive Services Task Force58. To assign an overall strength of evidence (good, fair, or poor) for each comparison and outcome, we examined the type, number, size and quality of studies; strength of association; consistency of results within and between study designs; and directness of evidence. Rating of good quality: Evidence includes consistent results from well-designed, wellconducted studies in representative populations that directly assess effects on health outcomes (at least two consistent, higher-quality RCTs or studies of diagnostic test accuracy). Rating of fair quality: Evidence is sufficient to determine effects on health outcomes, but the strength of the evidence is limited by the number, quality, size, or consistency of included studies; generalizability to routine practice; or indirect nature of the evidence on health outcomes (at least one higher-quality trial or study of diagnostic test accuracy of sufficient sample size; two or more higher-quality trials or studies of diagnostic test accuracy with some inconsistency; at least two consistent, lower-quality trials or studies of diagnostic test accuracy, or multiple consistent observational studies with no significant methodological flaws). Rating of poor quality: Evidence is insufficient to assess effects on health outcomes because of limited number or power of studies, large and unexplained inconsistency between higher-quality trials, important flaws in trial design or conduct, gaps in the chain of evidence, or lack of information on important health outcomes. Consistent results from higher-quality studies across a broad range of populations suggest a high degree of certainty that the results of the studies are true (that is, the entire body of evidence would be considered good-quality). Large effect sizes on important, patient-centered outcomes generally increases confidence in study findings, particularly when they are reported by large, higher-quality studies. For a fair-quality body of evidence, consistent results could be due to true effects or to biases present in some or all studies. Inconsistent results between studies can lower confidence that the results of any particular study are true because of methodological flaws or other issues, or reflect diversity between studies in the populations or interventions evaluated. For a poor quality body of evidence, reliable conclusions are not possible because of insufficient evidence. There is low certainty that the results are not due to bias or other methodologic shortcomings in the studies. To evaluate consistency, we classified conclusions of trials and systematic reviews as positive (the intervention is beneficial), negative (the intervention is harmful or not beneficial), or uncertain (imprecise estimates, unclear evidence, or inconsistent results)51. We defined inconsistency as >25% of higher-quality studies reaching discordant conclusions on efficacy (positive versus negative), two or more higher-quality systematic reviews reaching discordant conclusions, or unexplained heterogeneity (for pooled data). When results between systematic
14
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain reviews or individual trials were discordant, we investigated potential sources of discordance including differences in the populations, interventions, or outcomes addressed and (for systematic reviews) differences in methods for identifying, including, rating and synthesizing the evidence. Sparse data lowers confidence in conclusions from a body of evidence because of imprecise estimates of effects, lack of statistical power, and a greater likelihood that conclusions will be influenced by new evidence. When evaluating low back pain interventions, we defined sparse data as 2 placebo- or active-controlled trials (any sample size), or 3 trials with no trial having >100 subjects. If the body of evidence consisted of a single, small (N<100) study, we rated it poor quality, even if the study itself was rated higher-quality. We also downgraded studies using unvalidated assessment techniques because it is difficult to know how accurately or reliably they estimate the true magnitude of benefit or harm. Primarily relying on indirect evidence, including evidence from patients with other (non-low back) pain conditions or evidence involving indirect comparisons (effect of intervention A versus intervention C estimated from studies comparing intervention A to intervention B and studies comparing intervention B to intervention C) also generally lowers the overall quality of a body of evidence62. In the first stage of this review (focused on non-invasive therapies), data synthesis was primarily based on evaluation of evidence from higher-quality, previously published systematic reviews, supplemented by data from randomized trials not included in the reviews. In the second stage of this review, which focused on interventional therapies and surgery, we modified our approach to base our data synthesis on an independent abstraction and evaluation of placebo- and shamcontrolled randomized trials. We compared our synthesis with conclusions from previously published systematic reviews and evaluated for sources of discrepancy when inconsistency was present.
Assessing magnitude of benefits or harms

Although trials varied widely in how outcomes were assessed and reported, we used prespecified criteria to categorize magnitude of effects for the most commonly reported outcomes. For pain relief, we considered mean differences in effects of 5 to10 points on a 100 point VAS pain scale (or equivalent) as small/modest, 10 to 20 points as moderate, and >20 points as large/substantial. For back-specific functional status, we considered mean improvements in the RDQ of 2 to 5 points or 10 to 20 points on the ODI as moderate45. Mean improvements of >5 points on the RDQ and >20 points on the ODI were considered large/substantial. In order to compare and combine results across trials using different measures for the same outcome (such as pain relief or functional status), some systematic reviews report standardized mean differences (SMD). The SMD permits consistent interpretation across studies because mean differences are adjusted by within-group standard deviations. When SMDs were reported, we considered values from 0.2 to 0.5 small/modest, 0.5 to 0.8 moderate, and >0.8 large/substantial63. Though interpretation of the SMD can vary across different interventions and outcomes, there is some evidence that our classifications for SMDs and changes on pain
15
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain scores and functional status are roughly concordant. In trials of bed rest for low back pain, for example, an SMD between 0.2 and 0.3 was equivalent to 5 to 7.5 points on a 100 point VAS pain scale, and 1.2 to 1.8 points on the RDQ (all classified as small/slight)64, 65. A Cochrane review of spinal manipulation for low back pain estimated an SMD of 0.2 as equivalent to 5 mm on a 100 point VAS pain scale (both classified as small/slight using our system)66, 67 and two different systematic reviews of acupuncture calculated an SMD of 0.5468 and weighted mean difference of 17.8 on a 100 point pain scale69, 70 for the same treatment comparison (both classified as moderate). Because few trials reported the proportion of patients meeting specific thresholds (such as >50% reduction in pain score) for target outcomes, it was usually not possible to report numbers needed to treat or harm. When reported, we considered a relative risk (RR) of 1.25 to 2.00 for the proportion of patients reporting >30% pain relief or improvement in function (or similar outcome) a moderate benefit.
Size of effect Small/slight Definition Pain scales: Mean 5-10 mm improvement on a 100 mm visual analogue scale (VAS), or equivalent Back-specific functional status: Mean 5-10 mm improvement on the Oswestry Disability Index (ODI), 1-2 points on the Roland-Morris Disability Questionnaire (RDQ), or equivalent All outcomes: Standardized mean difference (SMD) 0.2 to 0.5 Pain scales: Mean 10-20 mm improvement on a 100 mm VAS, or equivalent Back-specific functional status: Mean 10-20 mm improvement on the ODI, 2-5 points on the RDQ, or equivalent All outcomes: SMD 0.5 to 0.8 Pain scales: Mean >20 mm improvement on a 100 mm VAS, or equivalent Back-specific functional status: Mean >20 mm improvement on the ODI, >5 points on the RDQ, or equivalent All outcomes: SMD >0.8
Moderate
Large/substantial
RESULTS
Size of literature reviewed
In the first stage of this review (searches performed through November 2006), the literature search for systematic reviews identified 913 citations. Search strategies are shown in Appendix 1. From these citations, we reviewed 265 full-text articles for inclusion; of those, 186 met inclusion criteria. A list of systematic reviews included for this report, along with our quality rating assignments, is shown in Appendix 5. A list of excluded reviews is shown in Appendix 6, along with reasons for exclusion. Main results of included systematic reviews are summarized at the end of Key Questions 3 (self-care therapies) and 4 (non-invasive therapies). We also identified 7591 citations from 44 searches for primary studies. From these citations, 202 primary studies were relevant and met inclusion criteria. Search strategies for primary studies are shown in Appendix 2. A list of included primary studies, along with quality rating assignments, is shown in Appendix 7. In the second stage of this review, we updated searches on interventional therapies and surgery through July 2008. For interventional therapies, the updated literature search yielded a total of 1,331 citations. We retrieved 174 articles based on examination of titles and abstracts. Of 116
16
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain full-text articles that potentially reported a relevant randomized controlled trial, we judged 105 to meet inclusion criteria. Of 58 full-text articles potentially reporting a relevant systematic review, we judged 30 (reporting 26 systematic reviews) to meet inclusion criteria71-100. Main results of included systematic reviews are summarized at the end of Key Question 8. 75 trials (reported in 83 articles) were included in one or more previously published systematic reviews. 22 trials not included in any previous systematic review also met inclusion criteria101-122. Of 97 total trials, 52 (reported in 56 articles) were placebo-controlled (Appendix 5 shows quality ratings)104-106, 108, 112, 123-145113, 114, 117-121, 146-169 . We excluded 28 potentially relevant reviews170-197 (Appendix 6) and twelve trials198-209 of interventional therapies. For surgical interventions, the updated literature search yielded a total of 1,449 citations. We retrieved 125 articles based on examination of titles and abstracts. Of 91 full-text articles that potentially reported a relevant randomized trial, we judged 85 to meet inclusion criteria. Of 36 full-text articles that potentially reported a relevant systematic review, we judged 26 (reporting 24 systematic reviews) to meet inclusion criteria (23 systematic reviews evaluated efficacy 72, 7982, 210-229 and one focused on harms230). Main results of included systematic reviews are summarized at the end of Key Question 9. 62 trials (reported in 71 articles) were included in one or more previously published systematic reviews. Twelve trials (reported in thirteen articles) not included in any previous systematic review also met inclusion criteria110, 231-242. Of 74 total trials, fourteen231, 236-239, 241, 243-251 compared surgery to non-surgical therapy and two252, 253 compared artificial disc replacement to fusion (Appendix 7 shows quality ratings). We excluded 12 potentially relevant reviews176, 177, 190, 191, 254-261 (Appendix 6) and four trials262-265 of surgery.
RESULTS
Key Question 1a How accurate are features of the history and physical exam for predicting presence of serious underlying conditions (red flags) or other conditions that may be responsive to specific therapies in patients with low back pain (such as nerve root compression or spinal stenosis)?
In primary care, about 0.7% of patients will have spinal malignancy (primary or metastatic), 4% compression fractures, and 0.01% spinal infection266. Estimates for prevalence of ankylosing spondylitis in primary care patients range from 0.3%266 to 5%267. Spinal stenosis and symptomatic herniated disc are present in about 3% and 4%, respectively268. Up to 90% of patients have non-specific low back pain, for which there is imprecise or poor correlation with any specific pathology268. Features of history and physical exam that can identify patients more likely to have serious conditions such as cancer or infection (red flags) or other conditions that may respond to specific treatments (such as nerve root compression from lumbar disc prolapse, spinal stenosis, ankylosing spondylitis, and vertebral compression fracture) are important for guiding diagnosis and therapy.
17
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: systematic reviews We identified five systematic reviews (four higher quality269-272, one lower-quality268) on the accuracy of history and physical exam for diagnosing various conditions associated with low back pain. We excluded three systematic reviews that were outdated273, did not clearly describe systematic methods for identifying or synthesizing the literature273, did not report diagnostic accuracy274, or reported duplicate information from another published review275. Studies of spinal palpatory maneuvers (to identify patients likely to benefit from manipulation) and physical exam maneuvers for sacroiliac joint pain are discussed in Key Question 5. Results of search: primary studies All of the systematic reviews noted important methodological shortcomings in the primary literature, such as spectrum bias (for example, only evaluating patients who underwent surgery, patients from referral settings, or those with more severe disease), little attention to inter- or intra-rater reliability, verification bias, non-blinded assessment of the index or reference tests, poor description of the index test, and lack of attention to reproducibility of findings over time. These deficiencies could explain some of the observed variation between studies in reported diagnostic accuracy. Another limitation of the literature is that the specific features of history and physical exam that were assessed varied, and for several features only a single or few studies are available. Only one systematic review (rated higher-quality), on the accuracy of the straight leg raise test for disc herniation, pooled data quantitatively270. We did not search for additional studies.
Accuracy of history and physical exam features for identifying specific diagnoses associated with low back pain Cancer
Two systematic reviews evaluated diagnostic accuracy of clinical history for identifying patients with cancer268, 271. Based on one higher-quality study266, both systematic reviews found failure to improve after 1 month of therapy, unexplained weight loss, and previous history of cancer each associated with high specificity (>0.90) for cancer. Previous history of non-skin cancer was associated with the highest positive likelihood ratio at 14.7, increasing the post-test probability of cancer from about 0.7% to 9%. Only age >50 years and no relief with bed rest were associated with sensitivities greater than 0.50 (0.77 and >0.90, respectively). Having any of the following was associated with a sensitivity of 1.0 and specificity of 0.60 for diagnosing vertebral cancer: age >50, history of cancer, unexplained weight loss, or failure of conservative therapy (positive likelihood ratio 2.5). For physical exam findings, one systematic review found the sensitivity of spinal tenderness for vertebral cancer varied widely across four studies (range 0.15 to 0.80), though specificity was relatively consistent (0.60 to 0.78)271. Other physical exam findings (such as muscle spasm, radiculopathy, Babinskis sign, or urinary retention) had poor sensitivity, though certain neuromuscular (weakness, atrophy, reflex changes) or sensory deficits were associated with high specificity in some studies.
18
EVIDENCE REVIEW
Infection
Few studies evaluated accuracy of history for diagnosing spinal osteomyelitis or other infections causing low back pain. One systematic review found a sensitivity of 0.40 for a history of intravenous drug abuse, urinary tract infection, or skin infection (specificity not reported)268.
Cauda equina syndrome

Cauda equina syndrome is most commonly caused by massive midline intervertebral disc herniation. Though there is little data on accuracy of history and physical exam for identifying patients with this condition, the most frequent finding in cauda equina syndrome is urinary retention (sensitivity 90%)273. In patients without urinary retention, the probability of cauda equina syndrome is approximately 1 in 10,000.
Compression fracture
For diagnosis of compression fracture, one systematic review included one unpublished study that found corticosteroid use associated with a higher predictive value (positive likelihood ratio 12.0) than age or history of trauma268. Age >50 years was associated with a sensitivity of 0.84 and specificity of 0.61 (positive likelihood ratio 2.2 and negative likelihood ratio 0.26) and age >70 years was associated with a sensitivity of 0.22 and specificity of 0.96 (positive likelihood ratio 5.5 and negative likelihood ratio 0.81).
Ankylosing spondylitis
Two systematic reviews evaluated diagnostic accuracy of history for identifying patients with ankylosing spondylitis268, 271. Both found younger age of onset associated with high sensitivity but poor specificity (sensitivity and specificity 0.92 and 0.30 for onset <35 years, 1.00 and 0.07 for onset <40 years). Most other historical features had only modest predictive value or gave inconsistent results. For example, the specificity of a history of sacral pressure varied from 0.68 to 0.92 in three studies. Combined historical findings (positive response to 4 of 5 of the following screening questions: onset before age 40, chronic onset, duration >3 months, morning stiffness, and improvement with exercise) did not improve diagnostic accuracy (positive likelihood ratio of 1.3 and negative likelihood ratio of 0.94). All physical exam findings (including Schobers test, degree of chest expansion, reduced lateral mobility, and sacral or lumbar pressure) were associated with poor sensitivity. In single studies, chest expansion 2.5 cm, Schobers sign <4 cm, and restricted anteroposterior compression, lateral compression, or hip extension were associated with relatively high specificities (>0.80). One recent study found a positive likelihood ratio of 3.7 for inflammatory low back pain associated with ankylosing spondylitis when at least two of the four criteria are met: morning stiffness of >30 minutes duration, improvement in back pain with exercise but not with rest, awakening because of back pain during the second half of the night only, and alternating buttock pain276. The positive likelihood ratio increased to 12.4 when at least 3 parameters are met.
19
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Revised criteria for diagnosing early ankylosing spondylitis (prior to the development of sacroiliac changes on imaging studies) have recently been proposed277. Their adoption is likely to affect estimates of diagnostic accuracy of history and physical exam findings for ankylosing spondylitis.
Herniated disc or radiculopathy

For diagnosing a herniated disc or radiculopathy, three systematic reviews found a history of sciatica had fairly high (79% to 99%) sensitivity and widely varying specificity (14% to 88%)268, 271, 272 . One systematic review also included one higher-quality study that found a typical distribution for radiculopathy on a pain drawing had modest sensitivity (46%), but high specificity (84%)272. The best-evaluated physical exam findings for herniated disc are the straight leg raise (Laseagues test) and the crossed straight leg raise tests. In a higher-quality, recent systematic review of 17 studies, the pooled sensitivity and specificity of the straight leg raise test for diagnosing herniated disc were 0.91 (95% CI 0.82 to 0.94) and 0.26 (95% CI 0.16 to 0.38)270. The pooled diagnostic odds ratio was 3.74 (95% CI 1.2 to 11.4). For the crossed straight leg raise test, the pooled sensitivity and specificity were 0.29 (95% CI 0.24 to 0.34) and 0.88 (95% CI 0.86 to 0.90), with a pooled diagnostic odds ratio of 4.39 (95% CI 0.74 to 25.9). Three other systematic reviews reached similar conclusions268, 271, 272. Other physical exam findings (such as decreased reflexes, strength, muscle atrophy, or sensory deficits) have been less well studied. In general, the presence of such neurological deficits is an insensitive finding for diagnosing radiculopathy or herniated disc268, 271, 272. Isolated studies found iliopsoas or tibialis anterior weakness associated with high (97% and 89%) specificity272. The specificity of gastrocnemius weakness, calf atrophy, and depressed ankle or knee jerks for diagnosing herniated disc ranged from slight to high268, 271, 272. The accuracy of combined history and physical examination findings to diagnose herniated disc varied across studies, in part because of inconsistencies in how the clinical findings were defined across studies272. For example, the sensitivity and specificity were 27% and 97% in one study that defined a positive cluster as two or more positive findings278, but 98% and 7% in another that defined a positive cluster as probable diagnosis based on clinical exam and history279.
Spinal stenosis
One recent, higher-quality systematic review found limited evidence (7 studies, 2 rated higherquality) on diagnostic accuracy of history and clinical findings or tests for spinal stenosis269. In the two higher-quality studies, the presence of radiating leg pain (sensitivity 94%) and changes in neurologic status on a downhill walking treadmill test (sensitivity 100%) were associated with the highest sensitivity, but neither finding was specific (21% and 33%, respectively)280, 281. Findings that were >80% specific (such as changing symptoms, bilateral paresis, or bilateral reflex changes on treadmill testing) were not sensitive (38% to 63%). The highest positive predictive value (3.1) was associated with changing symptoms during downhill treadmill testing.
20
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Pseudoclaudication and radiating leg pain were associated with positive likelihood ratios of 1.2 and 2.2, respectively. In one lower-quality study, lack of pain when seated and a wide-based gait were associated with positive likelihood ratios of 6.6 and 14.3 (respectively)282, though the positive likelihood ratio for pain relieved by sitting was only 0.96 in another lower-quality study283. This study also found age greater than 65 years associated with a positive likelihood ratio of 2.5 and negative likelihood ratio of 0.33. In another lower-quality study, a combination of factors on two stage treadmill test based on time to onset of symptoms and recovery time was associated with a positive predictive value of 15, but this finding has not yet been replicated283.
Summary of evidence
For diagnosis of cancer in primary care patients with acute low back pain, previous history of non-skin cancer (positive likelihood ratio 14.7), unexplained weight loss (positive likelihood ratio 2.7), and failure to improve after 1 month of therapy (positive likelihood ratio 3.0) were each associated with specificity >0.90. In a primary care setting, a history of non-skin cancer increased the likelihood of cancer from about 0.7% to 9% in one study (level of evidence: fair). For diagnosis of cancer in primary care patients with acute low back pain, the presence of any of the following was associated with a high sensitivity (1.00) and moderate specificity (0.60) in one higher-quality study: age >50 years, history of non-skin cancer, unexplained weight loss, or failure of standard non-invasive therapy (positive likelihood ratio 2.5, negative likelihood ratio 0.0). Physical exam findings (such as vertebral tenderness or neurologic deficits) generally have poor or inconsistent sensitivity for cancer, but high specificity in some studies (level of evidence: fair). For diagnosis of infection in patients with low back pain, few studies evaluated the accuracy of history and physical exam, though history of intravenous drug use, skin infection, or urinary tract infection only had modest sensitivity in one study (level of evidence: poor). For diagnosis of vertebral compression fracture, older age and history of corticosteroid use were the best predictors (level of evidence: fair). For diagnosis of ankylosing spondylitis, younger age at onset of back pain was sensitive but not specific. Physical exam findings for ankylosing spondylitis were generally associated with poor sensitivity and relatively high specificities. Presence of at least two of the following was associated with a positive likelihood ratio of 3.7: morning stiffness of >30 minutes duration, improvement in back pain with exercise but not with rest, awakening because of back pain during the second half of the night only, and alternating buttock pain (positive likelihood ratio of 12.3 when at least 3 findings present). Recently proposed changes in criteria used to diagnose early ankylosing spondylitis (i.e. prior to the development of sacroiliitis) are likely to affect estimates of diagnostic accuracy (level of evidence: fair). For diagnosis of radiculopathy, describing typical symptoms of sciatica has a relatively high sensitivity but inconsistent specificity. A positive straight leg raise (the best-studied physical exam maneuver) was associated with a pooled sensitivity of 0.91 and specificity of 0.26 in one higher-quality systematic review. A positive crossed straight leg raise was associated with a
21
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain pooled sensitivity of 0.29 and a specificity of 0.88. The specificity of neurologic deficits consistent with nerve root compression ranges from modest to high (level of evidence: fair). For diagnosis of spinal stenosis, higher-quality studies found features of the history and clinical exam associated with high sensitivity (such as radiating leg pain) generally associated with low specificity, or vice versa, resulting in modest or poor predictive values. Changing symptoms on downhill treadmill testing was associated with the highest positive likelihood ratio (3.1). In lower-quality studies, a wide-based gait and a combination of findings on treadmill testing were associated with higher likelihoods for spinal stenosis. Age greater than 65 years was associated with a positive likelihood ratio of 2.5 and negative likelihood ratio of 0.33. The predictive value of pain relieved by sitting ranged from poorly to highly (level of evidence: fair).
Recommendations and findings from other guidelines

The AHCPR guidelines recommend inquiring about features of the history and clinical exam suggestive of cancer or infection (history of cancer, unexplained weight loss, intravenous drug use, history of urinary infection, pain increased by rest, fever), particularly in patients over the age of 50 (strength of evidence: B). The AHCPR guidelines recommend inquiring about features suggestive of cauda equina syndrome such as bladder dysfunction, saddle anesthesia, and major limb motor weakness (strength of evidence: C). The AHCPR guidelines recommend inquiring about significant trauma or minor fall or heavy lift in potentially osteoporotic or older patients to avoid delays in diagnosing fractures (strength of evidence: C). The AHCPR guidelines recommend straight leg raise testing to assess sciatica in young adults, but notes that it may be normal in older patients with spinal stenosis (strength of evidence: B). The AHCPR guidelines recommend a focused neurologic exam emphasizing ankle and knee reflexes, ankle and great toe dorsiflexion strength, and distribution of sensory complaints to document the presence of neurologic deficits (strength of evidence: B). The VA/DoD and UK RCGP guidelines adopted an approach nearly identical to the one suggested by the AHCPR guidelines (history and physical with focus on identifying red flags and focused neurologic examination). The European COST guidelines also recommend diagnostic triage in patients with acute low back pain as recommended by other guidelines. In patients with chronic low back pain, diagnostic triage is recommended at the first assessment and at reassessment to exclude specific spinal pathology and nerve root pain.
22
EVIDENCE REVIEW
Key Question 1b How accurate are features of the history and physical exam for predicting the development of persistent low back pain and associated disability (yellow flags)?
Rapid improvements in low back pain typically occur in the first month after presentation. However, a small proportion of patients develop chronic and disabling back pain, and many patients continue to have back pain at lower intensity or recurrent low back pain episodes11-13. One systematic review found that 82% of those initially off work returned to work within one month, and 93% had returned to work by three to six months, with little subsequent improvement11. Yellow flags describe features of the history or physical examination that could help identify patients more likely to develop chronic and disabling low back pain. This identification can be used in order to provide interventions that might help retain or improve functionality. Results of search: systematic reviews We identified 14 systematic reviews on features of the clinical history or physical exam predictive of a high risk for persistent low back pain and related disability. Thirteen systematic reviews evaluated prognostic factors based on clinical history11, 284-295. Five were rated higherquality11, 286, 287, 291, 294. Four lower-quality systematic reviews evaluated prognostic factors based on the physical exam284, 293, 295, 296. Results of search: additional studies All of the systematic reviews reported important methodological shortcomings in the primary literature evaluating prognostic factors for low back pain including lack of blinding, small sample sizes, inadequate analyses of confounders, and incomplete follow-up of patients. In addition, the populations and settings were heterogeneous. Due in part to these limitations, only one systematic review quantitatively pooled trials294. We did not search for additional studies.
Accuracy of history and physical exam features for identifying patients more likely to develop chronic and disabling low back pain
The most comprehensive, higher-quality systematic review (based on 54 studies meeting minimum methodologic criteria) found strong evidence that each of the following was a predictor for persistent low back pain, non-return to work, or disability: low back pain associated with increased pain severity, longer duration of symptoms, associated disability, or leg pain; low back pain-related sick leave; history of spinal surgery; low job satisfaction; and poor general health (Table 1)286. There was moderate evidence that work-related and psychological factors (such as employment status, amount of wages, workers' compensation, and depression) and physical factors (such as time spent lifting per day and work postures) were also associated with worse outcomes. Findings of other systematic reviews were generally concordant. For example, a second higher-quality systematic review of 18 prospective cohort studies (six rated at least acceptable quality) found increased psychological distress, somatization, and poorer coping strategies associated with unfavorable outcomes291. Several systematic reviews found receipt of
23
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain benefits or workers compensation associated with poorer outcomes284, 290, 295. Other systematic reviews found modest evidence for an association between more severe pain290, 295, presence of radiating pain284, 293, 294 or presence of continuous pain293 and poorer outcomes. Evidence regarding associations between age or gender and poorer outcomes was mixed and inconsistent284, 290, 293, 294. Only a handful of studies assessed the usefulness of specific scales to predict poorer outcomes. One recent higher-quality systematic review found that the Vermont disability prediction questionnaire appeared promising11. Higher scores on the Vermont prediction questionnaire (>0.48) were associated with a positive likelihood ratio for return to work at 3 months of 5.7 (95% CI 3.9 to 8.5) and a negative likelihood ratio of 0.07 (95% CI 0.01 to 0.50) in one higherquality study297. Fear avoidance (avoidance of activity because of fears that it will worsen symptoms or outcomes) predicted worse outcomes in two284, 293 of four291, 292 systematic reviews. Evidence on the prognostic value of physical exam findings for prediction of poorer outcomes associated with low back pain is sparser than evidence regarding psychosocial factors. Presence of positive sham tests for pain (such as Waddells nonorganic signs) consistently predicted disability in three studies included in one systematic review293, though a more recent study found Waddells signs and symptoms inaccurate for predicting delayed return to work298. Other physical exam findings such as positive straight leg raise tests, absence of neurological signs, and intact range of motion were inconsistently associated with poorer outcomes284, 296. One systematic review found physical exam findings to be weaker predictors of outcomes than psychosocial factors295. Main results of the 14 systematic reviews on prognostic factors for low back pain are summarized in Table 1.
24
EVIDENCE REVIEW
Table 1. Systematic reviews of prognostic factors for identifying patients more likely to develop chronic and disabling low back pain
Number of included studies 10
Author, year Borge, 2001296
Prognostic factors evaluated Physical examination tests and physical examination observations
Crook, 2002284
19 studies of prognostic factors for low back pain
Grouped into categories: sociodemographics, medical/physical, history of back pain, pain, psychological, social/family, functional disability, health status, workplace, lifestyle, compensation, intervention Formal education
Dionne, 2001285
18 studies of prognostic factors for low back pain 54
Fayad, 2004286
Individual factors (medical/demographic, clinical exam, psychological characteristics, socio-cultural factors), professional factors (socio-professional and physical)
Main results Range of motion tests: 3 of 8 (or 3 of 9) studies found that lumbar range of motion tests predicted outcomes Nerve root tension tests: 1 study found no predictive value Neurologic symptoms (reflexes and sensitivity): 1 study found no predictive value Painful spots in the lumbar area: 1 study found no predictive value Palpation of spinous processes: 1 study found no predictive value 7 different tests: 1 study found no predictive value Spine-hip ratio or hip flexion: 1 study found no predictive value McKenzie protocol (centralizer vs. noncentralizer): 1 study found that protocol predicted outcome for self-reported pain intensity or return to work, but not for treatment outcome as measured by Oswestry scale or lifting capacity Predictors of slower return to work: psychological distress (1 study), older age and/or female gender (4), functional disabilities (4), job problems or problems with colleagues (3), previous hospitalization (1), previous episode of back pain (1) Predictors of faster return to work: availability of modified jobs (1), light mobilization (1), more than 2 years on the job or referral to occupational injury, and less than 30 days from injury to treatment (1), no pain (1), no sprain (1), good flexion (1), absence of neurological signs (1) Mixed results: workers compensation status (1 study negative predictor of return to work and 1 study positive predictor) Education as a predictor of outcomes of low back pain episodes (11 'major' studies): Worse outcomes significantly associated with low education for 20 outcomes in 11 studies, negative results for 5 outcomes in 2 studies, and no studies reported worse outcomes among better educated. Predictive factors for recurrence, chronicity of low back pain, and non-return to work: Strong evidence: History of low back pain (including pain severity, increased duration, associated disability, leg pain, related sickness leave, and history of spinal surgery), low job satisfaction, and poor general health. Moderate evidence: Socioprofessional and psychological factors (employment status, amount of wages, workers' compensation, and depression) and physical factors (lifting time per day and work postures)
Quality* 2/7
4/7
4/7
4/7
EVIDENCE REVIEW
Number of included studies 22 studies of prognostic factors for low back pain
Author, year Hartvigsen, 2004287
Prognostic factors evaluated 30 different psychosocial variables grouped into the categories: perception of work, organizational aspects of work, social support at work, and stress at work
Kuijer, 2006288
17
Linton, 2000289
McIntosh, 290 2000
16 studies of prognostic factors for low back pain 9
Sociodemographic factors, lifestyle, medical history, pain, observed disability, self reported disability, health belief, physical work demands, psychological work demands, emotions, and expectations Psychosocial factors (variously defined) Pain measurements, functional status, age, gender, occupational and/or industry measures Vermont disability prediction questionnaire; other factors not reported
Main results Predictive factors for 'consequences' of low back pain Perception of work: 3 of 6 higher-quality studies reported moderate positive associations (OR range 1.20 to 1.95), 3 of 13 lower-quality studies reported moderate positive associations (OR range 1.53 to 1.87) (insufficient evidence) Organizational aspects of work: 0 of 4 higher-quality studies reported positive associations; 2 of 5 lower-quality studies reported moderate positive associations (OR range 1.40 to 1.79) (strong evidence for no association) Social support at work: 2 of 5 higher-quality studies reported strong positive associations (OR range 3.40 to 5.75); 0 of 4 lower-quality studies reported positive associations (moderate evidence for no association) Stress at work: 0 of 3 studies (1 higher-quality) reported a positive association ) (moderate evidence for no association) Consistent evidence for patient expectations of recovery as a predictor for return to work. No other factors consistently predicted sickness absence or return to work.
Quality* 6/7
4/7
Pengel, 200311
6 studies of prognostic factors for low back pain
Acute or subacute pain (16 studies of LBP): 15 studies found a significant link between a psychological variable (including stress, family factors, coping, depression, avoidance, pain fear-avoidance, somatization, catastrophizing, hysteria) and outcome (Level A evidence) Pain measurements predictive in 3 studies, functional status predictive in 1 study, age predictive in 5 studies, gender predictive in 5 studies (2 found females slower to recover, 1 found opposite results), occupational and/or industry measures (not defined) predictive in 5 studies (including delayed working seen in construction workers (3 studies), benefits predictive in 1 study, Vermont disability prediction questionnaire (1 methodologically strong study), score >0.48: Odds ratio for return to work at 3 months 76.3 (95% CI 9.6 to 604.9), positive likelihood ratio 5.7 (95% CI 3.9 to 8.5), negative likelihood ratio 0.07 (95% CI 0.01 to 0.50) Other prognostic factors (not specifically stated) (2 studies): Odds ratios ranged from 0.04 to 10.4
3/7
4/7
5/7
EVIDENCE REVIEW
Author, year Pincus, 2002291
Prognostic factors evaluated Psychological distress/depressed mood, somatization, personality, and cognitive factors
Pincus, 2006292 Shaw, 2001293
Fear, fear avoidance, catastrophizing Age, gender, work (occupation, employer size), injury, symptom (pain vs. function), clinician exam (range of motion, nonorganic signs), psychological (job satisfaction, pain beliefs)
22
Main results Distress (8 studies, 4 rated unacceptable): Defined as composite of psychological distress, depressive symptoms, or depressed mood. Predictor of unfavorable outcome, especially in primary care (OR: approx 3; 2 high, 2 acceptable quality studies). Somatization (4 studies, 2 rated unacceptable): 1 high and 1 acceptable quality study found somatization scales to predict unfavorable outcome. Cohen's effect size statistic (d) 0.2 to 0.6 at 1 year and 0.9 at 2 year follow-up. Personality (3 studies, 2 rated unacceptable): In 1 acceptable quality study, the hysteria subscale of the MMPI was reported to be a predictor of return to work (OR 1.5), but this was considered statistically unreliable. Cognitive factors (6 studies, 5 rated unacceptable): 1 acceptable quality study found subscales from the Coping Strategies Questionnaire predicted unfavorable outcome: effect size 1.09 for praying/hoping, 1.88 for catastrophizing. Fear avoidance not significant in 1 study when entered into multivariate model. Fear avoidance: None of the studies that measured fear avoidance provided convincing evidence that fear-avoidance beliefs are a risk factor for poor outcomes; 6 of 9 studies failed to show a statistically significant association or only a week association. Factors predicting disability Age (16 studies): 8 supporting studies, 8 non-supporting Gender (16 studies): 12 non-supporting studies, 3 studies found females had slower recovery, 1 study found males had slower recovery Marital status (5 studies): 3 non-supporting studies Work environment (worker perceptions of coworker cohesion, problems with coworkers, social isolation, trouble at work) (3 studies): 3 supporting studies Occupation/industry (9 studies): 6 supporting studies (4 of 6 studies found construction associated with longer disability compared to other 'blue-collar' workers) Physical demands (11 studies): 5 supporting studies found association with worker selfreport (not objective measures) Tenure (6 studies): 2 supporting studies (newer employees) Greater work satisfaction (6 studies): 1 supporting study Salary (4 studies): 2 supporting studies Injury type (5 studies): 4 supporting studies Functional and overall clinical assessment (8 studies): 8 supporting studies, but substantial variation in types of functional tests related to prolonged worse absence. Consistent predictors of disability were Waddell nonorganic signs and other sham tests of pain (3 studies).
Quality* 5/7
4/7
3/7
EVIDENCE REVIEW
Author, year Steenstra, 2005294
Truchon, 295 2000
18
Prognostic factors evaluated 69 different prognostic factors related to characteristics of current episode, workers' health, psychosocial factors, work characteristics, and work organization Medical factors, ergonomic and psychosocial work-related factors, psychosocial factors not related to work, sociodemographic variables
Main results Longer duration of sick leave associated with: radiating low back pain, higher disability levels, older age, female gender, more social dysfunction and isolation, heavier work, and higher compensation. Not associated with duration of sick leave: history of low back pain, job satisfaction, educational level, marital status, number of dependents, smoking, working more than 8 hour shifts, occupation, and size of industry or company. "Promising" predictors of no return to work: previous history of low back pain, results of certain clinical tests, a subjective negative self-appraisal of ability to work, and job dissatisfaction.
Quality* 6/7
2/7
*Using Oxman criteria, maximum score 7 on a 1 to 7 scale
EVIDENCE REVIEW
Summary of evidence
There is consistent evidence from multiple systematic reviews that psychological distress or depression, impaired function, job dissatisfaction, high levels of fear avoidance beliefs, disputed compensation claims, and somatization are associated with worse low back pain outcomes (level of evidence: good). Increased duration or severity of pain and presence of leg pain are modestly associated with poorer outcomes (level of evidence: fair). Physical exam findings were inconsistently associated with outcomes and are weaker predictors of unfavorable outcomes than psychosocial factors (level of evidence: fair). Evidence on validated tools or scales for identifying patients likely to have poorer outcomes is sparse, though one study found the Vermont disability questionnaire promising (level of evidence: poor).

The AHCPR guidelines recommend inquiring about psychological and socioeconomic problems, as nonphysical factors can complicate assessment and treatment (strength of evidence: C). The AHCPR guidelines found that social, economic, and psychological factors can significantly alter a patients response to back symptoms and to treatment of those symptoms (strength of evidence: D). The VA/DoD, UK RCGP and European COST guidelines also recommend assessing psychological and socioeconomic factors and reviewing them if there is no improvement.
Key Question 1c How effective is identification and treatment of yellow flags for improving clinical outcomes in patients with low back pain?
Results of search: systematic reviews We found no systematic review on effects of interventions for identification and treatment of yellow flags in patients with acute or subacute low back pain. Although several systematic reviews evaluated interventions that addressed psychosocial issues in patients with subacute or mixed duration low back pain, identification and treatment of yellow flags was usually not the main goal of therapy or was included as part of an interdisciplinary biopsychosocial approach (see discussions of psychological, interdisciplinary, and functional restoration interventions in Key Question 4)299-303. Results of search: trials We identified two higher-quality trials on brief interventions for identifying and treating yellow flags304, 305. A third, higher-quality trial evaluated an intensive, interdisciplinary intervention in patients identified as higher-risk for developing chronic back pain with disability306. Two other trials (one lower-quality307) evaluated efficacy of fear-avoidance based therapy307, 308. All trials were conducted in patients with acute or subacute low back pain. We excluded two trials of
29
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain cognitive-behavioral interventions in patients who perceived themselves to be at high risk for developing chronic problems because they included all types of spinal pain (neck, upper back, and lower back) and did not clearly specify duration of symptoms309, 310.
Efficacy of interventions for identifying and treating yellow flags

Although several recent trials assessed interventions for identification and treatment of yellow flags in patients with low back pain, it is difficult to draw general conclusions about their effectiveness because of differences in the treatments (ranging from brief interventions administered by a primary care clinician to intensive, interdisciplinary interventions) and populations studied. Two higher-quality trials found brief interventions no more effective than standard practice or conventional physical therapy in patients with back pain of less than 12 weeks duration (Table 2)304, 305. One trial (n=314) found no differences through 12 months between usual care and a minimal (20 minute) intervention aimed at identifying, providing information about, and promoting self-care of psychosocial risk factors for any outcome including back-specific functional status (RDQ score), pain, sick leave, perceived general health (SF-36), or general practitioner visits305. The minimal intervention also failed to show a benefit in higher-risk subgroups of patients with increased baseline psychological distress or recurrent back pain. The second trial (n=402) found no differences on back-specific functional status (ODI score), pain, time off work, depression scores, use of health care resources, or satisfaction with care after either 3 or 12 months among patients randomized to a brief pain management program (aimed at identifying psychosocial risk factors, emphasizing return to normal activity through functional goal setting, and using educational strategies to overcome psychosocial barriers to recovery as well as a tailored exercise program) versus a physical therapy intervention (with an emphasis on spinal manipulation)304 The number of physical therapy sessions was slightly lower with the brief intervention. All patients improved regardless of which treatment they were randomized to.
30
EVIDENCE REVIEW
Table 2. Trials of brief interventions for identifying and treating yellow flags
Number of patients Duration of followup n=402 12 months
Author, year Hay, 2005304
Jellema, 2005305
n=314 12 months
Main results Brief pain intervention vs. manual physical therapy (results at 12 months unless otherwise noted) ODI score, mean change from baseline: 7.8 vs. 8.1 at 3 months, p=0.755; 8.8 vs. 8.8 at 12 months, p=0.994 Overall assessment 'much better' or 'completely better' at 12 months: 68% vs. 69% Back pain (0 to 100 scale): 78 vs. 70, p=0.401 Took time off work in last 12 months: 54% vs. 58%, p=0.45 Satisfaction with treatment (0 to 100 scale), median: 93 vs. 93 Minimal intervention vs. usual care (results at 12 months unless otherwise noted) RDQ score (0 to 24 scale): 1 vs. 1, mean difference 0.25 (0.77 to 1.28) No recovery (rated recovery as slightly improved, no change, slightly worse, much worse, or very much worse): 42/132 (32%) vs. 43/156 (28%), odds ratio 1.16 (0.63 to 2.17) Sick leave due to low back pain: 8/107 (8%) vs. 9/128 (7%), odds ratio 0.69 (0.43 to 1.13) Pain severity: mean difference 0.015 (-0.41 to 0.44)
Quality score* 7/9
6/9
*Excludes criteria involving blinding of patients and care providers, for maximum score of 9
Several factors could explain the lack of an effect in these two trials. In one study, patients randomized to the minimal intervention were not permitted to receive physical therapy for the first six weeks305. In addition, general practitioners randomized to the minimal intervention arm were only moderately successful in identifying psychosocial factors, and were no more effective than practitioners randomized to usual care in improving outcomes measured by psychosocial scales311. It is possible that additional training or a more intense intervention could result in more effective treatment. In addition, targeting the intervention to high-risk patients could improve outcomes compared to treating a less selected group of patients312. These hypotheses are supported in part by a third, small (n=70), higher-quality trial which found a more intense (including three physician evaluations and a total of up to 45 physical therapy, biofeedback/pain management, group didactic, and case manager/occupational therapy sessions), interdisciplinary functional restoration intervention associated with improved pain and decreased disability after 12 months (Table 3) compared to usual care in patients with acute (<8 weeks) low back pain identified as being at higher risk for chronic disability using a screening tool306.
31
EVIDENCE REVIEW
Table 3. Trials of intensive multidisciplinary functional restoration in patients at higher risk for chronic disability
Number of patients Duration of follow-up n=70 12 months
Author, year Gatchel, 2003306
Main results Multidisciplinary functional restoration vs. usual care Return to work at 12 months: 91% vs. 69% (p=0.027) Average number of healthcare visits: 26 vs. 29 (p=0.004) Average number of healthcare visits related to low back pain: 17 vs. 27, p=0.004 Average number of disability days due to back pain: 38 vs. 102, p=0.001 Average most intense pain" at 12 month follow-up: 46 vs. 67, p=0.001 Average self-rated pain over last 3 months: 27 vs. 43, p=0.001 Taking opioid analgesics: 27% vs. 44%, p=0.020
Quality score* 6/9
Two other trials evaluated interventions aimed at reducing fear avoidance behaviors (Table 4). In one lower-quality trial, 240 patients with persistent low back pain and activity limitations 8 to 10 weeks after the initial visit were randomized to four sessions of an individualized fear avoidance intervention with a psychologist and physical therapists versus usual care307. The fear avoidance intervention was superior for disability outcomes, with the proportion of patients experiencing a greater than one-third reduction in RDQ score: 28% vs. 13% at 2 months (p=0.0007) and 49% vs. 37% at 24 months (p=0.08). Average pain intensity was slightly better in patients randomized to the intervention after two months, though the difference was no longer significant at 24 months. There was no difference in SF-36 scores or ability to work, though a lower proportion of patients randomized to the fear avoidance intervention reported activity limitations due to back pain for 30 or more days after 24 months (8.5% vs. 14.3%, p=0.04). Patients randomized to the fear avoidance intervention also reported lower scores on fearavoidance and worry rating scales. The second, smaller (n=67), higher-quality trial found no differences on the ODI scale or pain intensity after 6 months between low back pain (less than 8 weeks duration) patients randomized to fear avoidance-based physical therapy (encouraging patient to take an active role in treatment and to view back pain as common, along with a selfcare booklet and graded exercise) and standard exercise308. The fear avoidance intervention was associated with lower fear avoidance beliefs in the subgroup of patients with high baseline fear avoidance scores.
32
EVIDENCE REVIEW
Table 4. Trials of fear-avoidance based interventions

Author, year George, 2003308
Von Korff, 2005307
n=240 24 months
Main results Fear avoidance exercise program vs. standard exercise ODI score (0 to 100), mean change: 18.0 vs. 17.1 at 4 weeks (NS), 23.9 vs. 23.0 at 6 months (NS) Present pain intensity (0 to 10), mean change: 2.4 vs. 2.0 at 4 weeks (NS), 2.6 vs. 3.0 at 6 months (NS) Fear Avoidance Beliefs Questionnaire, Physical Activity Scale (0 to 24), mean change: 5.0 vs. 1.8 at 6 months, p=0.037 Fear Avoidance Beliefs Questionnaire, Work Scale (0 to 42), mean change: 3.1 vs. 1.9 at 6 months, p=0.352 Fear avoidance intervention vs. usual care RDQ score (0 to 24): 10.2 vs. 11.5 at 2 months, p=0.0002; 8.1 vs. 9.1 at 24 months, p=0.0078 Proportion of patients with greater than one-third reduction in RDQ score: 28% vs. 13% at 2 months, p=0.0007; 49% vs. 37% at 24 months, p=0.08 Fear-avoidance (17-68): 36.4 vs. 39.9 at 2 months, p<0.0001; 34.3 vs. 38.4 at 24 months, p=0.0001 Average pain intensity (0 to 10): 4.9 vs. 5.3 at 2 months (p=0.020); 4.3 vs. 4.6 at 24 months (p=0.115) SF-36 social functioning and SF-36 mental health inventory: no differences Unable to work: No differences Unable to carry out usual activities due to back pain for 30 or more days: 24% vs. 26% at 2 months, p=0.06, 8.5% vs. 14.3% at 24 months, p=0.04
Quality score* 7/9
4/9
Harms No trial reported harms. Costs A cost-benefit analysis of the trial comparing an intensive, early multidisciplinary intervention in patients identified as higher risk for chronic disability calculated a net gain of $9,122, mostly related to a reduction in lost wages in the intervention group306.
Summary of evidence
In unselected patients with acute or subacute low back pain, two higher-quality trials found no benefits after 12 months from brief interventions designed to identify and treat yellow flags compared to usual care or physical therapy with an emphasis on manipulation or mobilization (level of evidence: good). In patients with back pain for less than 8 weeks identified as being at higher risk for chronic disability using a screening tool, one higher-quality trial found an intensive interdisciplinary functional restoration program more effective than usual care after 12 months (level of evidence: poor).
33
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain In patients with persistent activity limitations due to low back pain, one lower-quality trial found fear-avoidance based therapy slightly superior to usual care for back specific functional status after 24 months, though beneficial effects on pain were only short-lived (level of evidence: poor). For subacute (<8 weeks) low back pain, one higher-quality trial found no difference between fear-avoidance therapy and standard physical therapy after 6 months, though fear-avoidance beliefs were decreased in the intervention group (level of evidence: fair).
Recommendations and findings of other guidelines

No guidelines make recommendations about specific interventions in patients with acute or subacute low back pain identified as having yellow flags.
Key Questions 2a 2d Diagnostic testing

Because anatomic abnormalities of the spine are quite common in healthy persons, diagnostic imaging often identifies radiographic abnormalities that are only loosely associated with symptoms. In one systematic review of findings from plain radiography, degenerative changes (disc space narrowing, osteophytes, and sclerosis) were only modestly associated with low back pain (OR 1.2 to 3.3)313. Other findings, such as spondylolysis, spondylolisthesis, spina bifida, transitional vertebrae, spondylosis, and Scheuermanns disease also did not appear to be associated with symptoms. Another systematic review found advanced imaging methods (such as magnetic resonance imaging [MRI] or computed tomography [CT]) more likely to identify radiologic abnormalities in asymptomatic patients than plain radiography268. The proportion of asymptomatic patients with herniated disc on MRI, for example, ranged from 9% to 76%, degenerative disc from 46% to 93%, and stenosis from 1% to 21%. Greater use of advanced diagnostic imaging may therefore be associated with additional testing and interventions. For example, a significant proportion of the geographic variation in spinal surgery rates across the U.S. appears to correlate with differential rates of obtaining MRI314. On the other hand, patients and providers may be reassured by obtaining imaging tests, even if the findings dont necessarily alter management315.
Key Question 2a How accurate are different diagnostic tests for identifying serious underlying conditions (e.g., tumor, infection, compression fracture)?
Results of search: systematic reviews We identified one recent systematic review that evaluated diagnostic accuracy of plain radiography, MRI, CT, or radionuclide scanning for identifying serious underlying conditions associated with low back pain268. We also identified one higher-quality systematic review on diagnostic accuracy of erythrocyte sedimentation rate testing in patients with low back pain271. We excluded four other systematic reviews because they were outdated273, 316, 317 or reported duplicate information275 from another systematic review268.
34
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: primary studies The systematic review found numerous flaws in diagnostic studies, with the most common being failure to apply a single reference test to all patients, test review bias (study test was reviewed with knowledge of the final diagnosis), diagnosis review bias (determination of the final diagnosis was affected by the study test), and spectrum bias (only severe cases of disease were evaluated)268. Additional limitations of primary studies include heterogeneous populations, small sample sizes, and small numbers of studies. Estimates of diagnostic accuracy were therefore considered imprecise, and ranges rather than pooled estimates were reported. We did not search for additional primary studies.
Accuracy of imaging for diagnosing cancer

The accuracy of diagnostic imaging for diagnosing vertebral cancer is summarized in Table 5. Plain radiography was associated with lower sensitivity for metastatic cancer than MRI or radionuclide scanning (with planar imaging or single photon emission computed tomography [SPECT]), though it was associated with high specificity268. Magnetic resonance imaging and SPECT were associated with similar diagnostic accuracy. Planar imaging was less accurate than SPECT.
Table 5. Estimated accuracy of different imaging techniques for diagnosing cancer (ranges)
Positive likelihood ratio 12-120 8.3-31 3.9 9.7 Negative likelihood ratio 0.40-0.42 0.07-0.19 0.32 0.14
Technique Plain radiography MRI Radionuclide scanning with planar imaging SPECT
Sensitivity 0.6 0.83-0.93 0.74-0.98 0.87-0.93
Specificity 0.95-0.995 0.90-0.97 0.64-0.81 0.91-0.93
Source: Jarvik and Deyo, 2002268
Accuracy of imaging for diagnosing vertebral infection

For diagnosing vertebral infection, plain radiography was less accurate than MRI or radionuclide scanning (Table 6)268. MRI was more accurate than either plain radiography or radionuclide scanning.
Table 6. Estimated accuracy of different imaging techniques for diagnosing vertebral infection (ranges)
Positive likelihood ratio 1.9 12 4.1 Negative likelihood ratio 0.32 0.04 0.13
Technique Plain radiography MRI Radionuclide scanning
Sensitivity 0.82 0.96 0.90
Specificity 0.57 0.92 0.78
Source: Jarvik and Deyo, 2002268
Accuracy of imaging for diagnosing vertebral compression fracture

For vertebral compression fracture, plain radiography appears sensitive, but its ability to distinguish acute from chronic fracture is poor, and asymptomatic fractures are frequently
35
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain identified268. Although radionuclide scanning is insensitive for diagnosing fractures, it can help distinguish recent from old fractures. MRI can also provide additional information about the acuity of compression fractures.
Accuracy of elevated erythrocyte sedimentation rate for diagnosing cancer

One systematic review271 included one higher-quality study266 that found an ESR 20 mm/hr associated with a sensitivity of 0.78 and specificity of 0.67 for diagnosing vertebral cancer. Costs A decision analysis found that for diagnosing cancer in patients with low back pain, a strategy of selectively imaging patients with a positive clinical finding (history of cancer, age 50 years, weight loss, or failure to improve with conservative therapy) in combination with either an elevated ESR (50 mm/hr) or a positive x-ray was associated with the best cost-effectiveness ratio ($5,283 per case found)318. Using a similar strategy but directly imaging patients with a history of cancer resulted in similar estimates of cost-effectiveness. A decision analysis of diagnostic strategies for excluding cancer found rapid MRI associated with an incremental costeffectiveness of nearly $300,000/QALY relative to lumbar radiography319.
Summary of evidence
For diagnosing vertebral cancer, MRI and radionuclide scanning are more sensitive than plain radiography, though plain radiography is associated with high specificity (level of evidence: good). For diagnosing vertebral infection, MRI is more accurate than either lumbar radiography or radionuclide scanning (level of evidence: fair). For diagnosing vertebral compression fracture, lumbar radiography appears sensitive, but is unable to provide information about acuity (level of evidence: fair). For diagnosing vertebral cancer, an elevated erythrocyte sedimentation rate was associated with moderate sensitivity and specificity in one higher-quality study (level of evidence: fair).

The AHCPR guidelines state lumbar radiography in combination with CBC and ESR may be useful for ruling out tumor or infection in patients with acute low back problems when any of the following are present: prior cancer or recent infection, fever over 100 F, IV drug abuse, prolonged steroid use, low back pain worse with rest, or unexplained weight loss (strength of evidence: C). The AHCPR guidelines recommend prompt CT or MRI in the presence of red flags suggesting cauda equina syndrome or progressive motor weakness, preferably in consultation with a surgeon (level of evidence: C). The AHCPR guidelines recommend CT or MRI when clinical findings strongly suggest tumor, infection, fracture, or other space-occupying lesions of the spine (strength of evidence: C).
36
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The AHCPR guidelines state that in the presence of red flags, especially for tumor or infection, the use of other imaging studies such as bone scan, CT, or MRI may be clinically indicated even if lumbar radiography is negative (strength of evidence: C). The AHCPR guidelines recommend against CT-myelography and myelography because they are invasive and have an increased risk of complications, except in special situations for preoperative planning (strength of evidence: D). The European COST guidelines recommend MRI in patients with chronic low back pain with serious red flags.
Key Question 2b How accurate are different diagnostic tests for identifying other conditions (e.g. nerve root compression, herniated disc, spinal stenosis) that may respond to specific therapies?
Results of search: systematic reviews We identified the same systematic reviews described for Key Question 2a. In addition, we identified one other recent, higher-quality systematic review on diagnostic accuracy of imaging tests for spinal stenosis269, one higher-quality systematic review on accuracy of thermography for diagnosing lumbar radiculopathy320, and one lower-quality systematic review on accuracy of surface electromyogram321. Two other systematic reviews of surface electromyogram were excluded because they primarily evaluated whether the test could distinguish patients with low back pain from those without low back pain322, 323. Two outdated systematic reviews of imaging tests were also excluded324, 325. Results of search: trials We did not search for additional trials
Accuracy of imaging for diagnosing nerve root compression, herniated disc, and spinal stenosis
Plain radiography cannot directly visualize intervertebral discs and is therefore insensitive for diagnosis of disc herniation268. Similarly, facet osteophytes or severe spondylolisthesis on plain radiography can suggest nerve root impingement, but additional imaging is required to confirm the diagnosis. Plain radiography is also unable to detect compromise of the vertebral canal caused by soft tissue. One recent systematic review evaluated the accuracy of CT and MRI for diagnosis of herniated disc and spinal stenosis268. It found magnetic resonance imaging and computed tomography associated with similar accuracy for diagnosing either condition (Table 7). Another recent systematic review reached similar conclusions269. However, MRI is not associated with ionizing radiation and provides better visualization of soft tissues, vertebral marrow, and the spinal canal.
37
EVIDENCE REVIEW
Table 7. Estimated accuracy of different imaging techniques for diagnosing disc herniation and spinal stenosis (ranges reported)
Positive likelihood ratio 1.1-33 2.1-6.9 3.2-not defined 4.5-22 Negative likelihood ratio 0-0.93 0.11-0.54 0.10-0.14 0.10-0.22
Technique Sensitivity Herniated disc MRI 0.6-1.0 CT 0.62-0.9 Spinal stenosis MRI 0.9 CT 0.9 Source: Jarvik and Deyo, 2002268
Specificity 0.43-0.97 0.7-0.87 0.72-1.0 0.8-0.96
Accuracy of imaging for diagnosing ankylosing spondylitis

Evidence on diagnostic accuracy of different imaging methods for diagnosis of ankylosing spondylitis is sparse. In addition, interpretation of available evidence is difficult because new criteria for diagnosis of early disease prior to the development of radiographic findings of sacroiliitis have only recently been proposed277. One systematic review found plain radiography associated with a sensitivity of 0.26 to 0.45 and specificity of 1 for diagnosis of ankylosing spondylitis, but spectrum bias could have resulted in overestimates of accuracy268. Radionuclide scanning with planar imaging was associated with low sensitivity but high specificity in two studies (sensitivity 0.25 and 0.26, specificity 0.95 to 1.0)326, 327. In one other study, SPECT increased sensitivity to 0.85 but decreased specificity to 0.90326. MRI was associated with a sensitivity of 0.45 for diagnosis of ankylosing spondylitis in one study, but specificity could not be calculated327.
Diagnostic accuracy of other (non-imaging) tests

One higher-quality systematic review found an elevated ESR associated with a sensitivity of 0.69 and specificity of 0.68 for diagnosis of ankylosing spondylitis in patients suspected of having the disease271. Although the HLA-B27 antigen is present in approximately 90% of persons of western European ancestry who have ankylosing spondylitis, both the prevalence of HLA-B27 and the strength of its association with ankylosing spondylitis vary substantially in different ethnic groups328. A systematic review on diagnostic accuracy of thermography found major methodological flaws, inconsistent results, and no clear evidence supporting its use in diagnosis of radiculopathy320. Another systematic review found inconclusive and inadequate evidence to support the use of surface electromyography for diagnostic evaluation of low back pain321. Though electrophysiologic testing is often used when imaging and clinical exam findings are discordant or the diagnosis of radiculopathy or spinal stenosis is uncertain, we found no systematic reviews on its diagnostic accuracy329.
Summary of evidence
For diagnosis of herniated disc or spinal stenosis, MRI and CT scan are associated with similar diagnostic accuracy. However, MRI is associated with no ionizing radiation and
38
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain permits better visualization of soft tissues, vertebral marrow, and the spinal canal (level of evidence: good). For diagnosis of ankylosing spondylitis, evidence on diagnostic accuracy of different imaging methods (including MRI) is sparse. Plain radiography may have high specificity, but higherquality studies are needed. Estimates of diagnostic accuracy are likely to be affected by adoption of recently proposed criteria for diagnosis of early ankylosing spondylitis (prior to the development of radiographic evidence of sacroiliitis) (level of evidence: fair). For diagnosis of ankylosing spondylitis in patients suspected of having the disease, an elevated ESR was associated with moderate sensitivity and specificity. In persons of western European ancestry, the HLA-B27 antigen is associated with a sensitivity of about 90% (level of evidence: fair). There is no evidence supporting the use of thermography or surface electromyography for diagnosis of low back pain (level of evidence: fair). For diagnosis of radiculopathy or spinal stenosis, we found no systematic reviews evaluating diagnostic accuracy of electrophysiologic testing.

The AHCPR guidelines recommend against thermography for assessing acute low back problems (strength of evidence: C). The AHCPR guidelines recommend against electrophysiologic testing when the diagnosis of radiculopathy is obvious and specific on clinical examination (strength of evidence: D). The AHCPR guidelines recommend against surface electromyogram (EMG) and F-wave tests in patients with acute low back symptoms (strength of evidence: C). The AHCPR guidelines found needle EMG and H-reflex tests of the lower limb may be useful in assessing questionable nerve root dysfunction in patients with leg symptoms for longer than 4 weeks (regardless of presence of back pain) (strength of evidence: C). The AHCPR guidelines found that sensory evoked potentials (SEPs) may be useful in assessing suspected spinal stenosis and spinal cord myelopathy (strength of evidence: C). The VA/DoD recommendations for diagnostic imaging are essentially identical to the AHCPR recommendations. The European COST guidelines do not recommend EMG for evaluating chronic nonspecific low back pain. The European COST guidelines recommend MRI for evaluation of radicular symptoms, and plain radiography for evaluation of structural deformities. They recommend against MRI or CT for the diagnosis of facet joint pain.
39
EVIDENCE REVIEW
Key Question 2c In patients with red flags, how effective are different diagnostic tests for improving patient outcomes?
Results of search: systematic reviews We found no systematic reviews addressing this question. Results of search: trials No trials are available.
Efficacy of diagnostic testing in patients with red flags

We found no studies that compared outcomes associated with use of different diagnostic tests in patients with low back pain and associated cancer, vertebral infection, or cauda equina syndrome. All guidelines recommend prompt and appropriate work-up (including advanced imaging techniques) and management of patients strongly suspected of having these conditions or with a history of significant vertebral trauma, because delayed diagnosis and treatment can be associated with poorer outcomes24, 34, 35, 39.
Summary of evidence
There is no direct evidence on use of different diagnostic tests in patients with worrisome red flags, though all guidelines recommend prompt and appropriate work-up (including advanced imaging) because delayed diagnosis and treatment can be associated with poorer outcomes. Recommendations and findings of other guidelines The AHCPR, VA/DoD, UK RCGP, and European COST guidelines all recommend prompt work-up and immediate action in patients with low back pain suspected of having a red flag condition. The European COST guidelines recommend MRI in patients with chronic low back pain with serious red flags.
Key Question 2d In patients without red flags, how effective are different diagnostic tests or test strategies (including no testing) for improving patient outcomes?
Results of search: systematic reviews We found no systematic reviews addressing this question. Results of search: trials From 430 potentially relevant citations, we identified four randomized controlled trials on routine lumbar radiography versus clinical care without routine imaging in patients without red flags who present for initial evaluation of low back pain330-333. Routine lumbar radiography was compared to usual care in three trials (two higher-quality332, 333 and one lower-quality331) and to a brief educational intervention in one higher-quality trial330. Four other trials evaluated different strategies for using MRI in patients with low back pain. One higher-quality trial (n=782)
40
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain compared early routine versus delayed selective MRI or CT in patients presenting to surgical clinics for evaluation of low back pain334. A higher-quality trial performed MRI in all patients, and compared outcomes when MRI findings were routinely provided to clinicians and patients versus disclosure only if clinically indicated335, 336. Two higher-quality trials (conducted by the same investigators and using the same study design) evaluated effects of rapid MRI versus plain radiography in patients with low back pain in whom imaging was clinically felt appropriate337, 338.
Efficacy of routine, early plain radiography versus usual care or imaging only if clinically necessary (or without improvement)
For acute330, acute or subacute331, subacute or chronic332, or back pain of unspecified duration333, routine lumbar radiography in patients without red flags was not associated with improved patient functioning, time off work, severity of pain, or overall health status in any of the trials (Table 8). One higher-quality trial (n=153) found routine lumbar radiography slightly superior to usual care for psychological well-being333. Another large (n=421), higher-quality trial found routine radiography associated with increased physician visits in the 3 months after imaging and a trend towards a higher likelihood of pain at six months, but also increased patient satisfaction, though differences were small332. Results of a third, higher-quality trial found routine lumbar radiography was not associated with increased anxiety, dissatisfaction, dysfunction or differences in subsequent clinical treatments compared to a brief educational intervention and no routine imaging330. No serious missed diagnoses were identified in any patient enrolled in the three trials that recorded low-back pain diagnoses based on clinical follow-up through at least 6 months of follow-up330, 332, 333.
41
EVIDENCE REVIEW
Table 8. Trials of early plain radiography versus imaging only if clinically necessary
Author, year Deyo, 1987330
Djais, 2005331
n=101 3 weeks
Kendrick, 2001332
n=421 9 month
Kerry, 2002333
n=153 1 year
Main results Routine plain radiography vs. selective imaging + brief educational intervention Sickness Impact Profile (0 to 100, higher indicating worse function): 16.6 vs. 13.6 at 3 weeks (NS), 12.3 vs. 10.3 at 3 months (NS) Days of work absenteeism: 4.1 vs. 4.4 at 3 weeks (NS) Additional days of work loss: 0.28 vs. 0.05 at 3 months (NS) Self-rated improvement (1 to 6 scale): 2.7 vs. 2.7 at 3 weeks, 2.6 vs. 2.6 at 3 months Duration of pain: 9.4 vs. 10.8 days at 3 weeks (NS), 13.3 vs. 18.4 additional days of pain at 3 months (NS) Total physician visits: 1.07 vs. 0.42 at 3 months Overall satisfaction score (9 to 27 scale): 23.7 vs. 24.0 No differences for other measures of patient perceptions and attitudes (including worry that pain is due to serious illness) Routine plain radiography vs. usual care (median values, 3 week outcomes) RDQ: 6.5 vs. 4.5 (p=0.21) VAS pain score: 4 vs. 3 (p=0.07) EQ-5D: 0.63 vs. 0.74 (p=0.15) Health status scale: 70 vs. 80 (p=0.02) Pain "much improved": 25.5% vs. 40.0% (p=0.11) Routine plain radiography vs. usual care (9 month data) Still has pain at 6 months: 65% vs. 57% (p=0.11) Taken time off work: 13% vs. 13% (p=0.87) Median days off work: 11.5 vs. 8.5 (p=0.84) Median RDQ score: 3 vs. 2 (p=0.06) Median pain score: 1 vs. 1 (p=0.17) Median health status score: 80 vs. 80 (p=0.30) Median satisfaction with consultation: 21 vs. 19 (p<0.01, favors routine radiography) 3 visits to doctor: 5% vs. 5% Visited provider within 3 months: 53% vs. 30% (RR 1.62, 95% CI 1.33 to 1.97) Routine plain radiography vs. usual care (1 year data) SF-36, adjusted difference (not referred - referred): no subscale significant except for mental health -8, p<0.05 EuroQol, adjusted difference: 1 (NS) RDQ score (0 to 24), adjusted difference: -0.3 (NS) Consulted for back pain 6 weeks to 1 year: 32% vs. 39%, AOR 0.7 (0.3 to 1.4) Referred to other health professional 6 weeks to 1 year: 45% vs. 46%, AOR 1.1 (0.5 to 2.3) Very satisfied at 6 weeks: 33% vs. 28%, AOR 1.3 (0.6 to 3.0) Days off work, 0-12 months: 8.46 vs. 6.16 GP consultations: 1.6 vs. 1.1, p=0.06 Other consultations: 5.9 vs. 2.9, p=0.003
Quality score* 5/8
2/8
6/8
4/8
*Excludes criteria involving blinding of patients and providers and similarity of co-interventions, for maximum score of 8
42
EVIDENCE REVIEW
Efficacy of routine MRI versus MRI only if clinically necessary (or without improvement)
One higher-quality trial (n=782) found that in patients with low back pain of varying duration (40% with symptoms for >1 year) referred to surgeons with uncertain need for advanced imaging, routine early MRI or CT was associated with statistically significant but small differences in the Aberdeen Low Back Pain Score, SF-36 Bodily Pain Scale, and Euro-Qol after 8 and 24 months relative to delayed, selective imaging (Table 9)334. Effects on pain averaged about 3 points on 0 to100 scales. There were no differences in the proportion of patients who underwent surgery or received injections, or on other measures of health care use. A higherquality trial that obtained MRI in all patients with acute low back pain or radiculopathy found routine disclosure of MRI findings to patients and physicians was not associated with greater improvements in RDQ function scores compared to not disclosing MRI results unless clinically necessary335, 336. There were also no differences on any of the SF-36 subscales other than general health, which favored the no routine disclosure arm (6.0 vs. 4.2 point improvement at 6 weeks, p=0.008).
Table 9. Trials of early MRI versus imaging only if clinically necessary
Number of patients Duration of follow-up n=782 2 years
Author, year Gilbert, 2004334
Modic, 2005335, 336
n=246 6 weeks
Main results Early imaging (90% had MRI or CT) vs. delayed (30% had MRI or CT) (24 month data) Subsequent outpatient appointment: 84% vs. 68%, p<0.001 Total number of consultations: 1.91 vs. 1.88 (NS) Hospital admissions: 7.9% vs. 6.7% (NS) Surgical operation: 6.9% vs. 5.1% (NS) Injections: 17.8% vs. 19.3 % (NS) Aberdeen Low Back Pain score (0 to 100 scale), adjusted mean difference: -3.62, p=0.002 EQ-5D score (-0.59 to +1 scale), adjusted mean difference: 0.057, p=0.01 SF-36, bodily pain (0 to 100 scale), adjusted mean difference: 5.14, p=0.004 No differences on other SF-36 subscales Unblinded vs. blinded MR results >50% improvement in RDQ function: 60% vs. 67% (p=0.397) Proportion 'satisfied' with condition: 23% vs. 31% (p=0.207) Self-efficacy, fear-avoidance beliefs, and SF-36: similar between arms except for general health subscale of SF-36, mean improvement 4.2 vs. 6.0 at 6 weeks (p=0.008)
Quality score* 6/8
4/8
Efficacy of rapid MRI versus lumbar radiography in patients with low back pain referred for imaging
In the larger (n=380) of two higher-quality trials comparing rapid MRI to lumbar radiography in patients with low back pain referred for imaging (duration of symptoms not specified), there was no difference in any outcomes including functional status, pain intensity, or rate of spinal surgery (Table 10)337. There was a trend towards increased lumbar spine operations in the rapid MRI
43
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain group (risk difference=0.34, 95% CI -0.06 to 0.73). The smaller (n=62) trial (conducted by the same investigators) did not assess rates of lumbar spine operations, but otherwise reported similar findings338.
Table 10. Trials of rapid MRI versus lumbar radiography in patients referred for imaging
Author, year Jarvik, 1997338
Jarvik, 2003337
n=380 6 weeks
Main results Rapid MRI vs. lumbar radiography (3 month data) Modified RDQ score: 12.5 vs. 12.1 (p=0.40) SF-36: No differences Pain bothersomeness (0 to 24): 9.7 vs. 10.0 (p=0.79) Pain frequency (0 to 24): 10.1 vs. 9.9 (p=0.35) Disability days: No differences for number of home days, limited activity days, or bed days Patient satisfaction: Only differences among 12 questions about patient satisfaction were proportion who thought clinicians were concerned (75% vs. 100%, p=0.01) and proportion who felt reassured (72% vs. 37%, p=0.03) Proportion of patients referred to back specialists: 32% vs. 36% Rapid MRI vs. lumbar radiography RDQ Scale score, adjusted (12 month): 9.34 vs. 8.75 (NS) (score better for MRI at 3 months) SF-36: No differences at 12 months for bodily pain, physical functioning, role-physical Pain-bothersomeness: 9.68 vs. 9.75, NS Pain-frequency: 10.09 vs. 10.21, NS Lost work, days past 4 weeks: 1.57 vs. 1.26, NS Patient satisfaction: 7.04 vs. 7.34, NS Patient reassurance score: 3.18 vs. 2.50, p<0.05 favoring MRI Proportion reporting reassurance from imaging: 74% vs. 58% (p=0.002) Lumbar spine surgery: 6% vs. 2% (risk difference=0.34, 95% CI -0.06 to +0.73)
Quality score* 5/8
7/8
Costs Several recent RCTs of routine versus selective imaging also conducted cost-effectiveness analysis. In one trial,332 the cost-effectiveness of routine lumbar radiography was estimated at 20 (equivalent to about $39 U.S. in January 2007) per additional point on a patient satisfaction scale (scored between 9 and 27), the only outcome for which there was a difference in efficacy339. The increased cost was mostly related to direct costs associated with the imaging procedure itself. In another trial, early MRI or CT imaging was associated with a mean of 0.041 additional QALY during 24 months compared to selective MRI or CT, with an incremental costeffectiveness of $2,124/QALY334. An older decision analysis found that costs associated with routine lumbar radiography in patients with acute low back pain did not appear to justify the small benefits ($2,072 to avert one day of physical suffering)340. Finally, rapid MRI imaging was associated with additional costs of about $300 relative to lumbar radiography in patients with
44
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain low back pain referred for imaging, with nearly identical clinical outcomes (essentially a costminimization analysis)337.
Summary of evidence
For acute low back pain, the combination of delayed selective imaging with a brief educational intervention was not associated with differences in any outcomes relative to routine lumbar radiography, including patient satisfaction and psychological distress (one higher-quality trial) (level of evidence: fair). For acute or subacute back pain (one lower-quality trial), subacute or chronic back pain (one higher-quality trial) and back pain of unspecified duration (one higher-quality trial), routine lumbar radiography did not improve outcomes including pain and functional status, though small beneficial effects on patient satisfaction and psychological well-being were present in two trials (level of evidence: good). No serious missed diagnosis was identified in any patient enrolled in trials of routine lumbar radiography versus clinical care without routine imaging after at least 6 months of follow-up (level of evidence: fair). For back pain of varying duration, routine MRI was associated with only small benefits on pain and functional status outcomes compared to selective MRI in one higher-quality trial. For acute low back pain, one higher-quality trial found that in patients who underwent MRI, knowledge of imaging results was not associated with improved outcomes compared to nondisclosure unless clinically necessary (level of evidence: fair). In patients for whom imaging was felt to be indicated (duration of symptoms not specified), two higher-quality trials found rapid MRI associated with no significant benefits compared to plain radiography, and a trend towards increased surgeries in one of the trials (level of evidence: good).

The AHCPR guidelines recommend against lumbar radiography for routine evaluation of patients with acute low back problems within the first month of symptoms, unless a red flag is noted on clinical examination (strength of evidence: B). The AHCPR guidelines state that in patients without red flags, after 1 month of symptoms, an imaging test is acceptable when surgery is being considered (or to rule out a suspected serious condition) (strength of evidence: B). The UK RCGP guidelines recommend avoidance of unnecessary or repeated imaging, noting that lumbar spine radiography results in 150 times the gonadal radiation exposure of chest radiography in women (strength of evidence: ***). The European COST guidelines recommend against routine diagnostic imaging for acute or chronic nonspecific low back pain.
45
EVIDENCE REVIEW
Key Question 3 How effective are self-care advice, education, or other self-care interventions for improving patient outcomes?
Goals of patient education and patient self-care methods for low back pain are to reduce fear of normal activity, encourage exercise, and promote self-management of pain. A range of interventions have been defined as self-care for low back pain, including individual consultation with a professional or team of professionals, group treatment and/or education by professionals or trained lay leaders, group exercise classes, mini-back school and other approaches. For this report, we defined self-care advice and education as individual or group educational sessions that involve two sessions or fewer with a professional in a routine clinic visit and provides advice that is readily implemented independently by patients. We also included self-care groups led by trained or untrained non-medical lay persons. We defined self-care interventions as interventions that could be readily implemented by patients without seeing a health provider (lumbar supports, application of superficial heat or cold, use of different types of mattresses).
Self-care advice or education Advice to rest in bed

Search results: systematic reviews We identified a higher-quality Cochrane review (11 trials) on advice to rest in bed in patients with low back pain64, 65. We excluded an earlier version of the Cochrane review341 and seven other outdated systematic reviews193, 342-347. Search results: trials The Cochrane review included 11 trials of advice to rest in bed64, 65. Eight trials were rated higher-quality. All trials evaluated patients with acute low back pain, except for one trial of bed rest versus epidural anesthesia in patients with subacute or chronic sciatica348. Six trials compared advice to rest in bed versus advice to remain active, four versus exercise, two versus other interventions, and two compared different durations of bed rest. We did not search for additional trials.
Efficacy of advice to rest in bed versus advice to remain active

For acute non-specific low back pain, the Cochrane review included two-higher quality trials349, 350 that found advice to rest in bed associated with slightly inferior outcomes compared to advice to remain active for pain intensity (SMD=0.22, 95% CI 0.02 to 0.41 at 3 to 4 weeks and SMD=0.25, 95% CI 0.05 to 0.45 at 12 weeks) and functional status (SMD=0.29, 95% CI 0.09 to 0.49 at 3 to 4 weeks and SMD=0.24, 95% CI 0.04 to 0.44 at 12 weeks)64, 65. Differences were roughly equivalent to 5 to 7.5 points on a 100 point VAS pain scale and 1.2 to 1.8 points on the RDQ. Both trials also found advice to stay active superior to advice to rest in bed for sick leave and one trial349 found no difference in satisfaction with care or costs. For low back pain with sciatica, the Cochrane review included two higher-quality trials that found little or no difference between advice to rest in bed and advice to stay active for pain intensity, functional status, or sick leave351, 352.
46
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The Cochrane review excluded two lower-quality trials from pooled analyses because of low internal validity353, 354. In addition, one of the trials (n=80) evaluated army combat trainees and may not be applicable to patients encountered in routine practice353. It found bed rest superior to remaining ambulatory with restricted duties for rate of recovery, pain and days off work. The other trial found no statistically significant difference between advice to rest in bed and advice to remain active in family practice clinic patients with acute low back pain, though trends in rate of recovery and disability favored advice to remain active354.
Efficacy of advice to rest in bed versus other interventions

For acute non-specific low back pain, the Cochrane review64, 65 included two higher-quality trials that found no significant differences in pain intensity or functional status between advice to rest in bed and exercise therapy349, 355. A third, lower-quality trial found no difference on a combined pain, disability, and physical exam score between bed rest and manipulation, drug therapy, physiotherapy, back school, or placebo356. Another lower-quality trial found bed rest inferior to epidural anesthesia for time to recovery (31 versus 11 days, p<0.001)348. For back pain with sciatica, one higher-quality trial included in the Cochrane review64, 65 found physiotherapy (activity advice, mobilization, hydrotherapy, and disc unloading and loading exercises) slightly superior (WMD=6.9 points on a 0 to 100 scale) to advice to rest in bed for functional status at four weeks, though the difference was no longer significant at 12 weeks351. There were no differences in pain intensity.
Efficacy of different durations of bed rest

For acute357 or mixed duration358 low back pain, the Cochrane review64, 65 included two higherquality trials that found advice for shorter duration of bed rest (2 or 3 days) associated with similar effects on pain intensity compared to advice for longer duration of bed rest (7 days). One of the trials found advice for shorter bed rest associated with fewer days off work (mean 3.1 days) compared to advice for longer bed rest (mean 5.6 days) after 3 weeks, with benefits persisting through 12 weeks358. Only one-quarter of patients assigned to 7 days of bed rest actually rested for 7 days, which may have attenuated differences between advice for shorter versus longer durations of bed rest. Harms One trial reported one case of pulmonary embolus in patients assigned to bed rest351. Costs One trial found no significant differences in costs of health care and home help between advice for bed rest and either exercise or usual activities (usual activities associated with more rapid recovery in this trial)349.
Summary of evidence
For acute non-specific low back pain, advice to rest in bed was consistently associated with slightly inferior pain and functional status compared to advice to remain active in two higherquality trials (level of evidence: good).
47
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For acute non-specific low back pain, advice to rest in bed was associated with similar outcomes compared to exercise programs in three trials (two higher-quality) (level of evidence: good). For acute non-specific low back pain, there is insufficient evidence (one lower-quality trial) to accurately judge efficacy of advice to rest in bed compared to interventions other than exercise (level of evidence: poor). For back pain with sciatica, one higher-quality trial found no difference between advice to rest in bed and advice to remain active, and advice to rest in bed was associated with slightly inferior functional status at 3 weeks compared to a combined physiotherapy intervention, though this difference was no longer present after 12 weeks (level of evidence: good). Advice for seven days of bed rest was not associated with better pain outcomes compared to advice for two to three of bed rest in two higher-quality trials, and increased the number of days off work in one of these trials (level of evidence: fair). There is no evidence to judge efficacy of advice to rest in bed in patients with chronic low back pain.

The AHCPR guidelines found that a gradual return to normal activities is more effective than prolonged bed rest for treating acute low back problems (strength of evidence: B). The AHCPR guidelines found that prolonged bed rest for more than 4 days may lead to debilitation and is not recommended for treating acute low back problems (strength of evidence: B). The AHCPR guidelines found that the majority of low back patients will not require bed rest, though bed rest for 2 to 4 days may be an option for patients with severe initial symptoms of primarily leg pain (strength of evidence: D). The VA/DoD and UK RCGP guidelines are similar to the AHCPR guidelines, but found stronger evidence that bed rest for 2-7 days is inferior to placebo or ordinary activity (strength of evidence: A and ***, respectively). The European COST Guidelines recommend against prescribing bed rest for acute nonspecific low back pain.
Advice to remain active

For this section, we included studies of advice to remain active (maintaining usual activities as much as possible) or advice on exercises provided in a typical clinic visit or a clinic session lasting less than one hour. Advice and education that require more than an hour-long clinic session are reviewed in the section on brief educational interventions (Key Question 4).
48
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Search results: systematic reviews We identified one higher-quality Cochrane review on efficacy of advice to remain active359, 360. Another higher-quality Cochrane review on advice to rest in bed64, 65 included two additional trials (both higher-quality) that compared advice to remain active with bed rest350, 351. Search results: trials A total of 6 trials on advice to remain active were included in the two systematic reviews64, 65, 359, 360 . Five were rated higher-quality349-352, 354. All six trials compared advice to remain active to advice to rest in bed for acute low back pain. Two trials also compared advice to remain active to formal exercise therapy349, 351. We identified four additional trials not included in the systematic review. One higher-quality trial evaluated advice to stay active versus a combined physical therapy intervention361, one higher-quality trial evaluated advice to remain active, exercise, or both versus sham therapies362, one lower-quality trial evaluated exercise advice versus usual care or a self-care book363, and one lower-quality trial evaluated exercise advice versus supervised McKenzie exercise364, 365.
Efficacy of advice to remain active versus advice to rest in bed

Results of trials that compared advice to stay active with advice to rest in bed are discussed in the section on advice to rest in bed.
Efficacy of advice to remain active versus other interventions

For acute non-specific low back pain, the Cochrane review included one higher-quality trial349 that found advice to remain active associated with similar improvements in pain intensity compared to a formal exercise program359, 360. Short-term functional status (ODI) initially slightly favored advice to stay active (WMD=-8.6, 95% CI -13.9 to -3.3), but differences were no longer present after 3 weeks. Average length of sick leave was lower in the advice to remain active group, but differences were not statistically significant. One lower-quality trial not included in the Cochrane review found a single back education session with advice to remain active (45 minutes) associated with slower return from sick leave (22 vs.12 days, p<0.001) and more back pain recurrences through five years compared to supervised McKenzie exercise (Table 11)364, 365 . For subacute non-specific low back pain, a higher-quality trial not included in the Cochrane review found no clear differences between physiotherapist-provided advice to remain active and supervised exercise therapy (including aerobic exercise, stretches, functional activities, strength, speed, coordination, and endurance training) (Table 11)362. At 6 weeks, advice to remain active was slightly superior to sham advice for pain, the patient specific functional scale, and global perceived effect, and exercise was slightly superior to sham ultrasound plus diathermy on the same three outcomes. However, benefits with either intervention were no longer statistically significant by 12 months. Neither intervention was more effective than sham therapies on the RDQ or the Depression Anxiety Stress Scale at any assessment. For nonspecific back pain of more than 6 weeks duration, a higher-quality trial not included in the Cochrane review found no differences between advice to remain active and a physical
49
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain therapy intervention (consisting of any combination of stretching, spinal mobility, and strengthening exercises, manipulation and/or mobilization, superficial heat or cold, and advice) on pain or functional status through 12 months, though perceived benefit was greater in the physical therapy group (Table 11)361. For acute sciatica, one higher-quality trial found no differences in pain or functional status between advice to stay active and physical therapy (consisting of advice, mobilization, disc unloading and loading exercises, and hydrotherapy) through 6 months follow-up (Table 11)351.
Table 11. Trials of advice to remain active vs. exercise therapy not included in Cochrane review
Author, year Type of LBP Frost 2004361 Nonspecific low back pain
Pengel, 2007362 Nonspecific low back pain
n=259 12 months
Main results Advice to remain active vs. standard physical therapy (any combination of exercises, mobilization and/or mobilization, superficial heat or cold, and advice) ODI (0 to 100 scale), mean change: -1.33 vs. -2.65 at 2 months, -2.23 vs. -3.27 at 12 months (NS) RDQ (0 to 24 scale), mean change: -0.56 vs. -1.13 at 2 months, -0.99 vs. -1.36 at 12 months (NS) SF-36: No significant differences Perceived benefit (proportion reporting 'yes'): 60% vs. 77% at 2 months (p=0.002), 50% vs. 65% at 6 months (p=0.007) Perceived benefit (0 to 10 scale): 3.66 vs. 5.42 at 2 months (p<0.001); 4.13 vs. 5.02 at 12 months (p=0.011) Advice versus sham advice (mean change reported for all results) Pain (0 to 10 scale): -0.7 (95% CI -1.2 to -0.2) at 6 weeks, -0.4 (95% CI -1.0 to +0.3) at 12 months Patient-specific functional scale (0 to 10 scale): +0.7 (95% CI +0.1 to +1.3) at 6 weeks, +0.6 (95% CI +0.1 to +1.2) at 12 months Global perceived effect (-5 to +5 scale): +0.8 (95% CI +0.3 to +1.2) at 6 weeks, +0.3 (95% CI -0.2 to +0.9) at 12 months RDQ (0 to 24 scale): -0.5 (95% CI -1.6 to +0.5) at 6 weeks, -0.6 (95% CI -1.9 to +0.6) at 12 months Depression Anxiety Stress Scale (0 to 42 scale): +0.8 (95% CI -1.0 to +2.7) at 6 weeks, +0.3 (95% CI -1.7 to +2.2) at 12 months Exercise versus sham ultrasound plus sham diathermy (mean change reported for all results) Pain: -0.8 (95% CI -1.3 to -0.3) at 6 weeks, -0.5 (95% CI -1.1 to +0.2) at 12 months Patient-specific functional scale: +0.4 (95% CI -0.2 to +1.0) at 6 weeks, +0.5 (95% CI -0.1 to +1.0) at 12 months Global perceived effect: +0.5 (95% CI +0.1 to +1.0) at 6 weeks, +0.4 (95% CI -0.1 to +1.0) at 12 months RDQ: -0.8 (95% CI -1.8 to +0.3) at 6 weeks, -0.3 (95% CI -1.6 to +0.9) at 12 months Depression Anxiety Stress Scale: -0.7 (95% CI -2.5 to +1.2) at 6 weeks, -0.6 (95% CI -2.6 to +1.3) at 12 months
Quality score* 7/9
8/9
50
EVIDENCE REVIEW
Table 11. Trials of advice to remain active vs. exercise therapy not included in Cochrane review
Number of patients Duration of follow-up n=100 5 years Nonspecific low back pain
365
Author, year Type of LBP Stankovic, 1990, 1995364,
Main results Advice to remain active vs. McKenzie exercise Mean duration of sick leave: 22 vs. 12 days (p<0.001) Pain: decreased in exercise group (p<0.001), data not reported Recurrences: 74% (37/50) vs. 44% (22/50) after 1 year; 88% (37/42) vs. 64% (30/47) between 1 and 5 years (p<0.01) Sick leave between 1 and 5 years: 74% (31/42) vs. 51% (24/47) (p<0.03)
Quality score* 3/9
Efficacy of exercise advice versus usual care or a self-care book

For low back pain present less than 3 months, one lower-quality trial not included in the Cochrane review found advice for regular exercise superior to usual care for pain and function after one week (Table 12)363. Differences were no longer present after three weeks, when most patients in both groups reported resolved pain. Advice to exercise also improved patient satisfaction compared to usual care (p=0.03). There were no differences between advice to exercise and a self-care book. Adding a self-care education book to exercise advice also did not improve outcomes compared to either intervention alone.
Table 12. Trial of exercise advice vs. self-care book vs. usual care
Number of patients Duration of follow-up n=311 3 weeks
Author, year Little 2001363
Main results Self-care book vs. exercise advice vs. both vs. neither (control) (mean changes versus control) Pain/function scale (0 to 100): -8.7 vs -7.9 vs -0.1 at 1 week, -6.3 vs -1.4 vs -4.0 at 3 weeks (NS) Aberdeen pain and function scale (0 to 100): -3.8 vs -5.3 vs. -1.9 at 1 week (NS)
Quality score* 4/9
Harms No trial reported adverse events. Costs One of the trials included in the Cochrane review found no significant differences in costs of health care and home help between advice to remain active and either advice to rest in bed or an exercise program349.
51
EVIDENCE REVIEW
Summary of evidence
See section on advice for bed rest for summary of evidence on advice for bed rest versus advice to remain active. For acute low back pain, one higher-quality trial found advice to remain active associated with similar effects on functional status and pain compared to exercise therapy, but one lowerquality trial found more back pain recurrences. Effects on sick leave were mixed, with the no differences between advice to remain active and exercise therapy in the higher-quality trial (level of evidence: poor). For subacute low back pain or back pain present for longer than 6 weeks, advice to remain active was associated with similar effects on functional status and pain compared to exercise therapy or a combined physical therapy intervention in two higher-quality trials (level of evidence: fair). For acute sciatica, advice to remain active was not associated with clear benefits compared to a combined physiotherapy intervention in a single, higher-quality trial (level of evidence: fair). In patients with low back pain for less than 90 days, advice to exercise was superior to usual care in one lower-quality trial. There were no differences between advice to exercise and a self-care book, and combining the two interventions did not improve outcomes (level of evidence: poor).

The AHCPR guidelines recommend providing acute low back pain patients accurate information about expectations for both rapid recovery and recurrence of symptoms based on the natural history of low back symptoms; safe and effective methods of symptom control; safe and reasonable activity modifications; best means of limiting recurrent low back problems; the lack of need for special investigations unless red flags are present; and effectiveness and risks of commonly available diagnostic methods and further treatment measures to be considered should symptoms persist (strength of evidence: B). The AHCPR guidelines suggest that patients with acute low back problems may be more comfortable if they temporarily limit or avoid specific activities known to increase mechanical stress on the spine, especially prolonged unsupported sitting, heavy lifting, and bending or twisting the back while lifting (strength of evidence: D). The AHCPR guidelines suggest consideration of the patients age and general health, as well as the physical demands of required job tasks, when evaluating activity recommendations for employed workers with acute low back problems (strength of evidence: D). The VA/DoD and UK RCGP guidelines recommend an approach to patient advice and return to normal activity very similar to the AHCPR guidelines. The UK RCGP guidelines found stronger evidence than the AHCPR guidelines for advice to continue ordinary activity (strength of evidence: ***).
52
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The European COST guidelines recommend providing adequate information and reassurance to patients with acute low back pain. They also recommend advising patients to stay active and continue normal daily activities including work if possible.
Self-care back education books

We defined self-care back education books as reading material (books, booklets, or leaflets) that provides education and self-care advice for patients with low back pain. Although the specific content varies, self-care books are generally based on principles from published clinical practice guidelines and encourage return to normal activity, adoption of a fitness program and appropriate lifestyle modification, and provide advice on coping strategies and managing flares. Results of search: systematic reviews We identified no systematic reviews on efficacy of self-care books for low back pain. Results of search: trials From 611 potentially relevant citations, we identified ten trials (seven higher-quality309, 366-371) on self-care books for low back pain363, 372, 373. Nearly all of the trials evaluated patients with mixed acute, subacute, and/or chronic low back pain, or did not specify duration of symptoms. Five trials (two rated higher-quality368, 370) compared a self-care book to usual care and four trials (three rated higher-quality367, 369, 371) compared a self-care book to another intervention. Two trials compared different methods of providing information from self-care books366, 368. Three trials that compared a self-care book combined with another intervention versus a self-care book alone are reviewed in Key Question 10.
Efficacy of self-care books versus usual care

Five trials (two rated higher-quality368, 370) of patients with acute or subacute low back pain or back pain of unspecified duration found no significant differences between a self-care book versus usual care for pain or symptom bothersomeness scores (Table 13)363, 368, 370, 372, 373. There were also no differences in functional status363, 368 or time lost from work368, 370 in the trials that assessed these outcomes. Effects on health care use were mixed. One higher-quality trial found no difference between a self-care book and usual care on number of health care visits370, but one lower-quality trial found fewer patients receiving a self-care book consulted for back pain over a one-year period373. Effects of self-care books on self-reported behaviors were also mixed. One trial found patients randomized to a self-care book more likely to report recommended back care behaviors373. However, another trial found no difference between a self-care book and usual care in the proportion of patients who reported exercising, even though the self-care book group was associated with higher scores on perceived knowledge368.
53
EVIDENCE REVIEW
Table 13. Trials of a self-care book versus usual care

Number of patients Duration of follow-up n=300 1 year
Author, year Duration of LBP Cherkin, 1996368 Not specified
Hazard, 2000370 Not specified
n=1108 6 months
Main results Self-care book vs. nurse education + self-care book vs. usual care (mean change from baseline) RDQ score (0 to 24 scale): -5.4 vs -5.2 vs -5.3 (NS) at 1 week Symptom bothersomeness score (0 to 10 scale): -3.3 vs -3.3 vs -3.6 (NS) at 1 week Health care visits for low back pain: 45% vs. 46% vs. 47% in first 7 weeks after intervention (NS) Work loss days: 24% vs. 36% vs. 29% in first 7 weeks after intervention (NS) Self-care book vs. usual care Current pain severity, improvement in pain since maximum severity: no differences (data not reported) Number of health care visits: no differences (data not reported) Proportion not working at 6 months: 6.5% vs. 5.9% (p=0.84) Lost work days through 6 months: 19.1 vs. 18.1 Self-care book vs. exercise advice vs. both vs. neither (control) (mean changes versus control) Pain/function scale (0 to 100): -8.7 vs -7.9 vs -0.1 at 1 week, -6.3 vs -1.4 vs -4.0 at 3 weeks (NS) Aberdeen pain and function scale (0 to 100): -3.8 vs -5.3 vs. -1.9 at 1 week (NS) Self-care book vs. usual care Aberdeen Low Back Pain Scale (0 to 100): 42.7 vs. 42.6 at 2 days, 11.0 vs. 8.1 at 1 year (NS) Self-care book vs. usual care Patients initiating consultation for back pain: 23% vs 25% (NS) after 2 weeks,35.6% vs. 42.2% (p<0.05) over 1 year Days certified sickness absence: 10.3 vs 10.1 (NS) Referral to hospital or to physiotherapy: 19.9% vs. 24.7% (p>0.05)
Quality score* 6/9
5/9
Little, 2001363 Acute or subacute (<3 months) Roberts, 2002372 Acute (not defined) Roland, 1989373 Acute and chronic
n=311 3 weeks
4/9
n=64 12 months n=936 1 year
4/9
2/9
Efficacy of self-care books versus other interventions

Five trials (four higher-quality309, 367, 369, 371) compared a self-care book to other treatments (Table 14)363. For chronic low back pain, one higher-quality trial found a self-care book associated with moderately lower functional status at 26 weeks compared to yoga (difference of 3 to 4 points on the RDQ), and exercise therapy (difference of about 2 points on the RDQ)371. Yoga (but not exercise) was substantially superior to a self-care book on symptom bothersomeness scores at 26 weeks (by about 2 points on a 0 to 10 scale). For back pain of at least seven days duration, another higher-quality trial found no significant differences between a self-care book and either spinal manipulation or McKenzie exercise on symptom bothersomeness (0 to 10 scale) and RDQ scores, though trends favored spinal manipulation by about one point on both scales at 4 and 12 weeks367. For back pain of at least six weeks duration, a third higher-quality trial found massage, but not acupuncture, superior to a self-care book and videotape advice at 10 weeks in patients with low back pain for at least one week (difference of about 1 point on a 0 to 10 symptom bothersomeness scale and 2.5 points on the RDQ), though no differences between the self-care book and the other two interventions were observed after one year369. A fourth
54
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain higher-quality trial found a self-care book and weekly information packets inferior to weekly cognitive-behavioral therapy for long-term disability and the number of health care visits (but not for pain or functional status)309. Even though this trial was rated higher-quality because it met more than half of the quality criteria, it had an important flaw. About 20% of the patients randomized to cognitive-behavioral therapy never participated in the intervention and were excluded from the analysis. This could result in overestimates of benefits from cognitivebehavioral therapy if subjects who withdrew prior to receiving the intervention were less likely to respond to therapy. One lower-quality trial of patients with acute or subacute back pain found no short-term differences between a self-care book and physician advice to exercise on either a combined pain and function scale or the Aberdeen pain scale363.
Table 14. Trials comparing a self-care book to other interventions
Author, year Duration of LBP Cherkin, 1998367 >7 days
Cherkin, 2001369 Subacute or chronic
n=262 1 year
Main results Self-care book vs. spinal manipulation vs. McKenzie exercise Symptom bothersomeness (0 to 10 scale), mean scores: 3.1 vs. 1.9 vs. 2.3 at 4 weeks (NS), 3.2 vs. 2.0 vs. 2.7 at 12 weeks (NS), no differences at 2 years RDQ score (0 to 24 scale), mean scores: 4.9 vs. 3.7 vs. 4.1 at 4 weeks (NS), 4.3 vs. 3.1 vs. 4.1 at 12 weeks (NS), no differences at 2 years Proportion reporting reduced activity in 11 months after intervention: 36% vs. 33% vs. 35% Proportion needing bed rest: 9% vs. 8% vs. 11% Proportion who missed work: 17% vs. 7% vs. 13% Visits for back pain in second year after intervention: 24% vs. 29% vs. 20% Total costs over 2 years: $153 vs. $429 vs. $437 Self care book vs. acupuncture vs. massage Symptom bothersomeness (0 to 10 scale), mean scores: 4.6 vs. 4.0 vs. 3.6 at 10 weeks (p=0.01 for self care book versus massage, no other significant differences), 3.8 vs. 4.5 vs. 3.2 at 1 year (p=0.002 for acupuncture vs. massage, no other significant differences) RDQ score (0 to 24 scale), mean scores: 8.8 vs. 7.9 vs. 6.3 at 10 weeks (p<0.001 for self care book vs massage, p=0.01 for acupuncture vs. massage, p=0.75 for self care book vs. acupuncture), 6.4 vs. 8.0 vs. 6.8 at 1 year (p=0.05 for acupuncture vs. massage, no other significant differences) Provider visits:1.5 vs.1.9 vs.1.0 (p=0.17)
Quality score* 7/9
8/9
55
EVIDENCE REVIEW
Table 14. Trials comparing a self-care book to other interventions

Author, year Duration of LBP Linton, 2000309 Unspecified
Little, 2001363 Acute or subacute
n=311 3 weeks
Sherman, 2005371 Chronic
n=101 26 weeks
Main results Self care book vs. weekly information package vs. cognitive behavioral therapy Average pain (0 to 10, mean change from baseline): 0.8 vs. 0.8 vs. 0.9 Pain free days (0 to 7, mean change from baseline): 0.9 vs. 0.9 vs. 0.7 Days of sick leave in last six months (0 to 184, mean change from baseline): +10.0 vs. +14.4 vs. -0.4 Doctor visits in last six months (0 to 11, mean change from baseline): +0.5 vs. +0.4 vs. -0.5 Activities of Daily Living (0 to 60, mean change from baseline): -0.2 vs. +0.8 vs. +0.6 Modified Fear Avoidance Behavior Questionnaire (0 to 24, mean change from baseline): -2.0 vs. -2.7 vs. -3.5 Long-term disability: 10.4% (information package plus self-care book groups combined) vs. 1.1%, RR 9.3, 95% CI 1.2 to 70.8 Self-care book vs exercise advice vs. both (mean changes versus control) Pain/function scale (0 to 100): -8.7 vs -7.9 vs -0.1 at 1 week, -6.3 vs -1.4 vs -4.0 at 3 weeks (NS) Aberdeen pain and function scale (0 to 100): -3.8 vs -5.3 vs. -1.9 at 1 week (NS) Yoga vs. self-care book, mean differences RDQ score (0 to 24 scale): -2.6 (-4.6 to -1.6) at 6 weeks, -3.6 (-5.4 to -1.8) at 26 weeks Symptom bothersomeness score (0 to 10 scale): -1.6 (-2.6 to -0.5) at 6 weeks, and -2.2 (-3.2 to -1.2) at 26 weeks Exercise vs self-care book, mean differences RDQ score (mean difference): -1.7 (-3.7 to 0.4) at 6 weeks, -2.1 (-4.1 to -0.1) at 26 weeks Symptom bothersomeness score: -0.9 (-1.9 to -0.1) at 6 weeks, -0.8 (-2.1 to 0.5) at 26 weeks Yoga vs. exercise vs. self-care Visits to health care providers for low back pain: 4/34 (12%) vs 6/32 (19%) vs 9/29 (31%) at 26 weeks (NS) Medication use at week 26: 21% vs. 50% vs. 59% (p<0.05 for yoga vs. exercise or self-care) SF-36: No differences
Quality score* 5/9
4/9
8/9
Efficacy of different methods for providing information in self-care books

For back pain of unspecified duration, one higher-quality trial found no differences between a self-care book plus a 15-minute nurse education session and brief telephone follow-up versus a self-care book alone in RDQ scores, symptom bothersomeness scores, days lost from work, or number of health care visits (Table 15)368. However, patients in the nurse education group perceived themselves to be more knowledgeable and a higher proportion reported they had tried exercises recommended in the booklet (74% vs. 45%, p<0.001) in the first week after the intervention. A second higher-quality trial found no differences in pain or functional status through one year in patients with acute or subacute low back pain randomized to an experimental back book (the Back Book, developed to accompany the UKs 1996 Royal College of General Practitioners guidelines) aimed at changing beliefs and behaviors, compared to a
56
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain traditional self-care book mainly targeted at providing factual information366. However, patients randomized to the experimental book were more likely to report at least a 4-point reduction in fear avoidance beliefs, and patients with high baseline fear avoidance beliefs were more likely to report improvements of at least three points on the RDQ score.
Table 15. Trials evaluating different methods of providing information in a self-care book
Author, year Burton, 1999374
Cherkin, 1996368
n=300 1 year
Main results Experimental self-care book (The Back Book) vs. traditional self-care book Pain at worst (0 to 100), mean scores: 53.9 vs. 53.9 at 2 weeks, 50.9 vs. 50.8 at 1 year (NS) RDQ scores: No significant differences, data not reported Fear avoidance beliefs score, >4 point improvement: RR 2.72 (1.57 to 4.72) at 2 weeks, RR 1.47 (1.02 to 2.11) at 1 year Self-care book vs. nurse education + self-care book vs. usual care (mean change from baseline) RDQ score (0 to 24 scale): -5.4 vs -5.2 vs -5.3 (NS) at 1 week Symptom bothersomeness score (0 to 10 scale): -3.3 vs -3.3 vs -3.6 (NS) at 1 week Health care visits for low back pain: 45% vs. 46% vs. 47% in first 7 weeks after intervention (NS) Work loss days: 24% vs. 36% vs. 29% in first 7 weeks after intervention (NS)
Quality score* 6/9
6/9
Harms Three trials that reported information on adverse events reported none with a self-care book367, 369, 371 . Costs One trial estimated an average total cost (including the cost of the intervention and health care utilization) lower with a self-care book ($153) compared to either chiropractic therapy or physical therapy (around $430)367. Another trial found no significant differences in estimated costs between a self-care book ($200), massage ($139), and acupuncture ($252)369.
Summary of evidence
For acute or subacute low back pain or back pain of unspecified duration, five trials (two higher-quality) found no differences between a self-care book and usual care in pain or symptom bothersomeness scores (level of evidence: fair). In patients with back pain of varying duration, four trials (three higher-quality) that compared a self-care book to acupuncture, exercise, exercise advice, or manipulation found no significant differences, or the self-care book was only slightly inferior on symptom bothersomeness scores and functional status. Larger differences were seen in single higher-quality trials that found a self-care book inferior to yoga and to massage (level of evidence: good).
57
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For acute or subacute low back pain, there was no difference between a self-care book and advice to exercise in one lower-quality trial (level of evidence: poor).
Different methods for providing information in a self-care book were not associated with
significant differences in pain or functional status. A brief nurse education visit increased the proportion of patients who exercised compared to the self-care book without the nurse education visit in one higher-quality trial. In another higher-quality trial of patients with acute or subacute low back pain, an experimental self care book targeted at changing beliefs and behaviors reduced fear avoidance beliefs more than a traditional self-care book (level of evidence: fair).

The other guidelines do not specifically address self-care books. General recommendations on advice are listed in the bed rest section.
Self-care e-mail discussion groups

Results of search: systematic reviews We identified no relevant systematic reviews. Results of search: trials From 611 potentially relevant citations, we identified one lower-quality trial comparing a selfcare e-mail discussion group versus usual care for chronic low back pain375.
Efficacy of an e-mail discussion group versus usual care

One trial found that participation in a closed, moderated e-mail discussion group (along with a self-care book and videotape) was slightly superior to usual care for pain (p=.045), back-specific functional status (p=.02), role function (p=.007), and health distress (p=.001) after 12 months compared to usual care (Table 16)375. Differences averaged about 1 point on the 24 point RDQ and about 0.5 points on a 10 point pain scale. There were no differences in physician visits for back pain or average number of hospital days over a 12-month period.
Table 16. Trial of e-mail discussion group versus usual care
Author, year Lorig, 2002375
Main results E-mail discussion, book and video vs. usual care (mean changes from baseline at 12 months) RDQ (0 to 23): -2.77 vs -1.51 (p=.01) Health distress (0 to 5): -0.92 vs -0.57 (p=.001) Pain interference (0 to 10): -1.50 vs -1.02 (p=.05) Role function (0 to 7): -0.83 vs -0.53 (p=.007) Physician visits for back in last 6 months: -1.54 vs -0.65 (NS) Chiropractor visits for back in last 6 months: -1.32 vs -0.797 (NS) Physical therapist visits for back in last 6 months: -1.99 vs -1.31 (NS) Hospital days in recent 6 months: -0.198 vs 0.04 (NS)
Quality score* 2/9
58
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
For chronic low back pain, one lower-quality trial found an e-mail discussion group intervention plus a self-care book and videotape slightly superior to usual care for pain, disability, role function and health distress after one year (level of evidence: poor).

The other guidelines do not specifically address e-mail discussion groups. General recommendations on advice are listed in the bed rest section.
Self-care exercise videotape

Results of search: systematic reviews We identified no relevant systematic reviews. Results of search: trials From 611 potentially relevant citations, we identified one lower-quality trial comparing a selfcare exercise videotape to face-to-face instruction for back pain of unspecified duration376.
Efficacy of self-care exercise videotape versus face-to-face advice

One lower-quality trial found no differences between a self-care exercise video (featuring either the treating physiotherapist or an anonymous physiotherapist) and face-to-face advice in RDQ scores after 4 to 6 weeks (Table 17)376. On one subscale of the SF-36 (pain), the self-care video group had greater improvements than the face-to-face advice group (p<0.005, absolute differences not reported).
Table 17. Trial of self-care video advice versus face-to-face advice
Number of patients Duration of follow-up n=550 4-6 weeks
Author, year Miller, 2004376
Main results Self-care video with treating physiotherapist vs. selfcare video with anonymous physiotherapist vs. face-toface advice RDQ score (0 to 24), mean change: -3.58 vs -3.00 vs -2.47. Neither video group improved more than the face-to-face active group (p=.06) SF-36 pain subscale: Either video intervention experienced greater improvement compared to face-to-face advice (p<0.005, data not reported)
Quality score* 3/9
59
EVIDENCE REVIEW
Summary of evidence
For back pain of unspecified duration, one-lower quality trial found no differences in functional status between videotaped and face-to-face exercise advice through 4 to 6 weeks (level of evidence: poor).

The other guidelines do not specifically address videotaped exercise advice. General recommendations on advice are listed in the bed rest section.
Advice to restrict early morning flexion

Results of search: systematic reviews We identified no relevant systematic reviews. Results of search: trials From 611 potentially relevant citations, we identified one lower-quality trial that compared advice to restrict early morning lumbar flexion to sham exercise advice for chronic low back pain377, 378.
Efficacy of advice to restrict early morning flexion versus sham exercise advice
One lower-quality trial found a single 45-minute instructional session on restriction of early morning flexion (with supplemental videotape and written instructions) superior to sham exercise advice for mean pain intensity, days with disability, as well as medication use (Table 18)377, 378. Results are difficult to interpret because of large baseline differences between groups (baseline medication use and disability days two times higher in the sham exercise advice group).
Table 18. Trial of advice to restrict early morning flexion versus sham exercise advice
Author, year Snook, 1998377, 378
Main results Advice to restrict early morning flexion vs. sham exercise advice (mean at 6 months) Pain intensity (0 to 10): 1.52 vs. 1.36 (p<0.05) Pain days: 102 vs. 150 Disability days: 3.0 vs. 10.7 Impairment days: 3.0 vs. 10.7 Medication days: 16.7 vs. 49.9
Quality score* 2/10
*Excludes criteria involving blinding of care providers, for maximum score of 10
Harms No trial reported adverse events.

60
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Costs We found no studies evaluating costs.
Summary of evidence
For chronic low back pain, one lower-quality trial found advice to restrict early morning flexion superior to sham exercise advice for pain intensity and disability, but these findings are difficult to interpret because of marked baseline differences between groups (level of evidence: poor).

The other guidelines do not specifically address advice to restrict early morning flexion. General recommendations on advice are listed in the bed rest section.
Lay-facilitated groups for self-care

We defined lay-facilitated groups as sessions run by non-medical professionals, either with or without training in self-care group facilitation or specific self-care approaches for low back pain management. Results of search: systematic reviews We identified no relevant systematic reviews. Results of search: trials From 611 potentially relevant citations, we identified two trials that evaluated lay-facilitated selfcare groups versus wait-list control379 or usual care380.
Efficacy of lay-facilitated groups for self-care versus usual care

In patients invited to enroll 6 to 8 weeks after presentation with low back pain, one higher-quality trial found a four-session lay-facilitated self-care group supplemented by a self-care book and videotapes slightly superior to usual care plus a self-care book on RDQ scores (difference about 1.5 points) after 6 months, though not after 3 or 12 months380. A higher proportion of patients in the self-care group reported a >50% reduction in RDQ scores at 6 months, but there was no difference between groups in pain intensity. A lower-quality trial found a 6-week lay-facilitated self-care group (2 hours each session) no better than wait-list control for chronic low back pain in older adults (60 years or older)379.
61
EVIDENCE REVIEW
Table 19. Trials of lay-led self-care group versus wait-list control or usual care
Author, year Haas, 2005379
Von Korff, 1998380
n=255 12 months
Main results Lay-led group vs wait-list control Modified Von Korff pain score, mean (0 to 100): 41.4 vs 42.3 (p=.059),adjusted mean difference -1.0 (p=.835) at 6 months (NS) Modified Von Korff disability score, mean (0 to 100): 32.8 vs 35.8 (p=.303) at 6 months (NS) Lay-led group + self-care book vs usual care + selfcare book RDQ Questionnaire (0 to 24), mean score: 6.56 vs 7.40 at 3 months (NS), 5.83 vs 7.23 at 6 months (p=0.007), 5.75 vs 6.75 at 12 months (p=0.092). 50% decrease in RDQ score: 48% vs. 33% (p=0.02) at 6 months Pain intensity (0 to 10), mean score: 3.87 vs. 4.02 at 3 months, 3.22 vs. 3.79 at 12 months (NS)
Quality score 3/9
5/9
Harms No trial reported adverse events. Costs A cost analysis based on the trial estimated a mean cost of $9.70 per additional low-impact back day in the lay-led group relative to usual care381.
Summary of evidence
For subacute or chronic low back pain, a four-session lay-led self-care group was associated with small improvements in functional status (but not pain intensity) compared to usual care after 6 months (but not 3 or 12 months) in one higher-quality trial (level of evidence: fair). For elderly patients with chronic low back pain, a six-session lay-led self-care group was associated with no differences in pain or function compared to wait-list controls in one lowerquality trial (level of evidence: poor).

The other guidelines do not address lay-led self-care groups.
Self-help tools for back surgery decisions

Results of search: systematic reviews We identified no systematic reviews on tools for helping guide back surgery decisions. Results of search: trials From 611 potentially relevant citations, we identified one higher-quality trial that compared patient outcomes associated with a video and a self-care book for informing back surgery decisions versus a self-care book alone382. We excluded another higher-quality trial on a video program for informing surgery decisions because it did not evaluate clinical outcomes383.
62
EVIDENCE REVIEW
Efficacy of a video plus self-care book for informing back surgery decisions versus a self-care book alone
In potential back surgery candidates, one higher-quality trial found no difference in back-specific functional status between an interactive video plus self-care book and a self-care book alone through 1 year (Table 20)382. The video intervention was superior to the self-care book alone for the proportion of patients reporting extreme or quite a bit of pain (28% versus 37%, p=0.04). However, no difference was found between the interventions for resolution of back or leg pain at 3 months or 1 year. There was no difference in the proportion of patients who underwent surgery except for in those diagnosed with herniated disc, who were less likely to have surgery if randomized to the interactive video (32% vs. 47%, p=0.05).
Table 20. Trial of interactive video + self-care book versus self-care book alone for informing surgical decisions
Author, year Deyo, 2000382
Main results Videodisc program + booklet vs booklet alone RDQ Score: no differences between groups at 3 months or 1 year Back pain severity 'extreme' or 'quite a bit' at 1 year: 27.6% vs. 37.2% (p=0.04) Resolution of back or leg pain: no differences between groups at 3 months or 1 year Surgery rate: 26% vs 33% (p=0.08, NS). In those with herniated disks: 32% vs 47% (p=0.05) Health care utilization (Seattle patients only): Except for surgery data reported above, no differences between groups for number of physician visits, physical therapy, spine imaging, overall lab or pharmacy use, hospitalizations for back pain. Satisfaction with treatment, decision-making process: no differences Satisfaction with amount of information received: 71.8% vs 57.1% (p=0.005)
Quality score* 6/9
Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
In patients considered candidates for surgery, one higher-quality trial found no differences in function 1 year after randomization to an interactive video plus self-care book versus a selfcare book alone for informing back surgery decisions. A lower proportion of patients with herniated disc randomized to the interactive video underwent surgery. The video was associated with a lower proportion of patients with severe pain at one year, though there was no difference in rates of resolution of back or leg pain (level of evidence: fair).
63
EVIDENCE REVIEW

The other guidelines do not address tools to help patients with back care decisions.
Other self-care interventions Lumbar supports

We defined lumbar support as a back brace, corset, or orthotic device worn to passively support the back. Results of search: systematic reviews We identified one higher-quality Cochrane review evaluating effectiveness of lumbar supports for low back pain384, 385. We excluded two outdated systematic reviews343, 386 and one systematic review that didnt evaluate efficacy or safety387. Results of search: trials The Cochrane review included six trials (two higher-quality) of lumbar supports for treatment of low back pain384, 385. We did not search for additional trials.
Efficacy of lumbar supports versus no lumbar support

For low back pain of unspecified duration, the Cochrane review included one small (n=30), lower-quality trial388 that found a lumbar support superior to no intervention for improvement in pain after 1 hour, 3 weeks, and 6 weeks in patients384, 385.
Efficacy of lumbar supports versus other interventions

Three389-392 of four393 trials included in the Cochrane review384, 385 did not find lumbar supports to be more effective than a variety of other non-invasive interventions in reducing pain or improving functional outcomes, or rates of return to work. In the only higher-quality trial390, 391, a lumbar support was superior to minimal massage in patients with subacute or chronic low back pain on the RDQ, but there were no significant differences on the revised ODI or in pain relief. There were no differences between a lumbar support and spinal manipulation or transcutaneous muscular stimulation. Two lower-quality trials found no differences between lumbar supports and usual care (for chronic low back pain389) or either spinal manipulation, physiotherapy (any technique except manipulation), or acetaminophen (for back pain of varying duration393). One lower-quality trial found a lumbar support superior to advice on rest and lifestyle for pain relief, return to work, and overall improvement in patients with acute low back pain392.
Efficacy of one type of lumbar support versus another

The Cochrane review384, 385 included one higher-quality trial that found a lumbar support with a rigid insert associated with significantly more global improvement than a lumbar support without a rigid insert394. Harms No trial reported adverse events.
64
EVIDENCE REVIEW
Summary of evidence
For acute low back pain, one lower-quality trial found a lumbar support superior to advice on lifestyle and bed rest for pain relief, return to work, and overall improvement (level of evidence: poor). For low back pain of unspecified duration, there is insufficient evidence from one lower-quality trial to determine whether lumbar supports are effective compared to no intervention (level of evidence: poor). For low back pain of varying or unspecified duration, three trials (one higher-quality) found no clear differences between lumbar supports and other interventions (minimal massage, spinal manipulation, physiotherapy with any intervention other than manipulation, acetaminophen, TENS, or usual care). Most comparisons were evaluated in only one lower-quality trial (level of evidence: poor to fair). For chronic low back pain, one higher-quality trial found a lumbar support with a rigid insert associated with superior global improvement compared to a support without a rigid insert (level of evidence: fair).

The AHCPR guidelines found that lumbar corset and support belts had not been proven beneficial for treating patients with acute low back problems (strength of evidence: D). The VA/DoD and UK RCGP guidelines make similar recommendations. The European COST guidelines found insufficient evidence to recommend lumbar supports for nonspecific chronic low back pain.
Mattresses
Results of search: systematic reviews We identified no relevant systematic reviews. Results of search: trials From 198 potentially relevant citations, we identified two randomized395, 396 and one quasirandomized trial397 on efficacy of different mattress types for chronic low back pain. One trial was rated higher-quality395.
Efficacy of different mattress types

For chronic low back pain, one higher-quality trial (n=313) found a medium-firm mattress associated with greater likelihood for improvement in pain-related disability after 90 days compared to those randomized to a firm mattress (82% vs. 68%, p=0.005)395. In unadjusted analyses, there were no differences between mattresses in the proportion of patients with improvement in pain while lying in bed or on rising (Table 21). The medium-firm mattress was superior for pain while lying in bed when results were adjusted for perceived firmness of the new
65
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain mattress and baseline pain scores. One lower-quality randomized trial compared a soft interior sprung mattress to an isometric mattress396 and a quasi-randomized trial compared four different mattresses (orthopedic hard, standard, waterbed, hybrid water-foam)397. However, we could not reliably interpret results because of methodological flaws, use of nonstandardized outcome measures, and poor reporting of outcomes.
Table 21. Trials of different mattresses in patients with low back pain
Number of patients Duration of follow-up n=30 2 weeks followed by crossover n=15 2 weeks per intervention n=313 90 days
Author, year Atherton, 1983396
Main results Isometric versus soft inferior sprung mattress Proportion reporting pain least: 40% (10/25) vs. 28% (7/25) Proportion reporting comfort best: 40% (10/25) vs. 52% (13/25)
Quality score 4/11
Garfin, 1981397
Kovacs, 2003395
Orthopedic hard mattress versus standard box spring and mattress versus water-filled mattress versus hybrid (combination water-foam) mattress Results not interpretable Medium-firm versus firm mattress Proportion with improvement in pain-related disability: 82% vs. 68%, p=0.005; adjusted OR=2.10 (95% CI 1.24 to 3.56) Proportion with improvement in pain while lying in bed: 83% vs. 78%, p=0.29; adjusted OR=2.36 (95% CI 1.13 to 4.93) Proportion with improvement in pain on rising: 86% vs. 80%, p=0.20; adjusted OR=1.93 (95% CI 0.97 to 3.86)
0/11
11/11
Harms The higher-quality trial found firm mattress associated with a higher proportion of patients with worsening of pain in bed (17% vs. 9.0%) and worsening of disability (24% vs. 9%) compared to the medium-firm mattress395. Costs We found no studies evaluating costs.
Summary of evidence
For chronic low back pain, one higher-quality trial found a firm mattress slightly inferior to a medium-firm mattress for pain-related disability and pain while in bed. There were no differences in other pain outcomes (level of evidence: fair). There was insufficient evidence to judge the relative effectiveness of other mattress types or in patients with acute low back pain (level of evidence: poor).

The other guidelines do not address mattress types.
66
EVIDENCE REVIEW
Superficial heat or cold

Superficial heat or cold is the application of warm packs, heated blankets, cold packs, or ice superficially to the back. It may also be referred to generally as thermotherapy. Results of search: systematic reviews We identified one recent, higher-quality Cochrane review (9 trials, five rated higher-quality)398. We excluded one older systematic review that searched for but found no studies on superficial hot or cold399. Results of search: trials The Cochrane review included nine trials (five rated higher-quality)398. All nine trial evaluated application of superficial heat. Two trials also evaluated application of superficial cold. The same investigator led three of the trials400-402. We did not search for additional trials.
Efficacy of superficial heat versus placebo

For acute or subacute low back pain, the Cochrane review398 included two higher-quality trials400, 402 that found heat wrap therapy moderately more effective than placebo for short-term (5 days) pain relief (WMD=1.06, 95% CI 0.68 to 1.45 on a 0 to 5 scale) and improvement in disability (WMD=-2.10, 95% CI -3.19 to -1.01 on the RDQ). Another higher-quality trial403 found application of a heated blanket substantially decreased acute low back pain immediately following application compared to a non-heated blanket (WMD=-32.20, 95% CI -38.69 to -25.71 on a 100 point scale).
Efficacy of superficial heat versus other interventions

For acute low back pain, the Cochrane review398 included one higher-quality trial401 that found heat wrap therapy moderately superior to oral acetaminophen or ibuprofen for short-term pain relief (mean differences=0.68 and 0.49 points, respectively, on a 0 to 5 scale after 1 day and mean differences=0.66 and 0.93 after 3 to 4 days, p<0.05 for all differences) and RDQ scores (difference=2 and 2.2 points after 4 days, p<0.05). For subacute or acute low back pain, another higher-quality trial404 included in the Cochrane review398 found heat wrap therapy moderately superior to an educational booklet for early pain relief (WMD=0.60, 95% CI 0.05 to 1.15 after 2 days on a 0 to 5 scale and WMD=1.10, 95% CI 0.55 to 1.65 after 4 days) and improved function (WMD=0.40, 95% CI -1.15 to 0.95 after 2 days and WMD=0.30, 95% CI -0.41 to 1.01 after 4 days), though benefits were no longer present after a week. There were no significant differences between heat wrap therapy and McKenzie exercise.
Efficacy of superficial cold versus placebo

We identified no trials that compared superficial cold versus placebo or no treatment.
Efficacy of superficial cold versus other interventions

For chronic low back pain, the Cochrane review398 included one lower quality trial405 that found light ice massage and transcutaneous electrical stimulation similarly effective in reducing pain.
67
EVIDENCE REVIEW
Efficacy of superficial heat versus superficial cold

Two lower-quality, non-randomized trials406, 407 included in the Cochrane review398 reported conflicting results for superficial heat versus cold in patients. One trial found no significant differences between hot packs and ice massage for back pain of mixed duration406 and the other found ice massage superior to hot packs for chronic low back pain407. Harms No serious adverse events were reported in trials of heat wrap therapy398. Minor adverse events mainly consisted of skin irritation or increased pinkness. Costs One decision analysis compared the cost-effectiveness of heat wrap therapy relative to ibuprofen or acetaminophen in patients with acute low back pain408. It found heat-wrap therapy dominated over both drugs (decreased costs and superior efficacy), with conclusions insensitive to changes in parameters. This analysis relied on outcomes data from a single published trial401.
Summary of evidence
For acute or subacute low back pain, there is consistent evidence from three higher-quality trials that heat wrap therapy or a heated blanket is moderately superior to placebo or a nonheated blanket for short-term pain relief and back-specific functional status (level of evidence: good). For acute low back pain, heat wrap therapy was moderately superior to acetaminophen or ibuprofen for short-term pain relief in one higher-quality trial (level of evidence: fair). In patients with a mix of acute and subacute low back pain, heat wrap therapy was superior to a self-care booklet, but not exercise, in one higher-quality trial (level of evidence: fair). There is insufficient evidence (one lower-quality trial) to judge efficacy of superficial cold (level of evidence: poor). There is conflicting evidence from two lower-quality, non-randomized trials on efficacy of superficial heat versus superficial cold (level of evidence: poor).

The AHCPR guidelines found physical agents and modalities (including superficial heat or cold) of insufficiently proven benefit to justify their cost for acute low back pain (strength of evidence: C). However, they suggest that self-application of heat or cold to the back could be taught to the patient as an option. The VA/DoD and UK RCGP guidelines reached similar conclusions. The European COST guidelines make no recommendation for superficial heat or cold for acute low back pain, but note that three trials came from one research group with potential conflict of interest.
68
EVIDENCE REVIEW
The European COST guidelines found insufficient evidence to recommend superficial heat for
chronic low back pain.
69
EVIDENCE REVIEW
Table 22. Systematic reviews on efficacy of self-care therapies for low back pain
Overall quality using Oxman scale (1 to 7) 7
Author, year Hagen, 2002359
Type of systematic review Qualitative
Number of included trials (number rated higherquality) * 4 (3)
Number of trials not included in any other relevant systematic review 0 (see Hagen 2005)
Duration of treatment in included trials 4 to 12 weeks
Sample sizes in included trials 42 to 186
Interventions evaluated (number of trials) Advice to remain active (4)
Main conclusions Advice to remain active versus exercise (1 RCT): no differences in pain intensity; WMD=-8.6 points (95% CI -13.0 to -3.3) for ODI at 1-3 weeks (1 RCT), but no differences at 4 to 12 weeks; reduced sick leave at 1-3 weeks (WMD= -1.6 days; -3.5 to 0.3) and at 4 to 12 weeks (WMD=-2.5 days, 95% CI -5.6 to 0.6) Acute low back pain without sciatica: Advice to remain active vs. advice to rest in bed: SMD= 0.22 (95% CI 0.02 to 0.41) for pain at 3-4 weeks (2 RCTs) and SMD=0.25 (95% CI 0.05 to 0.45) at 12 weeks; SMD=0.29 (95% CI 0.09 to 0.49) for function (2 RCTs) at 3-4 weeks and SMD=0.24 (95% CI 0.04 to 0.44) at 12 weeks (2 RCTs); SMDs equivalent to 5 to 7.5 mm VAS and 1.2 to 1.8 points on RDQ; bed rest also increases length of sick leave during the first 12 weeks (high quality evidence) Sciatica: Advice to remain active vs. advice to rest in bed: SMD= -0.03 (95% CI -0.24 to +0.18) for pain at 3-4 weeks (2 RCTs) and 0.10 (95% CI -0.12 to 0.31) at 12 weeks (2 RCTs);SMD=0.19 (-0.02 to +0.41) for function at 3-4 weeks and SMD=0.12 (95% CI -0.10 to +0.33) at 12 weeks
Advice to remain active (six trials in two systematic reviews; see also Advice to rest in bed)
Advice to rest in bed (11 trials in one systematic review; see also Advice to remain active) Hagen, 200565 Quantitative 11 (8) Not applicable 9 days to 6 months (median=12 weeks) 40 to 398 (median= 186) Advice to rest in bed (11); advice to remain active (6) 7
EVIDENCE REVIEW
Table 22. Systematic reviews on efficacy of self-care therapies for low back pain
Author, year Jellema, 2001385; Van Tulder, 384 2000
Number of included trials (number rated higherquality) * 6 trials of treatment (2)
Number of trials not included in any other relevant systematic review Not applicable
Duration of treatment in included trials 3 to 8 weeks (median=3.5 weeks)
Sample sizes in included trials 19 to 334 (median= 190)
Interventions evaluated (number of trials) Lumbar support with rigid stay (2), pneumatic lumbar support (1), other or not specified (3) Superficial heat (9), superficial cold (2)
Main conclusions Insufficient evidence to assess efficacy of lumbar support versus no treatment (1 RCT); lumbar support superior to other interventions in 1 of 4 RCTs
Lumbar supports (six trials in one systematic review)
Superficial heat (9 trials in 1 systematic review) French, 398 2006 Quantitative 9 (5) Not applicable Single application to 7 days 36 to 371 (median= 90) Heat wrap versus oral placebo or non-heated wrap for acute or subacute LBP (4 RCTs): WMD=1.06 (95% CI 0.68 to 1.45 on a 0 to 5 scale) for pain relief up to day 5 (2 RCTs); WMD=-2.10 (95% CI -3.19 to -1.01) for RDQ (2 RCTs) Insufficient evidence to assess efficacy of superficial heat versus superficial cold 7
*Trials adequately meeting at least half of the quality rating criteria or rated as good or higher-quality if the number of criteria met was not reported CI=confidence interval, ODI=Oswestry Disability Index, LBP=low back pain, OR=odds ratio, RCT=randomized controlled trial, RDQ=Roland-Morris Disability Questionnaire, RR=relative risk, TENS=transcutaneous electrical nerve stimulation, WMD=weighted mean difference
EVIDENCE REVIEW
Table 23. Summary of evidence on self-care therapies for acute low back pain
Number of trials (number rated higherquality by at least one systematic review) 6 (5) Effective vs. placebo, sham, wait list, or no treatment? No evidence
Intervention Advice to remain active
Net benefit* Small (no significant harms or burdens) Not effective
Inconsistency? No
Directness of evidence? Direct
Overall quality of evidence Good
Comments Advice to remain active superior to advice to rest in bed in 6 trials Advice to rest in bed inferior to advice to remain active in 6 trials
Advice to rest in bed Lumbar supports Self-care education book
8 (6)
No evidence
No
Direct
Good
1 (0) 10 (7) See comments
Unable to estimate Small (no significant harms or burdens)
No evidence No evidence
Not applicable No
Direct Direct
Poor Fair Nearly all trials evaluated patients with low back pain of mixed (acute, subacute, or chronic) or unspecified duration. Self-care education book similar to usual care in 5 trials One poor-quality trial evaluated self-care exercise videotape in patients with low back pain of unspecified duration
Self-care exercise videotape
No evidence
Unable to estimate
No evidence
Not applicable
Direct
Poor
Superficial heat
5 (5)
Moderate
Yes (2 trials)
No
Direct
Good
* Based on evidence showing medication is more effective than placebo, and/or evidence showing medication is at least as effective as other medications or interventions thought to be effective, for one or more of the following outcomes: pain, functional status, or work status. Versus placebo, small benefit defined as 5-10 points on a 100-point Visual Analogue Scale (VAS) for pain (or equivalent), 1-2 points on the Roland Morris Disability Questionnaire (RDQ), 10-20 points on the Oswestry Disability Index (ODI), or a standardized mean difference (SMD) of 0.2-0.5. Moderate benefit defined as10-20 points on a VAS for pain, 2-5 points on the RDQ, 10-20 points on the ODI, or a SMD of 0.5-0.8. Large benefit defined as >20 points on a 100 point VAS for pain; >5 points on the RDQ, >20 points on the ODI, or a SMD of >0.8. Inconsistency defined as <75% of trials reaching consistent conclusions on efficacy (no effect vs. positive effect considered inconsistent)
EVIDENCE REVIEW
Table 24. Summary of evidence on self-care therapies for chronic or subacute low back pain
Number of trials (number rated higherquality by at least one systematic review) 2 (2) Effective vs. placebo, sham, wait list, or no treatment? No evidence
Intervention Advice to remain active Advice to rest in bed Advice to restrict early morning flexion Lay-led selfcare groups
Net benefit* Small
Important Inconsistency? No
Overall quality of evidence Fair
Comments Advice to remain active similar in effectiveness to exercise therapy in 2 trials
No evidence 1 (0)
No evidence Unable to estimate Unable to estimate
No evidence Unable to estimate (1 trial) Unable to estimate (1 trial)
Not applicable Not applicable
Not applicable Direct
No evidence Poor
2 (1)
Yes
Direct
Poor
Lay-led self-care group superior to usual care on some outcomes in 1 higher-quality trial, but no differences versus wait-list control in 1 lowerquality trial
Lumbar supports Mattresses
2 (1) 3 (1)
Unclear Not effective
Some inconsistency No
Direct Direct
Poor Fair Medium-firm mattress slightly superior to firm mattress in one higher-quality trial Nearly all trials evaluated patients with low back pain of mixed (acute, subacute, or chronic) or unspecified duration. Self-care education book similar to usual care in 5 trials
Self-care education book
10 (7) See comments
Small (no significant harms or burdens)
No evidence
No
Direct
Fair
Self-care e-mail discussion group Self-care exercise videotape
1 (0)
Unable to estimate Unable to estimate
No evidence
Not applicable
Direct
Poor
No evidence
No evidence
Not applicable
Direct
Poor
One poor-quality trial evaluated self-care exercise videotape in patients with low back pain of unspecified duration
EVIDENCE REVIEW
Table 24. Summary of evidence on self-care therapies for chronic or subacute low back pain
Number of trials (number rated higherquality by at least one systematic review) 1 (1) Effective vs. placebo, sham, wait list, or no treatment? Yes (1 trial)
Intervention Self-help tools for back surgery decisions Superficial heat
Net benefit* Small
Important Inconsistency? Not applicable
Comments No effect on functional outcomes, though fewer patients using self-help tool underwent surgery Three lower-quality trials
3 (0)
Unable to estimate
Unclear (3 trials)
No
Direct
Poor
EVIDENCE REVIEW
Table 25. Summary of evidence on self-care therapies for radiculopathy or sciatica

Number of trials (number rated higherquality by at least one systematic review) 2 (2)
Intervention Advice to remain active Advice to rest in bed Traction
Net benefit* Small (no significant harms or burdens) Not effective
Effective vs. placebo, sham, or no treatment? No evidence
Important Inconsistency? No
Comments Advice to remain active superior to advice to rest in bed in 2 trials Advice to rest in bed inferior to advice to remain active in 2 trials Other trials of traction included patients with back pain of varying duration
3 (2)
No evidence
No
Direct
Good
16 (4)
Not effective (continuous or intermittent traction) Small to moderate (autotraction)
No for continuous or intermittent traction (8 trials), yes for autotraction (2 trials)
Some inconsistency (for autotraction versus continuous or intermittent traction)
Direct
Fair
EVIDENCE REVIEW
Key Question 4 How effective are different non-invasive interventions for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? Medications Acetaminophen
Acetaminophen (paracetamol) is an anti-pyretic and analgesic medication without significant anti-inflammatory properties. It is believed to work in part by indirectly decreasing production of prostaglandins through inhibitory effects involving cyclo-oxygenase enzymes409, 410. Search results: systematic reviews We identified one lower-quality systematic review of multiple medications for low back pain that included trials of acetaminophen411. In addition, a higher-quality Cochrane review of NSAIDs for low back pain included trials comparing acetaminophen to NSAIDs412, 413. The systematic reviews each included three to five short-term (four weeks or less in duration) trials, only one of which was rated higher-quality414. We excluded two relevant but outdated systematic reviews193, 415 . Search results: trials A total of six unique trials353, 414, 416-419 of acetaminophen were included in two systematic reviews411-413. From 134 potentially relevant citations, we identified one higher-quality401 and two lower-quality393, 420 trials of acetaminophen for low back pain that met inclusion criteria and were not included in the systematic reviews. All three compared acetaminophen to other active interventions. Among all trials of acetaminophen, the longest was four weeks in duration. We excluded 13 trials that either did not specifically evaluate low back pain patients421, 422 or compared dual therapy with acetaminophen plus another drug to a different drug or drug combination423-433. One other trial is discussed in the section on dual therapy versus monotherapy434.
Efficacy of acetaminophen versus placebo

For acute low back pain, one lower-quality trial included in the Cochrane review found no difference between acetaminophen (three grams/day) and no treatment417.
Efficacy of acetaminophen versus NSAIDs

For acute low back pain, the Cochrane review included three lower-quality trials which reported conflicting results on efficacy of acetaminophen (up to four grams/day) versus NSAIDs412, 413. Two trials353, 417 found no differences, but a third trial found two out of four evaluated NSAIDs superior to acetaminophen416. One trial not included in the systematic reviews found acetaminophen 4000 mg/day similarly effective compared to ibuprofen 1200 mg/day401. For chronic low back pain, one higher-quality trial (included in both systematic reviews) found acetaminophen inferior to diflunisal for the proportion of patients reporting good or excellent
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain efficacy after four weeks (10 of 16 vs. 4 of 12, p=0.01), though the proportion reporting no or mild low back pain was similar (13 of 16 vs. 7 of 12)414. Although there are no other trials of acetaminophen versus NSAID for chronic low back pain, there is consistent evidence from higher-quality systematic reviews of patients with osteoarthritis that acetaminophen is slightly inferior for pain relief435-438. One trial of patients with back pain of mixed acute and chronic duration found no differences between acetaminophen and flurbiprofen418.
Efficacy of acetaminophen versus other interventions

Acetaminophen was moderately inferior to heat wrap therapy and similarly effective compared to ibuprofen in one higher-quality trial401. In single, lower-quality trials, acetaminophen was inferior to amitriptyline419 and electroacupuncture420. Other, mostly lower-quality trials found no difference between acetaminophen, codeine, phenylbutazone, or the combination of aspirin plus oxycodone for rates of return to work353 or between acetaminophen and either physical therapy, a corset, or spinal manipulation for pain or other assessed outcomes393. Harms Adverse events associated with acetaminophen were poorly reported in trials of patients with low back pain. In two higher-quality systematic reviews of osteoarthritis patients, acetaminophen was superior to NSAIDs for gastrointestinal tolerability and other GI side effects436, 438. There exists a perceived safety advantage of acetaminophen, but evidence from clinical trials on serious side effects such as bleeding, hypertension, and myocardial infarction are sparse. Observational data suggest that acetaminophen is associated with a lower rate of GI bleeding compared to NSAIDs439, 440, but may be associated with modest increases in blood pressure441-443 and renal dysfunction444. One recent analysis from the observational, large Nurses Health Study suggests that heavy use of acetaminophen may be associated with an increased risk of cardiovascular events similar in magnitude to heavy use of NSAIDs445. A recent trial found 31% to 44% of healthy patients randomized to treatments that included four grams of acetaminophen daily experienced serum alanine aminotransferase elevations greater than three times the upper limit of normal, compared to 0% with placebo, though the clinical significance of this finding is unknown446. Costs We found no studies evaluating costs.
Summary of evidence
For acute low back pain, there is conflicting evidence from four lower-quality trials regarding efficacy of acetaminophen versus NSAIDs, with three finding no difference in outcomes (level of evidence: fair). For chronic low back pain, one higher quality trial found acetaminophen inferior to an NSAID on an overall assessment of efficacy (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Multiple trials of patients with osteoarthritis consistently found acetaminophen slightly inferior to NSAIDs for pain relief (less than 10 points on a 100 point visual analogue pain scale) (level of evidence: good). There is insufficient evidence from single, lower quality trials that compared acetaminophen to other interventions (such as other medications, physical therapy, superficial heat, a corset, or spinal manipulation) to accurately judge relative efficacy (level of evidence: poor). Acetaminophen is associated with a lower risk of serious gastrointestinal adverse events compared to NSAIDs based primarily on observational data (level of evidence: fair). Acetaminophen is better tolerated than NSAIDs (level of evidence: good). Additional studies are required to evaluate whether high-dose acetaminophen is associated with increased cardiovascular risk (results available from a single observational study) (level of evidence: poor). Acetaminophen at 4 grams daily is associated with elevations in aminotransferase levels of 31% to 44% in healthy subjects, though the clinical significance of this finding is not known (level of evidence: fair).

The AHCPR guidelines found acetaminophen reasonably safe and acceptable for treating patients with acute low back problems (strength of evidence: C). The VA/DoD guideline recommendations are identical to AHCPRs. The UK RCGP guideline found that comparisons of effectiveness between acetaminophen and NSAIDs are inconsistent (strength of evidence: **). The European COST guidelines recommend acetaminophen as first choice when needed for pain relief in patients with acute low back pain.
Non-Steroidal Anti-Inflammatory Drugs (NSAIDs) Non-selective NSAIDs

NSAIDs are thought to reduce joint and muscle pain primarily by blocking the cyclo-oxygenase (COX)-2 enzyme447. However, non-selective NSAIDsor NSAIDs that block both the COX-1 and COX-2 enzymesalso cause gastrointestinal (GI) bleeding because the COX-1 enzyme helps protect the lining of the stomach from acid. Results of search: systematic reviews We identified two higher-quality systematic reviews evaluating efficacy of non-selective NSAIDs for non-specific low back pain411-413. The more comprehensive study (51 included trials) was a Cochrane review412, 413. A third, higher-quality systematic review evaluated efficacy of NSAIDs in patients with sciatica100. We excluded four outdated systematic reviews193, 346, 415, 448.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials Fifty-seven unique trials of NSAIDs were included in three systematic reviews411-413. We did not search for additional trials. Almost all of the trials were short-term. Only six of the 51 trials included in the Cochrane review were longer than two weeks in duration (the longest evaluated six weeks of therapy)412, 413.
Efficacy of non-selective NSAIDs versus placebo

The Cochrane review (51 trials, 16 higher-quality) estimated a pooled relative risk for global improvement in patients with acute low back pain of 1.24 (6 trials, 95% CI 1.10 to 1.41) after one week of NSAIDs relative to placebo and 1.29 (3 trials, 95% CI 1.05 to 1.57) for not requiring additional analgesics412, 413. Qualitatively, two of four higher-quality trials included in the Cochrane review reported better pain relief with NSAIDs compared to placebo and two found no differences. In a single trial of patients with chronic low back pain (rated higher-quality), an NSAID (ibuprofen) was superior to placebo449. The second, lower-quality systematic review was not as comprehensive (21 trials) as the Cochrane review and synthesized evidence qualitatively411. It also concluded that NSAIDs are effective for acute low back pain The third systematic review, which focused on a subset of three trials (two higher-quality) that evaluated patients with sciatica, found no difference between NSAIDs and placebo (OR 0.99, 95% CI 0.6 to 1.7)100.
Efficacy of non-selective NSAIDs versus other interventions

The Cochrane review found moderate evidence that NSAIDs are not more effective than opioid analgesics or muscle relaxants (6 trials, 1 higher-quality)412, 413. However, small sample sizes (n=19 to 44) could have limited the power of trials to detect differences. The Cochrane review also included two trials that found NSAIDs no more effective than physiotherapy or spinal manipulation and two trials that reached discordant conclusions about efficacy of NSAIDs relative to bed rest in patients with acute low back pain.
Efficacy of one non-selective NSAID versus another NSAID

The Cochrane review found no evidence from 24 trials that any one NSAID is superior to others for pain relief412, 413. Most comparisons were between different oral NSAIDs. One lower-quality trial found no difference between intramuscular and oral administration of tenoxicam450. Harms The Cochrane review found NSAIDs associated with similar risk of adverse events compared to placebo (RR=0.83, 95% CI 0.64 to 1.08)412, 413. The trials were generally short-term and not designed to evaluate risks of serious harms such as GI bleeds and CV events. In studies of NSAIDs taken for a variety of indications, NSAIDs are associated with an increased risk for serious GI complications (such as bleeding and perforation) that rises with age451, 452. The association between NSAIDs and cardiovascular events remains an active, ongoing area of investigation. A recent meta-analysis of over 130 randomized trials with cardiovascular safety data for NSAIDs found all non-selective NSAIDs other than naproxen associated with an
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain increased rate of myocardial infarction (about 1 additional myocardial infarction for every 300 patients treated for one year with an NSAID versus non-use)453. Due to concerns about potential cardiovascular risks, the FDA recently required labeling revisions to include additional warnings for all prescription and over-the-counter non-selective NSAIDs454. Costs We found no studies evaluating costs.
Summary of evidence
For acute low back pain, non-selective NSAIDs are associated with moderate short-term pain relief and global improvement compared to placebo (six trials) (level of evidence: good). For chronic low back pain, a single higher-quality trial found non-selective NSAIDs more effective than placebo (level of evidence: fair). Most trials evaluated mixed populations of patients with and without sciatica. Three trials (two higher-quality) that specifically evaluated patients with sciatica found no differences between non-selective NSAIDs and placebo (level of evidence: fair). Non-selective NSAIDs have not been shown to be more effective than other medications (opioids, skeletal muscle relaxants) or non-invasive interventions (spinal manipulation, physical therapy, bed rest) for low back pain (level of evidence: fair). There is no evidence that any non-selective NSAID is more effective than any other (level of evidence: good). Non-selective NSAIDs are associated with an increased risk of serious GI complications compared to non-use (level of evidence: good). The association between non-selective NSAIDs and cardiovascular events is an active area of research. In one recent meta-analysis of over 130 randomized controlled trials, non-selective NSAIDs other than naproxen were associated with a modest increase in risk of cardiovascular complications relative to non-use (about 1 additional myocardial infarction for every 300 patients treated for one year) (level of evidence: fair).

The AHCPR guidelines found NSAIDs acceptable for treating patients with acute low back problems (strength of evidence: B). The AHCPR guidelines found that NSAIDs have a number of potential side effects, with the most frequent gastrointestinal irritation. They recommend the decision to use these medications be guided by comorbidity, side effects, cost, and patient and provider preference (strength of evidence: C). The VA/DoD and UK RCGP guidelines for NSAIDs are similar to the AHCPR recommendations. Both the VA/DoD (strength of evidence: B) and UK RCGP (strength of evidence: ***) guidelines found various NSAIDs equally effective for low back pain.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The UK RCGP guidelines also found NSAIDs less effective for the reduction of nerve root pain (strength of evidence: **). The European COST guidelines recommend NSAIDs as second choice (after paracetamol) when needed for pain relief in patients with acute low back pain. They also recommend NSAIDs for pain relief in patients with chronic low back pain, but only for exacerbations or short-term periods (up to 3 months).
COX-2 selective NSAIDs

COX-2 selective NSAIDs could theoretically cause fewer GI complications than non-selective NSAIDs because they dont block the COX-1 enzyme, which helps protect the stomach lining. However, rofecoxib and valdecoxib were both voluntarily withdrawn from the market due to concerns about increased cardiovascular risk and other adverse events455, 456. Celecoxib is currently the only COX-2 selective NSAID available in the U.S. Results of search: systematic reviews No trials of COX-2 inhibitors were included in the Cochrane review of NSAIDs412, 413. We identified no other systematic reviews evaluating COX-2 inhibitors in patients with low back pain. Results of search: trials From 85 potentially relevant citations, we identified no trials on celecoxib for low back pain. We excluded eleven trials that evaluated selective NSAIDs not available in the U.S.457-466 and three trials467-469 that evaluated celecoxib in post-operative settings.
Efficacy of COX-2 inhibitors

In trials of patients with osteoarthritis and rheumatoid arthritis, there was no clear difference in efficacy between celecoxib and non-selective NSAIDs for pain relief, functional outcomes, or other measures of clinical efficacy470, 471. Harms In a meta-analysis of primarily short-term randomized trials, celecoxib was associated with a lower rate of discontinuations due to gastrointestinal side effects (RR=0.75, 95% CI 0.7-0.8) and clinical ulcers or bleeds (RR=0.61, 95% CI 0.46-0.81) compared to non-selective NSAIDs in patients with a variety of underlying conditions471. In the only long-term study designed to assess risk of ulcer complications (the CLASS trials), celecoxib was associated with fewer gastrointestinal complications after 6 months compared to diclofenac, but not compared to ibuprofen472. However, this benefit was no longer present after longer follow-up, in part due to high loss to follow-up473. No gastrointestinal safety advantage was observed with celecoxib in the subgroup of patients taking aspirin. The most comprehensive meta-analysis (over 130 randomized trials) found an increased risk of myocardial infarction with celecoxib compared to placebo when given for a variety of indications, though most events were observed in long-term trials using higher doses of celecoxib453. Other
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain than naproxen, which was neutral with respect to cardiovascular events, the risk of myocardial infarction with selective and non-selective NSAIDs in this meta-analysis was similar, with an estimated 1 additional myocardial infarction for every 300 patient-years of treatment compared to non-use of NSAIDs. Costs We found no studies evaluating costs.
Summary of evidence
Systematic reviews of COX-2-selective NSAIDs given for a variety of indications found no clear differences in efficacy (pain relief) compared to non-selective NSAIDs (level of evidence: good). Celecoxib is associated with a lower risk of discontinuations due to GI adverse events and serious GI complications compared to non-selective NSAIDs in trials of patients with a variety of underlying conditions, but most of the evidence comes from short-term trials (level of evidence: good).
In the largest meta-analysis of randomized trials, celecoxib was associated with an increased
risk of myocardial infarction compared to placebo (about 1 additional myocardial infarction for every 300 patients treated for one year). Most events were observed in trials of longer duration and that evaluated higher doses (level of evidence: good).

The other guidelines do not address COX-2-selective NSAIDs
Aspirin
Like the non-aspirin NSAIDs, aspirin (acetylsalicylic acid) has anti-inflammatory and analgesic effects. An important distinction between aspirin and non-aspirin NSAIDs is that aspirin also induces irreversible functional defects in platelets. Aspirin is therefore also used for primary and secondary prevention of thrombotic events, though usually in lower doses than considered most effective for pain relief. Results of search: systematic reviews We identified no systematic reviews evaluating efficacy of aspirin for low back pain. Results of search: trials From 74 potentially relevant citations, we identified one lower-quality trial that evaluated efficacy of aspirin versus multiple comparator drugs in patients with acute low back pain416. We excluded three trials that did not report results specifically for patients with low back pain422, 474 or were in a foreign language475.
Efficacy of aspirin versus other analgesics

The only trial that met inclusion criteria found aspirin at 3600 mg/day associated with a lower mean daily pain index score (1.425 vs. 1.713 on a 3 point scale, p<0.05) than the combination
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain of dextropropoxyphene plus acetaminophen, but found no differences between aspirin and indomethacin, mefenamic acid, acetaminophen alone, or phenylbutazone416. Aspirin also received the highest patient preference rating, though the difference was only significant compared to mefenamic acid and phenylbutazone (2.37 vs. 1.75 and 1.68, respectively, on a 3point scale). Harms Most trials that evaluated gastrointestinal bleeding risk and cardioprotective effects with aspirin were conducted in patients who received aspirin for cardiovascular prophylaxis, typically at lower doses (50 mg to 1500 mg/day) than considered most effective for analgesic and antiinflammatory effects. In a higher-quality meta-analysis of 24 such randomized trials with nearly 66,000 participants, the risk of any gastrointestinal bleeding was 2.47% with aspirin compared with 1.42% with placebo (OR=1.68, 95% CI 1.51 to 1.88), based on an average of 28 months therapy476. The risk of major gastrointestinal bleeding is probably substantially lower477. There was no association between gastrointestinal hemorrhage and dose, and modified release formulations did not attenuate risk for bleeding. Costs We found no studies evaluating costs.
Summary of evidence
There is insufficient evidence to judge efficacy of aspirin for low back pain (level of evidence: poor). Aspirin is associated with an increased risk of gastrointestinal bleeding even at low doses (level of evidence: good). Unlike non-aspirin NSAIDs, aspirin does not increase risk of cardiovascular events, and it is used for primary and secondary prevention of cardiovascular events (level of evidence: good).

The AHCPR guidelines do not consider aspirin separately from other NSAIDs.
Other Medications Antidepressants

Therapeutic effects of antidepressants on depression are thought due to their effects on different neurotransmitters. Certain antidepressants (particularly those that inhibit norepinephrine uptake) are also thought to have pain-modulating properties independent from effects on depression478. Results of search: systematic reviews We identified three higher-quality systematic reviews on efficacy of antidepressants for low back pain411, 479, 480. Two systematic reviews included seven480 and nine479 placebo-controlled trials. The third systematic review (seven trials)411 also included one head-to-head trial of
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain antidepressants481 and one trial that compared an antidepressant to acetaminophen419. We excluded four older systematic reviews193, 482-484 and one systematic review that evaluated antidepressants for a variety of pain conditions485 Results of search: trials Ten unique trials were included in the three systematic reviews of antidepressants411, 479, 480. In all of the trials, the duration of therapy ranged from four to eight weeks. We did not search for additional trials.
Efficacy of antidepressants versus placebo

No trial evaluated efficacy of antidepressants versus placebo for acute low back pain. For chronic low back pain, the overall conclusions of the three systematic reviews appeared consistent411, 479, 480. The first, qualitative systematic review found tricyclic or tetracyclic antidepressants slightly to moderately superior to placebo for at least one pain-related outcome measure in four of five trials (SMD=0.43486 and SMD=0.69487 in the two highest quality trials)480. Effects on functional outcomes were inconsistently reported and did not show clear benefits. The only tetracyclic antidepressant evaluated was maprotiline, a drug not available in the U.S.486. None of the trials evaluated norepinephrine-serotonin reuptake inhibitors such as duloxetine or venlafaxine. There were no beneficial effects associated with antidepressants without inhibitory effects on norepinephrine uptake (paroxetine and trazodone) compared to placebo in three trials. Maprotiline, the only tetracyclic antidepressant evaluated in the systematic reviews, is not available in the U.S. A second, quantitative systematic review found all antidepressants pooled together slightly effective for improving pain severity (SMD=0.41, 95% CI 0.22 to 0.61, 9 trials), though not for improving functional status (SMD=0.25, 95% CI -0.21 to 0.69, 5 trials)479. Although conclusions were interpreted as insensitive to antidepressant class (test for heterogeneity or stratified results not reported), effects on pain did not appear consistent across antidepressants. The point estimates indicate that paroxetine and trazodone (three trials) are associated with the least pain improvement (no statistically significant benefit in any of the trials). A third (qualitative) systematic review also concluded that tricyclic antidepressants are effective for chronic low back pain411.
Efficacy of one antidepressant versus another antidepressant

For chronic low back pain, two head-to-head trials provided somewhat conflicting evidence on the relative efficacy of different antidepressant classes. One higher-quality trial487 found maprotiline superior to paroxetine for pain relief (-45% vs. -27%, p=0.013), but one lower-quality trial481 found similar proportions of patients randomized to amitriptyline and fluoxetine reported at least moderate pain relief (82% vs. 77%).
EVIDENCE REVIEW
Efficacy of antidepressants versus other interventions

There is little evidence on efficacy of antidepressants versus other medications for low back pain. For acute low back pain, a single, small (n=39), lower-quality trial included in one of the systematic reviews411 found amitriptyline superior to acetaminophen for pain relief (p=0.045)419. Harms Though adverse events were generally not well reported, one systematic review found antidepressants associated with a higher risk for any adverse event compared to placebo (22% vs. 14%, p=0.01)479. Drowsiness (7%), dry mouth (9%), dizziness (7%) and constipation (4%) were the most commonly reported events. The trials were not designed to assess risk of serious adverse events such as overdose, increased suicidality, and arrhythmias. Costs We found no studies evaluating costs.
Summary of evidence
For acute low back pain, there is insufficient evidence (one lower-quality trial) to judge efficacy of antidepressants (level of evidence: poor). For chronic low back pain, tricyclic antidepressants are slightly to moderately more effective than placebo for pain relief in higher-quality trials, but do not significantly improve functional outcomes (level of evidence: good). For chronic low back pain, several trials found paroxetine and trazodone not effective or marginally effective compared to placebo (level of evidence: fair). There is insufficient evidence from head-to-head trials (one lower-quality trial) to judge relative effectiveness of tricyclic antidepressants and selective serotonin reuptake inhibitors (level of evidence: poor). There are no trials on effectiveness of other antidepressants venlafaxine or duloxetine for low back pain (level of evidence: poor). Although serious adverse events were not observed in the trials, the selected populations evaluated in clinical trials may decrease generalizability to general practice (level of evidence: fair).

The AHCPR guidelines recommend against use of antidepressant medications for acute low back problems (strength of evidence: C). The VA/DoD guidelines do not address antidepressant medications, and the UK RCGP guidelines found little evidence on their effectiveness for chronic low back pain, and none for acute low back pain (strength of evidence: *). The European COST guidelines recommend consideration of noradrenergic or noradrenergicserotoninergic antidepressants as co-medications for pain relief in patients with chronic low
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain back pain without renal disease, glaucoma, pregnancy, chronic obstructive pulmonary disease, or heart failure.
Benzodiazepines
Benzodiazepines are a class of medications that act on gaba-aminobutyric acidA (GABAA) receptors and have sedative, anxiolytic, and antiepileptic effects. They are commonly used as muscle relaxants, though they are not approved by the U.S. Food and Drug Administration for this indication. Results of search: systematic reviews We identified one higher-quality Cochrane review of skeletal muscle relaxants for low back pain that included trials of benzodiazepines488, 489. We excluded two relevant but outdated systematic reviews193, 415. Results of search: trials Eight trials of benzodiazepines were included in the Cochrane review488, 489. The trials ranged from 5 to 14 days in duration. We did not search for additional trials.
Benefits of benzodiazepines versus placebo

For acute low back pain, one higher-quality trial included in the Cochrane review488, 489 found no differences between diazepam and placebo490, but another, lower quality trial found diazepam superior to placebo for short-term pain relief and overall improvement491. For chronic low back pain, pooled results from two higher-quality trials found tetrazepam (not available in the U.S.) associated with better short-term pain relief (RR=0.71, 95% CI 0.54 to 0.93) and overall improvement (RR=0.63, 95% CI 0.42 to 0.97) compared to placebo after 10-14 days492, 493. A third, lower-quality placebo-controlled trial of diazepam for chronic low back pain found no benefit494.
Efficacy of benzodiazepines versus skeletal muscle relaxants

In two head-to-head trials included in the Cochrane review488, 489, there were no differences between diazepam and tizanidine for acute low back pain (one higher-quality trial495) or between diazepam and cyclobenzaprine for chronic low back pain (one lower-quality trial494). For acute low back pain, a third, higher-quality trial found diazepam inferior to carisoprodol for muscle spasm, global efficacy (excellent or very good 70% vs. 45%), and functional status496. One study which pooled data from 20 trials (n=1553) found no difference between diazepam and cyclobenzaprine for short-term (14 days) global improvement, but included trials of patients with either back or neck pain (mixed duration)497. Harms Adverse events such as somnolence, fatigue, and lightheadedness were reported more frequently with benzodiazepines compared to placebo488, 489. No trial evaluated risks with longterm use of benzodiazepines for low back pain such as addiction, abuse, overdose, or development of tolerance.
EVIDENCE REVIEW
Summary of evidence
For acute low back pain, evidence on efficacy of benzodiazepines versus placebo is mixed from two trials (1 higher-quality) (level of evidence: poor). For acute low back pain, evidence on efficacy of diazepam compared to skeletal muscle relaxants is mixed, with diazepam inferior to carisoprodol in one higher-quality trial, but no differences compared to tizanidine in another higher-quality trial (level of evidence: fair). For chronic low back pain, two higher-quality trials found benzodiazepines moderately effective for short-term outcomes, but a third found no benefit (level of evidence: fair). For chronic low back pain, there is insufficient evidence to judge efficacy of benzodiazepines relative to skeletal muscle relaxants (1 lower-quality trial) (level of evidence: poor). In patients with back or neck pain of mixed duration, there was no difference in short-term global improvement between diazepam and cyclobenzaprine in one analysis of 20 trials (n=1553) (level of evidence: fair). Benzodiazepines are associated with increased short-term central nervous system adverse events (level of evidence: good). Risks of addiction, abuse, development of tolerance, and overdose, particularly with long-term use, are unknown.

The UK RCGP guidelines note that use of benzodiazepines for more than two weeks carry a significant risk of habituation and dependency (strength of evidence: **) The European COST guidelines recommendations for muscle relaxants and benzodiazepines are the same.
Antiepileptic drugs
Gabapentin and pregabalin are antiepileptic drugs similar in structure to the neurotransmitter gamma-aminobutyric acid (GABA). They have been shown to be effective in patients with neuropathic pain498-500 and are approved by the FDA for treatment of diabetic neuropathy and postherpetic neuralgia. Other antiepileptic drugs have also been used to treat neuropathic pain, though they are not FDA-approved for this indication. The efficacy of antiepileptic drugs specifically for radicular (or non-radicular) low back pain has not been well studied. Results of search: systematic reviews We identified no systematic reviews evaluating efficacy of antiepileptic drugs for low back pain. Results of search: trials From 94 potentially relevant citations, we identified two trials of gabapentin for radiculopathy that met inclusion criteria501, 502. One was rated higher quality501. We also identified two higherquality randomized trials of topiramate for chronic radiculopathy503 or for chronic low back pain
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain with or without radiculopathy504. The trials ranged from six to ten weeks in duration. We identified no other trials of antiepileptic drugs for low back pain.
Efficacy of gabapentin versus placebo for radiculopathy

In one higher-quality trial, neither gabapentin nor placebo was associated with an improvement in resting back pain after six weeks (Table 26)501. However, gabapentin (but not placebo) was associated with small improvements compared to baseline on assessments of back pain with movement and for leg pain. It was not clear if between-group differences were significant. In the other, lower-quality trial, which used higher doses of gabapentin, patients with radiculopathy had greater improvement in pain at rest with gabapentin versus placebo after eight weeks502.
Table 26. Trials of gabapentin versus placebo in for chronic radicular low back pain
Author, year McCleane, 2001501
Yildirim, 2003502
n=50 8 weeks
Main results Gabapentin titrated to 1200 mg/day versus placebo Back pain at rest (mean change from baseline on 0-10 VAS): -0.51 (NS) vs. 0.1 (NS) Back pain with movement (mean change from baseline on 0-10 VAS): -0.47 (p<0.05) vs. +0.01 (NS) Leg pain (mean change from baseline on 0-10 VAS): -0.45 (p<0.05) vs. -0.24 (NS) Gabapentin titrated to 3600 mg/day versus placebo Back pain at rest (mean change from baseline on 0-3 scale): -1.04 vs. -0.32, p<0.01
Quality score 8/11
3/11
Efficacy of topiramate versus placebo for chronic low back pain with or without radiculopathy
One small (n=41), higher-quality crossover trial in patients with radiculopathy found topiramate more effective than diphenhydramine (used as an active placebo) for improving back and overall pain, though mean differences were small (less than one point on a 0 to 10 scale)503. There was no significant difference in leg pain, ODI scores, or SF-36 scores. Topiramate was also associated with a higher proportion of patients reporting moderate to complete pain relief (54% vs. 24%, p=0.005). A second higher-quality trial (n=96) of patients with chronic low back pain with or without leg pain found topiramate moderately more effective than placebo for improving Pain Rating Index scores (about 13 points on a 0 to 100 scale)504. Topiramate was also slightly more effective than placebo for improving scores on all SF-36 subscales. The largest difference was on the physical function subscale (9.1 point difference, range 0.6 to 8.3 for other subscales).
EVIDENCE REVIEW
Table 27. Trials of topiramate versus placebo for chronic low back pain with or without radiculopathy
Number of patients Duration of follow-up n=41 6 weeks, followed by crossover
Author, year Type of LBP Khoromi, 2005503 Radiculopathy
Muehlbacher, 2006504 Chronic low back pain with or without radiculopathy
n=96 10 weeks
Main results Topiramate titrated to 400 mg/day (average dose 208 mg/day) vs. diphenhydramine titrated to 50 mg/day (average dose 40 mg/day) Average leg pain (mean change from baseline on 0 to 10 scale): -0.98 vs. -0.24 (p=0.06) Average back pain: -1.36 vs. -0.49 (p=0.017) Average overall pain: -0.33 vs. +0.49 (p=0.02) Global pain relief moderate or better: 15/29 (54%) vs. 7/29 (24%) (p=0.005) Global pain relief 'lot' or 'complete': 9/29 (31%) vs. 1/29 (3.4%) ODI: -5 vs. -3 (NS) Beck Depression Inventory: No difference SF-36: No differences for any subscale after correction for multiple comparisons Topiramate titrated to 300 mg/day versus placebo Pain Rating Index (mean change from baseline on 0 to 100 scale): -12.9 vs. -1.5 (p<0.001) SF-36 Physical functioning subscale (mean change from baseline on 0 to 100 scale): +8.7 vs. -0.4 (p<0.01, favors topiramate) SF-36, Bodily pain subscale (0 to 100): +4.1 vs. +0.9 (p<0.01, favors topiramate) SF-36, other subscales: Differences in change compared to baseline ranged from 0.6 (Role-emotional) to 8.3 (Rolephysical) points, favoring topiramate for all comparisons at p<0.05
Quality score 7/11
7/11
Harms Withdrawal due to adverse events occurred in 2 of 25 patients randomized to gabapentin versus none of 25 randomized to placebo in one trial502. No withdrawals due to adverse events occurred in the other trial501. However, drowsiness (6%), loss of energy (6%), and dizziness (6%) were reported with gabapentin501. A higher proportion of patients randomized to topiramate compared to diphenhydramine withdrew due to adverse events in one trial (33% vs. 15%)503, but there was no difference in rates of withdrawal due to adverse events in the other (4% vs. 4%)504. Topiramate was also associated with higher rates of withdrawal due to adverse events (33% vs. 15%), sedation (34% vs. 3%) and diarrhea (30% vs. 10%) compared to diphenhydramine in one trial503. Costs We found no studies evaluating costs.
Summary of evidence
In patients with radiculopathy, two small (n=50 and n=80) trials (one higher-quality) found gabapentin slightly superior for short-term pain relief compared to placebo (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain No trials evaluated efficacy of gabapentin in patients with non-radicular low back pain. In patients with radiculopathy, one small (n=42), higher-quality trial found topiramate slightly superior to diphenhydramine (used as an active placebo) for short-term pain relief, but not functional status. Topiramate was associated with more withdrawals due to adverse events, sedation and diarrhea than diphenhydramine (level of evidence: poor). For chronic low back pain with or without radiculopathy, one small (n=96), higher-quality trial found topiramate moderately superior to placebo for short-term pain relief and slightly superior for functional status (level of evidence: poor).

The European COST guidelines found insufficient evidence to recommend gabapentin in patients with chronic nonspecific low back pain, but does not address use of gabapentin for radiculopathy.
Skeletal muscle relaxants

The term skeletal muscle relaxants is commonly used to refer to a heterogeneous group of pharmacologically unrelated medications that are FDA-approved to treat two distinct underlying conditions: spasticity from upper motor neuron syndromes and pain or spasms from musculoskeletal conditions such as non-specific low back pain505. The muscle relaxants carisoprodol, chlorzoxazone, cyclobenzaprine, metaxalone, methocarbamol, and orphenadrine carry FDA-approved indications for treatment of musculoskeletal conditions. Although the other drugs in this class (baclofen, dantrolene, and tizanidine) are approved only for the treatment of spasticity, there is some overlap in clinical usage. In particular, tizanidine has also been studied in patients with musculoskeletal conditions such as low back pain. Benzodiazepines are commonly used as muscle relaxants, though they are not FDA-approved for this indication (see section on benzodiazepines). Results of search: systematic reviews We identified four higher-quality systematic reviews on efficacy and safety of muscle relaxants for low back pain100, 411, 488, 489, 506. Of these, a recent higher-quality Cochrane review was the most comprehensive (26 trials of skeletal muscle relaxants)488, 489. We excluded two outdated systematic reviews193, 415. Results of search: trials Thirty-six unique trials of skeletal muscle relaxants were included in the four systematic reviews100, 411, 488, 489, 506. The duration of therapy in all trials was two weeks or less, with the exception of one three-week trial. We did not search for additional trials.
Efficacy of skeletal muscle relaxants versus placebo

For acute low back pain, the Cochrane review included eight trials that found skeletal muscle relaxants superior to placebo for short-term (2 to 4 days) pain relief (at least a two-point or 30% improvement on an 11 point pain rating scale) and global efficacy488, 489. The relative risk for pain relief was 1.25 (95% CI 1.12 to 1.41) after 2 to 4 days and 1.72 (95% CI 1.32 to 2.22) after
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain 5 to 7 days, based on three higher-quality trials and one lower-quality trial that could be pooled. Skeletal muscle relaxants were also superior to placebo for short-term improvement in global efficacy (RR=2.04, 95% CI 1.05 to 4.00 after 2 to 4 days), though differences were no longer significant after 5 to 7 days (RR=1.47, 95% CI 0.88 to 2.44). The Cochrane review also included three trials of skeletal muscle relaxants for chronic low back pain. Only onea lower-quality trial of cyclobenzaprine that did not report pain intensity or global efficacy outcomesevaluated a skeletal muscle relaxant available in the U.S.494. Two other systematic reviews were less comprehensive than the Cochrane review, but reached consistent conclusions411, 506. One systematic review of cyclobenzaprine included trials of patients with back or neck pain. 506. It found cyclobenzaprine slightly to moderately superior to placebo (SMD=0.38 to 0.58) for pain, muscle spasm, tenderness to palpation, range of motion, and activities of daily living, with the greatest benefit seen within the first few days of treatment. It included two lower-quality trials of cyclobenzaprine for chronic or subacute low back or neck pain that reported mixed results versus placebo and were excluded from the Cochrane review488, 489. A systematic review100 on various treatments for sciatica included one higherquality trial507 that found no difference between tizanidine and placebo.
Efficacy of one skeletal muscle relaxant versus another skeletal muscle relaxant
The Cochrane review found insufficient evidence to conclude that any muscle relaxant is more beneficial or less harmful compared to any other488, 489. A systematic review of muscle relaxants for various musculoskeletal conditions reached similar conclusions505. Cyclobenzaprine is the most-studied skeletal muscle relaxant in published trials506. There is sparse evidence (two trials) on effectiveness of the antispasticity drugs dantrolene and baclofen for either chronic or acute low back pain488, 489. Tizanidine (the other antispasticity skeletal muscle relaxant) was effective for low back pain in eight trials. Harms The Cochrane review found skeletal muscle relaxants associated with more total adverse events (RR=1.50, 95% CI 1.14 to 1.98) and central nervous system (primarily sedation) adverse events (RR=2.04, 95% CI 1.23 to 3.37) than placebo, though most events were self-limited, and serious complications appeared rare488, 489. Certain skeletal muscle relaxants are associated with other specific safety issues. For example, carisoprodol is a controlled substance in some states because of its metabolism in part to meprobamate, a drug associated with abuse and overdose. Dantrolene carries a black box warning on its label about potentially fatal hepatotoxicity. Chlorzoxazone and tizanidine are associated with usually self-limited and mild hepatotoxicity505. Costs We found no studies evaluating costs.
EVIDENCE REVIEW
Summary of evidence
For acute low back pain, multiple trials found skeletal muscle relaxants moderately more effective than placebo for short-term (less than one week) pain relief and global response (level of evidence: good). For chronic low back pain or sciatica, there is insufficient evidence (one lower-quality trial) to judge efficacy of skeletal muscle relaxants (level of evidence: poor). Although there is no evidence showing one skeletal muscle relaxant is superior to others (level of evidence: fair), the number of available trials varies considerably for different drugs, with cyclobenzaprine the most-studied drug in published trials. Only two trials evaluated the efficacy of the antispasticity drugs baclofen and dantrolene (level of evidence: poor). Skeletal muscle relaxants are associated with an increased rate of adverse events (mostly sedation) compared to placebo, though they are usually mild and self-limited (level of evidence: fair). Specific safety issues are associated with carisoprodol (metabolism to meprobamate), dantrolene (potentially fatal hepatotoxicity), chlorzoxazone and tizanidine (usually reversible and mild hepatotoxicity).

The AHCPR guidelines recommend muscle relaxants as an option in the treatment of low back pain problems. While they found muscle relaxants probably more effective than placebo, they also found muscle relaxants had not been shown to be more effective than NSAIDs (strength of evidence: C). The AHCPR guidelines recommend balancing potential side effects (particularly drowsiness) associated with muscle relaxants against a patients intolerance for other agents when considering the optional use of muscle relaxants (strength of evidence: C). The VA/DoD guidelines are identical to the AHCPR guidelines. The UK RCGP guidelines are similar to the AHCPR recommendations, but rated evidence on the effectiveness of muscle relaxants for acute back pain more highly (strength of evidence: ***). The European COST guidelines recommend the addition of a short course of muscle relaxants on its own or added to NSAIDs in patients with acute low back pain, if acetaminophen or NSAIDs failed to reduce pain. The European COST guidelines recommend consideration of muscle relaxants for short-term pain relief in chronic low back pain, but suggests caution because of side effects and to use medications with fewer side effects first.
Opioid analgesics
Opioid analgesics are derivatives of morphine that bind to opioid receptors. Some are available in immediate-release and sustained-release formulations, and opioids can be administered via a variety of routes (most commonly oral or transdermal). Opioids are the most potent medications
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain available for treatment of most types of severe pain. However, they are also associated with significant adverse events, including nausea, somnolence, respiratory depression (including risk of overdose), abuse, and addiction. Results of search: systematic reviews We identified no systematic reviews of opioids for low back pain. We excluded four reviews that did not clearly use systematic methods508, 509 or were not specific for low back pain510, 511. Results of search: trials From 600 potentially relevant citations, we identified nine trials (one higher-quality512) of opioids that met inclusion criteria353, 426, 512-518. Two trials were placebo-controlled512, 514, two compared opioids to either NSAIDs or acetaminophen353, 517, and the remainder compared different opioid drugs or formulations (sustained-release versus immediate-release). All of the trials were less than 3 weeks in duration except for two (one 16 weeks517, the other 13 months513). We excluded twelve trials416, 423, 425, 428-431, 433, 519-521 that evaluated dual therapy with an opioid plus another medication versus a different medication (or medication combination), one trial522 that evaluated single-dose therapy, two trials523, 524 that did not report efficacy of opioids specifically for low back pain, and two trials525, 526 that did not evaluate any included outcome.
Efficacy of opioids versus placebo

For chronic low back pain, a single higher-quality trial found either sustained-release oxymorphone or sustained-release oxycodone superior to placebo for pain relief after 18 days (average difference in pain relief 18 points on a 100 point scale, Table 28)512. The active treatments were also superior to placebo for measures of interference with pain on activities. A problem with interpreting these results is that all patients were titrated to stable pain control on opioids prior to allocation to continued opioids or placebo, so poorer outcomes in the placebo group could have been due in part to cessation of opioids and withdrawal. In addition, although patients were initially randomized to sustained-release oxycodone or sustained-release oxymorphone, it was not clear if patients were randomly re-allocated to continued opioids or placebo. A second, lower-quality placebo-controlled trial found the less potent opioid propoxyphene no better than placebo for improvement in pain scores or assessments of global improvement in patients with acute or chronic low back pain514. Propoxyphene was superior to placebo on only one of three sleep parameters (difficulty falling asleep).
EVIDENCE REVIEW
Table 28. Trials of an opioid versus placebo in patients with low back pain
Number of patients Duration of follow-up n=61 14 days Hale, 2005512 n=235 18 days
Author, year Baratta, 1976514
Main results Propoxyphene versus placebo Pain on active improvement (mean improvement from baseline): 0.8 vs. 0.4, NS Global improvement at least satisfactory: 22% vs. 14% (NS) Sustained-release morphine versus sustained-release oxycodone versus placebo Pain intensity (100 point VAS), mean differences versus placebo: -18.21 vs. -18.55 (p=0.0001 for each comparison) Global assessment at least good: 59% vs. 63% vs. 27%
Quality score 5/11
7/11
Two systematic reviews of fifteen511 and thirty510 placebo-controlled trials of opioids for various non-cancer pain conditions (most commonly osteoarthritis and neuropathic pain) found opioids moderately effective. They estimated a mean decrease in pain intensity with opioids in most trials of at least 30%511 or an SMD for pain relief of -0.60 (95% CI -0.69 to -0.50)510. In one of the reviews, opioids were also slightly superior to placebo for functional outcomes (SMD=-0.31, 95% CI -0.41 to -0.22)510. Estimates of benefit were similar for neuropathic and non-neuropathic pain.
Efficacy of opioids versus NSAIDs or acetaminophen

For low back pain, opioids have only been directly compared to NSAIDs in two lower-quality trials (Table 29). One small, lower-quality trial of patients with chronic low back pain found adding an opioid to naproxen alone associated with superior outcomes for average pain, current pain, and anxiety or depression scores after 16 weeks517. Differences in pain relief were small, ranging between 5 and 10 points on a 100-point scale. In addition, results are difficult to interpret because doses of naproxen werent clearly reported. Another trial (n=50) found similar mean number of days before return to work in patients with acute low back pain randomized to codeine or acetaminophen (10.7 vs. 13.0 days)353.
EVIDENCE REVIEW
Table 29. Trials of an opioid versus an NSAID or acetaminophen

Author, year Jamison, 1998517
Wiesel, 1980353
n=50 14 days
Main results Sustained-release morphine + immediate-release oxycodone (titrated dose) + naproxen versus immediaterelease oxycodone (set dose) + naproxen versus naproxen alone (mean scores over 16 weeks, all outcomes on 0 to 100 scales) Average pain: 54.9 vs. 59.8 vs. 65.5 Anxiety: 11.2 vs. 15.0 vs. 31.6 Depression: 10.8 vs. 16.4 vs. 26.9 Level of activity: 49.3 vs. 49.3 vs. 51.5 Hours of sleep (means): 5.9 vs. 5.9 vs. 6.1 Codeine versus acetaminophen Mean number of days before return to work: 10.7 vs. 13.0 (NS)
Quality score 3/11
3/11
One systematic review that included trials of opioids for a variety of chronic pain conditions (8 trials, only one of low back pain patients) found no difference between all opioids and other drugs (NSAIDs, tricyclic antidepressants, or acetaminophen) for pain relief (SMD=-0.05, 95% CVI -0.32 to 0.21), though more potent opioids (oxycodone and morphine) were slightly superior to other drugs (SMD=-0.34, 95% CI -0.67 to -0.01) in stratified analyses510. There were no differences in functional outcomes.
Efficacy of different opioids and opioid formulations

There was no evidence from five lower-quality trials that sustained-release opioid formulations are superior to immediate-release formulations for pain, functional status, or other measured outcomes in patients with low back pain (Table 30)426, 515-518.
EVIDENCE REVIEW
Table 30. Trials of a sustained-release opioid versus an immediate-release opioid

Number of patients Duration of follow-up n=61 2 weeks followed by crossover n=104 5 days Hale, 1999516 n=57 4-7 days followed by crossover n=36 16 weeks
Author, year Gostick, 1989515
Main results Sustained- versus immediate-release dihydrocodeine No differences for pain intensity, rescue drug use, global efficacy, patient preference Sustained-release codeine plus acetaminophen versus immediate-release codeine plus acetaminophen Long-acting codeine superior for pain intensity, but nonequivalent codeine use (200 mg vs. 71 mg) Sustained- versus immediate-release oxycodone No differences for overall pain intensity, mean pain intensity, or rescue drug use Sustained-release morphine + immediate-release oxycodone (titrated dose) + naproxen versus immediate-release oxycodone (set dose) + naproxen versus naproxen alone (mean scores over 16 weeks, all outcomes on 0 to 100 scales) Average pain: 54.9 vs. 59.8 vs. 65.5 Anxiety: 11.2 vs. 15.0 vs. 31.6 Depression: 10.8 vs. 16.4 vs. 26.9 Level of activity: 49.3 vs. 49.3 vs. 51.5 Hours of sleep (means): 5.9 vs. 5.9 vs. 6.1 Sustained- versus immediate-release oxycodone No differences for pain intensity, time to stable pain control, mean number of dose adjustments
Quality score 5/11
Hale, 1997426
5/11
5/11
Jamison, 1998517
3/11
Salzman, 1999518
n=57 10 days
2/11
In two head-to-head trials of opioids for chronic low back pain (Table 31), there were no differences in efficacy between sustained-release oxymorphone and sustained-release oxycodone512 or between transdermal fentanyl and sustained-release morphine513. The latter study is the longest (13 months) and largest (n=683) trial of opioids for low back pain available.
Table 31. Head-to-head trials of sustained-release opioids
Number of patients Duration of follow-up n=683 13 months Hale, 2005512 n=235 18 days
Author, year Allan, 2005513
Main results Transdermal fentanyl versus sustained-release oral morphine No differences for pain scores, rescue medication use, quality of life, loss of working days Sustained-release morphine versus sustained-release oxycodone No differences for pain intensity, pain relief, pain interference with activities, global assessment
Quality score 4/11
7/11
A systematic review of opioids for various non-cancer pain conditions also found no clear differences between sustained- and immediate-release opioids or different sustained-release opioids527.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Harms In the single higher-quality trial, a large proportion of patients on opioids had adverse events (85%), with constipation and sedation the most commonly reported symptoms512. Few serious adverse events were reported, and withdrawal due to adverse events was low in all groups, probably due at least in part to the use of a run-in period prior to randomization. In trials that compared opioids to other analgesics (NSAIDs or acetaminophen), constipation, dry mouth, somnolence, and nausea were all more common in the opioid arms353, 517. One lower-quality trial reported a higher rate of constipation with oral sustained-release morphine compared to transdermal fentanyl (65% vs. 52%)513. However, sustained-release morphine was also associated with a non-significant trend towards a lower rate of withdrawal due to any adverse event (31% vs. 37%). In systematic reviews of opioids for various non-cancer pain conditions, 50% to 80% of patients experienced at least one adverse event. Constipation (41%), nausea (32%), and somnolence (29%) were the most common adverse events510, 511, 528. Relative to placebo, the rate of constipation was 10% to 16% higher with opioids, nausea 15% higher, dizziness or vertigo 8% to 9% higher, somnolence or drowsiness 9% to 10% higher, vomiting 5% to 8% higher, and dry skin, itching or pruritus 4% to 11% higher510, 528. About 22% to 24% of patients randomized to opioids withdrew due to adverse events, a rate about two-to-threefold higher than in patients randomized to placebo510, 528. Abuse and addiction were rarely reported in the trials, but because of short follow-up, enrollment of selected populations, and use of insensitive or poorly defined methods for detecting abuse and addiction, reliable conclusions about risks for these outcomes were not possible even when such data (few or no cases) were reported510, 511. In trials with longer-term (longer than seven months) open-label follow-up, less than half of patients remained on opioids511. Costs We found no studies evaluating costs.
Summary of evidence
For either acute or chronic low back pain, evidence that demonstrates efficacy of opioids versus placebo is sparse (one higher-quality trial showing moderate effects on pain) (level of evidence: fair). Multiple trials of patients with various non-cancer pain conditions consistently found opioids moderately superior to placebo for pain relief in primarily short-term trials (level of evidence: good), though effects on functional outcomes appear small and evidence on long-term effects is sparse. There is insufficient evidence from single, lower-quality trials to judge efficacy of opioids versus acetaminophen or in addition to NSAIDs (level of evidence: poor). For chronic low back pain, consistent evidence from lower-quality trials found no differences between sustained- and immediate-release opioids on a variety of outcomes (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain There were no clear differences in efficacy or safety between different sustained-release opioids in two head-to-head trials (one higher-quality) (level of evidence: fair). Although adverse events are common with opioids, few serious adverse events were reported in published trials (level of evidence: fair). However, reliable estimates of long-term harms, rates of abuse or addiction, overdose, or other serious adverse events are not available (level of evidence: poor).

The AHCPR guidelines recommend opioids as an option for a time-limited course in patients with acute low back problems, with the decision guided by consideration of potential complications (which can lead to discontinuation in as many of 35% of patients) relative to other options (strength of evidence: C). The AHCPR guidelines found opioids no more effective in relieving low back symptoms than safer analgesics such as acetaminophen, aspirin, or other NSAIDs (strength of evidence: C). The AHCPR guidelines recommend warning patients about potential physical dependence and the danger associated with the use of opioids while operating heavy equipment or driving (strength of evidence: C). The VA/DoD and UK RCGP guideline recommendations are essentially identical to the AHCPR recommendations. The UK RCGP guidelines also suggest that pain of such severity that it requires opioids for longer than two weeks requires further investigation and assistance with management (strength of evidence: *). The UK RCGP guidelines suggest combinations of paracetamol plus a weak opioid as an alternative when paracetamol or NSAIDs alone do not give adequate pain control, though adverse effects include constipation and drowsiness (strength of evidence: **). The European COST guidelines recommend weak opioids in patients with nonspecific chronic low back pain who do not respond to other treatment modalities. Due to the risk of addiction, they recommend slow-release over immediate-release formulations and scheduled rather than as-needed dosing.
Tramadol
Tramadol is a synthetic centrally-active analgesic that has weak affinity for opioid -receptors. It also appears to have effects on the noradrenergic and serotoninergic systems. Results of search: systematic reviews We identified one higher-quality systematic review of various medications for low back pain411 that included three short-term trials of tramadol519, 529, 530 for low back pain. Two of the trials were rated higher-quality529, 530.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials Three trials519, 529, 530 of tramadol were included in the systematic review411. From 147 potentially relevant citations, we identified two additional trials of tramadol for low back pain that met inclusion criteria. Both compared sustained-release to immediate-release tramadol531, 532. All trials ranged between one and four weeks in duration. We excluded three trials that evaluated dual therapy with tramadol plus another medication versus another medication or medication combination432, 520, 533, one trial because it is only available as a conference abstract534, and one small (n=40) trial cited in an electronic database that we could not locate535
Efficacy of tramadol versus placebo

The systematic review411 included one higher-quality trial530 that found tramadol moderately more effective than placebo for chronic low back pain on mean pain scores at 4 weeks (3.5 vs. 5.1 on 10 point scale, p0.001) as well as the McGill Pain Questionnaire (p=0.0007) and the RDQ (p=0.0001).
Efficacy of tramadol versus other interventions

No trial compared tramadol to opioid analgesics for low back pain. The systematic review included two trials comparing tramadol to other drugs411. For acute low back pain, one higherquality trial found tramadol inferior to the NSAID dextroprofen-trometamol for pain relief (p=0.044) and need for rescue medication (p=0.011)529. For chronic low back pain, one lowerquality trial found tramadol associated with similar outcomes compared to the combination of paracetamol plus codeine519.
Efficacy of sustained-release versus immediate-release tramadol

Two short-term (three weeks), lower-quality trials found no differences in efficacy between sustained-release and immediate-release tramadol for chronic low back pain (Table 32)531, 532.
EVIDENCE REVIEW
Table 32. Trials of sustained-release tramadol vs. immediate-release tramadol

Number of patients Duration of follow-up n=248 9 days
Author, year Raber, 1999531
Sorge, 1997
532
n=205 3 weeks
Main results Tramadol sustained-release versus tramadol immediaterelease Pain relief, improvement in VAS (0 to 100): -25 vs. -25 for perprotocol analysis; ITT results stated as similar but data not reported Functional assessment 'without pain' or 'slight pain possible': >80% in both intervention groups for putting on jacket, putting on shoes, and climbing/descending stairs No awakenings due to low back pain: 41% vs. 47% Global assessment 'good' or 'moderately good': 80% (84/105) vs. 81% (80/99) Global assessment 'good': 47% (49/105) vs. 46% (45/99) Tramadol sustained-release versus tramadol immediaterelease Pain relief 'complete', 'good', or 'satisfactory': 88% (52/59) vs. 86% (49/57; results only reported for persons who completed three-week course Pain relief 'complete': 8.5% (5/59) vs. 5.3% (3/57); results only reported for persons who completed three-week course
Quality score 4/11
5/11
Harms In two trials included in the systematic review411, tramadol was associated with similar rates of withdrawal due to adverse events compared to placebo530 or the combination of paracetamol plus codeine529. There were also no differences in adverse events between sustained-release and immediate-release tramadol531, 532. Costs We found no studies evaluating costs.
Summary of evidence
No trial evaluated efficacy of tramadol versus placebo for acute low back pain. For acute low back pain, tramadol was inferior to the NSAID dextroprofen-trometamol (not available in the U.S.) for pain relief and need for rescue medications in one higher-quality trial (level of evidence: fair). For chronic low back pain, tramadol was moderately more effective than placebo for shortterm pain relief and improvement in functional status in one higher-quality trial (level of evidence: fair). For chronic low back pain, tramadol was no better than the combination of paracetamol plus codeine in one lower-quality trial (level of evidence: poor). There is insufficient evidence to judge efficacy of tramadol compared to acetaminophen, opioid analgesics, or NSAIDs available in the U.S. (no trials). There was no difference in benefits or harms between sustained- and immediate-release tramadol in two lower-quality trials (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain In single trials, tramadol was associated with similar rates of withdrawal due to adverse events (a marker for intolerable or severe adverse events) compared to placebo or the combination of paracetamol plus codeine (level of evidence: fair).

The European COST guidelines recommend weak opioids (including tramadol) in patients with nonspecific chronic low back pain who do not respond to other treatment modalities. Due to the risk of addiction, they recommend slow-release over immediate-release formulations and scheduled rather than as needed dosing.
Systemic corticosteroids
Results of search: systematic reviews We identified no systematic reviews on efficacy of systemic corticosteroids for low back pain. Results of search: trials From 418 potentially relevant citations, we identified three small (n=33 to 65), higher-quality trials of systemic corticosteroids for radiculopathy of acute or unspecified duration536-538. One other higher-quality trial evaluated efficacy of systemic corticosteroids for acute low back pain without radiculopathy539. We excluded three trials540-542 of systemic corticosteroids in operative or post-operative settings and one German-language trial543.
Efficacy of systemic corticosteroids versus placebo

For acute sciatica, one higher-quality trial found a single large (500 mg) bolus of intravenous methylprednisolone associated with small (average 6 mm on a 100 mm scale) early improvement in short-term leg pain compared to placebo, but the benefit was no longer observed after the first 3 days (Table 33)536. There were no differences in degree of pain relief, functional disability, the proportion requiring spine surgery within the first month, or medication use. In two other higher-quality trials, seven day tapering courses of either oral537 or intramuscular538 dexamethasone (initial dose 64 mg/day) were not associated with differences in any outcomes including overall effect (either early or after up to 4 years of follow-up), hospitalization length, or subsequent surgery. For acute low back pain with a negative straight leg raise test, one higher-quality trial found no differences between a single intramuscular injection of 160 mg of methylprednisolone versus a placebo injection for pain relief or function539. The excluded German-language trial also reported no significant difference between a 10-day course of intramuscular steroids and placebo in patients with sciatica in likelihood of a successful outcome (OR=2.0, 95% CI 0.8 to 4.9)543.
EVIDENCE REVIEW
Table 33. Trials of systemic corticosteroids versus placebo

Number of patients Duration of follow-up n=65 (acute sciatica) 30 days
Author, year Finckh, 2006536
Friedman, 2006539
n=88 (acute low back pain with negative straight leg raise) 1 month n=33 (sciatica, duration of symptoms unclear) 1 to 4 years n=52 (sciatica, duration of symptoms unclear) 9 days or longer
Main results Methylprednisolone 500 mg IV bolus versus placebo Leg pain, difference between interventions in VAS pain scores (0 to 100 scale): 5.7 (favors methylprednisolone) at day 3, (p=0.04), not significant after 3 days (p=0.22) Proportion with >20 mm improvement in VAS pain score after 1 day: 48% vs. 28% (p=0.097) Methylprednisolone 160 mg IM bolus vs. placebo Pain, mean change from baseline (0 to 10 scale): -4.1 vs. -4.8 (NS) after 1 week, -5.1 vs. -5.8 (NS) after 1 month RDQ-18, mean score (0 to 18): 2.6 vs. 3.4 after 1 week, 2.6 vs. 3.1 after 1 month Dexamethasone 64 mg followed by 1 week oral taper versus placebo Early improvement: 33% (7/21) vs. 33% (4/12) Sustained improvement (1 to 4 years): 50% (8/16) vs. 64% (7/11) Dexamethasone 64 mg followed by 1 week intramuscular taper versus placebo Positive effect: 52% (13/25) vs. 58% (14/24) Subsequent surgery: 32% (8/25) vs. 25% (6/24)
Quality score 10/11
11/11
Haimovic, 1986537
6/11
Porsman, 1979538
6/11
Harms In one trial, a large (500 mg) intravenous methylprednisolone bolus was associated with two cases of transient hyperglycemia and one case of facial flushing536. In another trial, a smaller (160 mg) intramuscular methylprednisolone injection was associated with no cases of hyperglycemia requiring medical attention, infection, or gastrointestinal bleeding539. Although there was a higher rate of adverse events (primarily gastrointestinal) in the placebo group in this trial, these findings are difficult to interpret because both groups also were given naproxen and oxycodone and use of those medications was not reported. Adverse events were poorly reported in the other trials. Costs We found no studies evaluating costs.
Summary of evidence
For acute sciatica, systemic corticosteroids were consistently associated with no clinically significant benefit when given as a single large parenteral bolus or as a short oral or intramuscular taper (three higher-quality trials) (level of evidence: good). For acute non-radicular low back pain, one higher-quality trial found no benefit from a single intramuscular injection of methylprednisolone (160 mg) (level of evidence: fair). Serious adverse events after single large boluses of corticosteroids were not reported in two trials (level of evidence: fair). However, systemic corticosteroids are associated with
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain hyperglycemia, systemic infections, bleeding, osteoporosis, avascular necrosis, and psychosis, particularly with higher doses and longer courses of treatment.

The AHCPR guidelines recommend against systemic steroids for acute low back problems (strength of evidence: C). The AHCPR guidelines found a potential for severe side effects with extended use of oral steroids or short-term use of high-dose steroids (strength of evidence: D). The UK RCGP guidelines on systemic steroids are similar.
Topical lidocaine
Results of search: systematic review We found no systematic reviews of topical lidocaine for low back pain. Results of search: trials From 278 potentially relevant citations, we identified one open-label, randomized trial, but it did not meet inclusion criteria because results are only available as a conference abstract544. It found no differences between lidocaine 5% patch and celecoxib 200 mg for low back pain (with or without radiation) after four weeks on the Brief Pain Index, ODI, or proportion of patients with >30% reduction in pain. The trial was terminated early because of concerns about potential cardiovascular risks associated with celecoxib, and results were only reported for 76 of the 97 patients randomized. Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
There is insufficient evidence to evaluate efficacy of topical lidocaine for low back pain (one open-label trial, terminated early, only available as an abstract) (level of evidence: poor).

The other guidelines do not address topical lidocaine.
Herbal therapies
Results of search: systematic reviews We identified one higher-quality Cochrane review on efficacy of devils claw, white willow bark, or topical cayenne for low back pain herbal therapies for low back pain545, 546. We excluded an earlier version of this systematic review547.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials Ten trials were included in the systematic review545. Five of the ten trials were rated higherquality (met more than half of the 12 quality criteria), but all assessed short-term (<6 weeks) outcomes and more than half either had authors with potential conflicts of interest or did not report potential conflicts. In addition, the same investigator led half of the trials. We did not search for additional trials.
Efficacy of harpagoside (devils claw) versus placebo

For acute episodes of chronic non-specific low back pain, the Cochrane review545, 546 included two higher-quality trials548, 549 that found devils claw (harpagoside) superior to placebo for the proportion pain-free (9% and 17% with devils claw versus 2% and 5% with placebo). However, significant differences were not seen for Arhus Index scores (a measure of physical impairment, disability, and pain) or concomitant analgesic (tramadol) use.
Efficacy of salix alba (white willow bark) versus placebo

For chronic low back pain, the Cochrane review545, 546 included one higher-quality trial that found white willow bark superior to placebo for the likelihood of becoming pain-free, with a significant dose trend (5.7% with placebo, 21% with low dose willow bark, 39% with high dose), as well as for improvements in Arhus Index scores550.
Efficacy of capsicum frutescens (cayenne) versus placebo

Capsaicin is the main active ingredient in cayenne. For acute low back pain, one lower-quality trial included in the Cochrane review545, 546 found topical cayenne (in combination with topical salicylate) superior to placebo cream (mean improvement 3.79 cm on a 10 cm VAS after 14 days in the cayenne group)551. For chronic low back pain, two lower-quality trials found cayenne associated with a higher likelihood of at least 50% improvement in pain compared to placebo (35% versus 17% in one trial552 and 45% versus 24% in the other553). Arhus Index scores also decreased more in the cayenne groups (33% vs. 22% in one trial552 and 42% vs. 31% in the other553). However, a fourth, lower-quality trial of cayenne versus homeopathic treatment (Spiroflor SLR homeopathic gel) for back pain of mixed duration found no differences in pain relief, proportion using acetaminophen, proportion unable to work, or assessments of overall efficacy554.
Efficacy of herbal therapies versus other interventions

Two trials included in the Cochrane review545, 546 compared either devils claw (higher-quality459) or willow bark (lower-quality458) to low-dose (12.5 mg) rofecoxib, a COX-2 selective NSAID no longer on the market. Both found no statistically or clinically significant differences between herbal therapies and rofecoxib for pain, Arhus Index scores, or other outcomes. Harms Devils claw was not consistently associated with a higher rate of adverse events compared to placebo in the Cochrane review545, 546. Serious adverse events were rare in published trials, though a severe allergic reaction was reported in one study of willow bark550. Because of its
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain mechanism of action, cayenne is associated with burning or itching upon initial administration that decreases after repeated applications. Costs We found no studies evaluating costs.
Summary of evidence
For acute exacerbations of chronic low back pain, two higher-quality trials found devils claw slightly superior to placebo for short-term pain relief and one higher-quality trial found devils claw equivalent to low-dose rofecoxib. Because all of the trials were led by the same investigator, reproducibility of findings has not been established (level of evidence: fair). For acute exacerbations of chronic low back pain, one higher-quality trial found white willow bark superior to placebo and one lower-quality trial found white willow bark equivalent to lowdose rofecoxib (level of evidence: fair). For acute low back pain or acute exacerbations of chronic low back pain, one lower-quality trial found cayenne moderately superior to placebo for pain relief and other outcomes, but one other lower-quality trial found no benefit compared to a homeopathic gel (level of evidence: poor). For chronic low back pain, one higher-quality trial found willow back moderately superior to placebo for short-term pain relief (level of evidence: fair). For chronic low back pain, two lower-quality trials found cayenne moderately superior to placebo (level of evidence: fair). Serious adverse reactions with herbal therapies appear uncommon (level of evidence: fair). No trials evaluated long-term outcomes.

The European COST guidelines make no recommendation for herbal therapies for acute low back pain, but note that most of the available trials came from the same research group and primarily involved patients with acute exacerbations of chronic low back pain.
The European COST guidelines recommend consideration of capsicum pain plasters for short-term symptomatic pain relief in chronic low back pain.
EVIDENCE REVIEW
Table 34. Systematic reviews on efficacy of medications and herbal therapies for low back pain
Number of included trials (number rated higherquality) * 3 (1) Number of trials not included in any other relevant systematic review 1
Drug Acetaminophen (6 unique trials in two systematic reviews)
Author, year Schnitzer, 411 2004
Type of systematic review Qualitative (efficacy of multiple medications) Qualitative
Duration of treatment in included trials 7 days to 5 weeks (median=4 weeks) 7 days to 4 weeks (median=2 weeks)
Sample sizes in included trials 30 to 60 (median=39)
Interventions evaluated (number of trials) Acetaminophen 4 grams/day (2), 2 grams/day (1) Acetaminophen 4 grams/day (3), 2 grams/day (1), dose not specified (1)
Main conclusions Does not draw specific conclusions regarding acetaminophen Acetaminophen vs. NSAIDs for acute LBP (3 lower-quality RCTs): No differences in 2 rd trials; in 3 trial 2 of 4 evaluated NSAIDs superior to acetaminophen Acetaminophen vs. diflunisal for chronic LBP (1 RCT): Diflunisal superior for proportion reporting no or mild low back pain after 2-4 weeks and for global assessment of efficacy Antidepressant vs. placebo for chronic low back pain (9 RCTs): SMD=0.41 (95% CI 0.22 to 0.61) for pain (9 RCTs); SMD=0.24 (95% CI 0.21 to 0.69) for activities of daily living (5 RCTs) Antidepressants vs. placebo for chronic low back pain (7 RCTs): Antidepressants superior to placebo in 5 of 7 trials
Van Tulder, 412 2000
5 (1)
30 to 70 (median=50)
Antidepressants (10 unique trials in three systematic reviews)
Salerno, 2002479
Quantitative
9 (5)
4 to 8 weeks (median=6 weeks)
16 to 103 (median=50)
Nortriptyline (1), imipramine (2), amitriptyline (1), desipramine (1), doxepine (2), maprotiline (1), paroxetine (2), trazodone (1) Nortriptyline (1), imipramine (1), amitriptyline (2), maprotiline (1), paroxetine (2), fluoxetine (1) trazodone (1)
Schnitzer, 411 2004
Qualitative (efficacy of multiple medications)
7 (4)
4 to 8 weeks (median=8 weeks)
16 to 103 (median=50)
EVIDENCE REVIEW
Drug Antidepressants (10 unique trials in three systematic reviews)
Author, year Staiger, 480 2003
Duration of treatment in included trials 4 to 8 weeks (median=8 weeks)
Interventions evaluated (number of trials) Nortriptyline (1), imipramine (2), amitriptyline (1), maprotiline (1), paroxetine (2), trazodone (1)
Main conclusions Tricyclic and tetracyclic antidepressant vs. placebo for chronic low back pain (5 RCTs): 3 of 5 trials, including the two highest quality trials, found mild to moderate, significant benefits for pain; insufficient evidence on functional status Paroxetine or trazodone vs. placebo for chronic low back pain (3 RCTs): No consistent benefits on pain (SMD ranged from -0.13 to +0.32 in 3 RCTs) Diazepam vs. placebo for acute LBP (1 RCT): Diazepam superior for shortterm pain and overall improvement Tetrazepam vs. placebo for chronic LBP (3 RCTs): RR=1.41 (95% CI 1.08 to 1.85, 2 RCTs) for pain relief of >20% or >16 on a 100 point VAS after 8 to 14 days and RR=1.59 (95% CI 1.03 to 2.38) for global improvement after 8-14 days (2 RCTs) Benzodiazepine vs. skeletal muscle relaxants (3 RCTs): No differences in higherquality trials
Benzodiazepines (8 unique trials in one systematic review)
Van Tulder, 2003488
Qualitative and quantitative
8 (5)
6 to 14 days (median=8 days)
50 to 152 (median=73)
Diazepam (6), tetrazepam (2)
EVIDENCE REVIEW
Drug Herbal therapies (10 unique trials in one systematic review)
Author, year Gagnier, 545 2006
Duration of treatment in included trials 1 to 6 weeks (median=4 weeks)
Interventions evaluated (number of trials) Devils claw (3), white willow bark (3), cayenne (4)
Main conclusions Devils claw more effective than placebo for short-term improvement in pain in 2 RCTs, white willow bark more effective than placebo for short-term improvement in pain in 2 RCTs, topical cayenne more effective than placebo in 3 trials but no more effective than homeopathic gel in one trial NSAIDs for acute LBP (14 RCTs): NSAIDs superior to placebo in 2 of 3 RCTs; 9 of 11 RCTs of NSAID vs. active control found significant improvements from baseline in NSAID group NSAID for chronic LBP (4 RCTs): NSAIDs superior to placebo in 1 RCT. In 3 of 3 RCTs of NSAID vs. active control found significant improvements from baseline in NSAID group NSAID vs. placebo for acute LBP (9 RCTs): RR=1.24 (95% CI 1.10 to 1.41) for global improvement after 1 week (6 RCTs) and RR=1.29 (95% CI 1.05 to 1.57) for not requiring additional analgesics after 1 week (3 RCTs)
Non-steroidal antiinflammatory drugs (57 unique trials in three systematic reviews)
Schnitzer, 2004411
21 (10)
7 days to 8 weeks (median=14 days)
30 to 282 (median=73)
Naproxen (4), ibuprofen (1), indomethacin (4), diclofenac (3), piroxicam (6), diflunisal (6), others (9)
Non-steroidal antiinflammatory drugs (57 unique trials in three systematic reviews)
Van Tulder, 413 2000
51 (15)
34
1-2 days to 6 weeks (median=12 days)
20 to 459 (median=72)
Naproxen (4), ibuprofen (6), indomethacin (10), diclofenac (15), piroxicam (7), diflunisal (8), others (18)
EVIDENCE REVIEW
Drug Non-steroidal antiinflammatory drugs (57 unique trials in three systematic reviews) Skeletal muscle relaxants (38 unique trials in four systematic reviews)
Author, year Vroomen, 100 2000
Type of systematic review Quantitative
Duration of treatment in included trials 2-4 days to 17 days (median=10 days)
Interventions evaluated (number of trials) Indomethacin (1), piroxicam (1), others (2)
Main conclusions NSAID vs. placebo for sciatica (3 RCTs): OR=0.99 (95% CI 0.6-1.7)
Browning, 506 2001
Quantitative (efficacy of cyclobenzaprine for back or neck pain)
14 (5)
11
48 to 1153 (median= 100)
Cyclobenzaprine (14)
Cyclobenzaprine vs. placebo for acute or chronic LBP or neck pain: OR=4.7 for global improvement (10 RCTs, 95% CI 2.7-8.1), SMD=0.41 (95% CI 0.29 to 0.53) for local pain at 1 to 4 days (7 RCTs), SMD=0.54 (95% CI 0.34 to 0.74) for function at 1-4 days (6 RCTs); results similar at >9 days SMR vs. placebo for acute LBP (5 RCTs): SMR superior in 4 of 5 RCTs (no benefit in 1 of 3 RCTs of tizanidine); benefit mostly short-term and early (<7 days)
Schnitzer, 2004411
5 (4)
49 to 361 (median=112)
Tizanidine (3), baclofen (1), other (1)
EVIDENCE REVIEW
Drug Skeletal muscle relaxants (38 unique trials in four systematic reviews)
Author, year Van Tulder, 488, 489 2003
Type of systematic review Qualitative and quantitative
Duration of treatment in included trials Single dose to 21 days (median=7 days)
Interventions evaluated (number of trials) Cyclobenzaprine (5), carisoprodol (3), chlorzoxazone (1), orphenadrine (4) methocarbamol, tizanidine (8), dantrolene (1), baclofen (1), others (5)
Main conclusions Skeletal muscle relaxant (SMR) vs. placebo for acute low back pain (8 RCTs): RR=1.25 (95% CI 1.12 to 1.41) for pain relief of >20% or >16 on a 100 point VAS after 2-4 days (3 RCTs), RR=1.72 (95% CI 1.32 to 2.22) for pain relief after 5-7 days (2 RCTs), RR=2.05 (95% CI 1.05 to 4.00) for global improvement after 2-4 days (4 RCTs) and RR=1.47 (95% CI 0.88 to 2.44) for global improvement after 5-7 days (4 RCTs) Tizanidine vs. placebo for sciatica (1 higher-quality RCT): No difference
Vroomen, 100 2000
Qualitative (efficacy of medications for sciatica)
1 (1)
7 days
112
Tizanidine (1)
*Trials adequately meeting at least half of the quality rating criteria or rated as good or higher-quality if the number of criteria met was not reported CI=confidence interval, NSAID=non-steroidal anti-inflammatory drug, RCT=randomized controlled trial, RR=relative risk, OR=odds ratio, SMD=standardized mean difference, VAS=visual analogue scale
EVIDENCE REVIEW
Table 35. Summary of evidence on medications and herbal therapies for acute low back pain
Number of trials (number rated higher-quality by at least one systematic review) 3 (0)
Drug Acetaminophen
Net benefit* Moderate
Effective versus placebo? Unable to determine (1 lower-quality trial showing no difference) No evidence No evidence Unable to determine (2 trials with inconsistent results)
Inconsistency? Some inconsistency (versus NSAIDs) Not applicable Not applicable Some inconsistency (versus placebo and versus skeletal muscle relaxants)
Comments Little data on serious adverse events
Antidepressants Antiepileptic drugs Benzodiazepines
0 0 5 (3)
No evidence No evidence Moderate
Not applicable Not applicable Direct, with supporting indirect evidence from mixed populations with back and neck pain Direct
Not applicable Not applicable Fair Only evaluated in patients with radicular low back pain No reliable data on risks of abuse or addiction No differences between diazepam and cyclobenzaprine for short-term global efficacy (both superior to placebo) in one large, short-term trial of patients with back or neck pain (mixed duration) Most trials evaluated patients with acute exacerbations of chronic low back pain
Herbal therapies
7 (5)
Moderate
Yes for devils claw (2 trials) and white willow bark (1 trial), unable to determine for cayenne (1 lower-quality trial) Yes (7 trials)
Non-steroidal anti-inflammatory drugs
31 (10)
Moderate
Some inconsistency for cayenne (effective versus placebo but not versus homeopathic gel) No
Fair for devils claw and white willow bark, poor for cayenne Good
Direct
Opioids
1 (1)
Moderate
No evidence
Not applicable
Data available from trials of opioids for other acute pain conditions
Fair
May cause serious gastrointestinal and cardiovascular adverse events. Insufficient evidence to judge benefits and harms of aspirin or celecoxib for low back pain No reliable data on risks of abuse or addiction
EVIDENCE REVIEW
Table 35. Summary of evidence on medications and herbal therapies for acute low back pain
Drug Skeletal muscle relaxants
Effective versus placebo? Yes (19 trials)
Inconsistency? No
Systemic corticosteroids Topical lidocaine Tramadol
1 (1) 0 1 (1)
Not effective No evidence Unable to estimate
No (1 trial) No evidence No evidence
No Not applicable Not applicable
Direct Not applicable Direct
Fair Not applicable Poor
Comments Little evidence on efficacy of antispasticity skeletal muscle relaxants baclofen and dantrolene for low back pain Mostly evaluated in patients with radicular low back pain The only trial compared tramadol to a non-steroidal antiinflammatory drug not available in the U.S.
* Based on evidence showing medication is more effective than placebo, and/or evidence showing medication is at least as effective as other medications or interventions thought to be effective, for one or more of the following outcomes: pain, functional status, or work status. Versus placebo, small benefit defined as 5-10 points on a 100-point Visual Analogue Scale (VAS) for pain (or equivalent), 1-2 points on the Roland Morris Disability Questionnaire (RDQ), 10-20 points on the Oswestry Disability Index (ODI), or a standardized mean difference (SMD) of 0.2-0.5. Moderate benefit defined as10-20 points on a VAS for pain, 2-5 points on the RDQ, 10-20 points on the ODI, or a SMD of 0.5-0.8. Large benefit defined as >20 points on a 100 point VAS for pain; >5 points on the RDQ, >20 points on the ODI, or a SMD of >0.8. Inconsistency defined as >25% of trials reaching discordant conclusions on efficacy (no effect vs. positive effect considered discordant)
EVIDENCE REVIEW
Table 36. Summary of evidence on medications and herbal therapies for chronic or subacute low back pain
Drug Acetaminophen
Effective versus placebo? No trials in patients with low back pain Yes (9 trials)
Inconsistency? No
Directness of evidence? Data available from trials of acetaminophen for osteoarthritis Direct
Comments Asymptomatic elevations of liver function tests at therapeutic doses. Only tricyclic antidepressants have been shown effective for low back pain No evidence on duloxetine or venlafaxine One small trial evaluated topiramate for back pain with or without radiculopathy No reliable data on risks of abuse or addiction
Antidepressants
10 (5)
Small to moderate
No
Good
Antiepileptic drugs Benzodiazepines
1 (1)
Small to moderate Moderate
Yes (1 trial of topiramate) Mixed results (3 trials) Yes for willow bark (1 trial) and cayenne (2 trials), no evidence for devils claw Yes (1 trial)
Not applicable
Direct
Poor
3 (2)
Some inconsistency (versus placebo) No
Direct
Fair
Herbal therapies
3 (0)
Moderate
Direct
Fair
Non-steroidal anti-inflammatory drugs
6 (3)
Moderate
No
Direct
Good
May cause serious gastrointestinal and cardiovascular adverse events. Insufficient evidence to judge benefits and harms of aspirin or celecoxib for low back pain No reliable data on risks of abuse or addiction
Opioids
7 (1)
Moderate
Yes (1 trial)
No
Most trials compare different opioids or opioid formulations
Fair
EVIDENCE REVIEW
Table 36. Summary of evidence on medications and herbal therapies for chronic or subacute low back pain
Drug Skeletal muscle relaxants
Net benefit* Unable to estimate
Effective versus placebo? Unable to determine (5 trials)
Inconsistency? Not applicable
Directness of evidence? Most trials evaluated skeletal muscle relaxants not available in the U.S. or mixed populations of patients with back and neck pain Not applicable Not applicable Direct
Overall quality of evidence Poor
Comments The two higher-quality trials evaluated skeletal muscle relaxants not available in the U.S.
Systemic corticosteroids Topical lidocaine Tramadol
0 0 4 (1)
No evidence No evidence Moderate
No evidence No evidence Yes (1 trial)
Not applicable Not applicable No
Not applicable Not applicable Fair
Mostly evaluated in patients with radicular low back pain
* Based on evidence showing medication is more effective than placebo, and/or evidence showing medication is at least as effective as other medications or interventions thought to be effective, for one or more of the following outcomes: pain, functional status, or work status. Versus placebo, small benefit defined as 5-10 points on a 100-point Visual Analogue Scale (VAS) for pain (or equivalent), 1-2 points on the Roland Morris Disability Questionnaire (RDQ), 10-20 points on the Oswestry Disability Index (ODI), or a standardized mean difference (SMD) of 0.2-0.5. Moderate benefit defined as10-20 points on a VAS for pain, 2-5 points on the RDQ, 10-20 points on the ODI, or a SMD of 0.5-0.8. Large benefit defined as >20 points on a 100 point VAS for pain; >5 points on the RDQ, >20 points on the ODI, or a SMD of >0.8. Inconsistency defined as >25% of trials reaching discordant conclusions on efficacy (no effect vs. positive effect considered discordant)
EVIDENCE REVIEW
Table 37. Summary of evidence on medications for sciatica or radicular low back pain
Drug Antiepileptic drugs Non-selective non-steroidal anti-inflammatory drugs Systemic corticosteroids
Net benefit* Small
Effective versus placebo? Yes (2 trials of gabapentin and 1 trial of topiramate) No (3 trials)
Inconsistency? No
Comments No trials of antiepileptic drugs other than gabapentin or topiramate NSAIDs more effective than placebo in mixed populations of patients with low back pain with or without sciatica
4 (2)
Not effective
No
Direct
Fair
3 (3)
Not effective
No (3 trials)
No
Direct
Good
* Based on evidence showing medication is more effective than placebo, and/or evidence showing medication is at least as effective as other medications or interventions thought to be effective, for one or more of the following outcomes: pain, functional status, or work status. Versus placebo, small benefit defined as 5-10 points on a 100-point Visual Analogue Scale (VAS) for pain (or equivalent), 1-2 points on the Roland Morris Disability Questionnaire (RDQ), 10-20 points on the Oswestry Disability Index (ODI), or a standardized mean difference (SMD) of 0.20.5. Moderate benefit defined as10-20 points on a VAS for pain, 2-5 points on the RDQ, 10-20 points on the ODI, or a SMD of 0.5-0.8. Large benefit defined as >20 points on a 100 point VAS for pain; >5 points on the RDQ, >20 points on the ODI, or a SMD of >0.8. Inconsistency defined as >25% of trials reaching discordant conclusions on efficacy (no effect vs. positive effect considered discordant)
EVIDENCE REVIEW
Acupuncture and related interventions Acupuncture and dry needling

Acupuncture involves the insertion of needles at specific acupuncture points in order to treat or prevent symptoms and conditions. Dry needling also involves the insertion of needles, but targets painful trigger points rather than acupuncture points. Neither technique involves injection of medications through the needle. Results of search: systematic reviews We identified two higher-quality systematic reviews (33 and 35 trials) on efficacy of acupuncture (including electroacupuncture) for primarily chronic low back pain68-70. One review69, 70 also evaluated efficacy of dry needling. We also identified two lower-quality systematic reviews555, 556 , one of which focused on safety of acupuncture556. We excluded an outdated Cochrane review557, four other outdated systematic reviews558-561, and three systematic reviews that did not specifically assess efficacy of acupuncture for low back pain562-564. Results of search: trials Fifty-one unique trials on efficacy of acupuncture were included in three systematic reviews68-70, 555 . Both higher-quality systematic reviews identified significant methodological shortcomings in trials of acupuncture (10 of 33 and 14 of 35 studies rated as higher-quality)68-70. In addition, about one-third of the trials were conducted in Asian settings, which could limit generalizability of findings to the U.S., due to different patient expectations for benefit565 or other factors. Treatment duration varied from one to 20 session of acupuncture. We also identified three recent, large (n=241, 298 and 2,841), higher-quality trials566-568 of acupuncture for low back pain not included in the systematic reviews.
Efficacy of acupuncture versus placebo or sham treatment

Results of the three systematic reviews on efficacy of acupuncture were generally consistent6870, 555 . For acute low back pain, the two higher-quality systematic reviews found sparse evidence (two trials, one higher-quality) on efficacy of acupuncture versus sham acupuncture6870 . Results are inconclusive because of small sample sizes and inconsistent results. One lower-quality trial569 found acupuncture superior to sham acupuncture, but one higher-quality trial570 found no differences. The higher-quality trial only evaluated a single session of single point acupuncture. For chronic low back pain, both higher-quality systematic reviews found acupuncture moderately more effective than no treatment (8 trials, SMD=-0.69, 95% CI -0.98 to -0.4068 and 2 trials, SMD=-0.73, 95% CI -1.19 to -0.2869, 70). Both systematic reviews also found acupuncture moderately more effective than sham treatments (acupuncture or TENS) (7 trials, SMD=-0.54, 95% CI -0.73 to -0.3568 and 2 trials, WMD=-17.8, 95% CI -25.5 to -10.169, 70) for short-term (defined as <6 weeks or <3 months) pain relief. Versus sham acupuncture alone, one of the systematic reviews estimated an SMD=-0.58 (4 trials, SMD=-0.80 to -0.36), equivalent to a WMD=-14.5 on a 100 point VAS68. However, these results may overestimate benefits of
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain acupuncture because they are based on a re-calculation of effect sizes from one trial (resulting in a larger estimate of effect) and also included data from a trial in which most patients randomized to sham acupuncture were excluded from analysis because they crossed over to receive true acupuncture571, 572. A re-analysis based on published effect sizes that excluded the trial with high crossover estimated an SMD for short-term pain intensity of -0.425 (95% CI -0.66 to -0.19) for acupuncture versus sham acupuncture, equivalent to a WMD=-10.6 on a 100 point VAS571. Both systematic reviews found acupuncture associated with moderate short-term improvements in functional status compared to no treatment (SMD=-0.62, 95% CI -0.95 to 0.3068 and SMD=-0.63, 95% CI -1.08 to -0.1969, 70), but not compared to sham therapies. For short- and long-term assessments of overall improvement, acupuncture was superior to either sham treatments or no treatment. A recent, higher-quality trial (n=298) not included in the systematic reviews found acupuncture substantially superior to wait list control for short-term pain relief (mean difference 21.7 points on a 100 point scale), but was inconsistent with the systematic reviews because it found no differences between acupuncture and sham acupuncture (superficial needling at nonacupuncture points) on any outcome at either 8 weeks or with longer follow-up on (through 52 weeks) (Table 38)566. In general, evidence on longer-term (more than 6 weeks after treatment) benefits of acupuncture is sparser and more inconsistent than evidence on shortterm benefits. In one systematic review acupuncture was associated with moderately superior long-term pain relief compared to sham TENS in two trials (SMD=-0.62, 95% CI -1.22 to -0.03) and to no additional treatment in five trials (SMD=-0.74, 95% CI -1.47 to -0.02), but was no better than sham acupuncture in two trials (SMD=-0.59, 95% CI -1.29 to +0.10)68. One higherquality trial included in the systematic reviews that evaluated outcomes one year after treatment found no differences between acupuncture and a self-education book for pain (SMD=-0.35, 95% CI -0.09 to +0.51), and acupuncture slightly inferior to massage (SMD=+0.40, 95% CI +0.09 to +0.71)369. A large, higher-quality trial not included in the systematic reviews found substantial differences between acupuncture and no acupuncture in back function (20 points on a 100 point scale) and back pain (27 points on a 100 point scale) at 3 months, but clinically insignificant differences (less than 5 points) at 6 months (Table 38)568. On the other hand, another higherquality trial not included in the systematic reviews found that some beneficial effects of acupuncture may extend beyond a year567. In this trial, acupuncture was associated with small sustained improvements in SF-36 pain scores compared to usual general practitioner care 24 months after a short course of treatment (mean adjusted difference -8.0 on a 100 point scale, p=0.032) and decreased use of low back pain medications in the last 4 weeks (60% vs. 41%, p=0.03). There were no differences in ODI scores, McGill Present Pain Intensity scores, or other SF-36 dimension scores.
EVIDENCE REVIEW
Table 38. Recent trials of acupuncture not included in systematic reviews

Number of patients Duration of follow-up n=298 8 weeks (versus wait list control) to 52 weeks (versus sham acupuncture)
Author, year Brinkhaus, 566 2006
Thomas, 2006567
n=241 24 months
Witt, 2006568
n=2,841 6 months
Main results Acupuncture vs. sham acupuncture vs. wait list control at 8 weeks; acupuncture vs. sham acupuncture at 52 weeks Pain intensity (difference from baseline, 0 to 100 scale): -28.7 vs. -23.6 vs. -6.9 at 8 weeks (p=0.26 for acupuncture vs. sham; p<0.001 for acupuncture vs. wait list control); 39.2 vs. 44.9 at 52 weeks (p=0.20) Back function (mean, 0 to 100 scale): 66.8 vs. 62.9 vs. 57.7 at 8 weeks, 66.0 vs. 63.1 at 52 weeks (NS) Pain Disability Index (mean, 0 to 100 scale): 18.8 vs. 21.5 vs. 27.1 at 8 weeks, 19.0 vs. 23.0 at 52 weeks (NS) SF-36 physical health scale (mean): 40.5 vs. 36.2 vs. 33.9 at 8 weeks (p=0.004 for acupuncture vs. sham and p<0.001 for acupuncture vs. wait list control); 38.9 vs. 36.1 at 52 weeks (p=0.07) SF-36 mental health scale: No differences at 8 weeks, 50.5 vs. 47.2 at 52 weeks (p=0.04) SF-36 pain scale (mean): 58.8 vs. 50.7 vs. 39.9 at 8 weeks (p=0.01 for acupuncture vs. sham), 52.4 vs. 44.0 at 52 weeks Depression: No significant differences Acupuncture versus usual care SF-36 Pain score, mean adjusted difference between interventions: -5.6 (95% CI -11.4 to +0.2) at 12 months, -8.0 (95% CI -13.2 to -2.8) at 24 months (favors acupuncture) McGill Present Pain Intensity: No difference at 12 or 24 months ODI Score: No difference at 12 or 24 months Pain-free in last 12 months: 18% vs. 8% (p=0.06) Use of low back pain medication in last 4 weeks: 60% vs. 41% (p=0.03) Acupuncture vs. no acupuncture (difference in change from baseline, positive values favor acupuncture) Back function loss (Hannover Functional Assessment Questionnaire, 0 to 100 scale): 22.0 (95% CI 19.3 to 24.7) at 3 months, 3.7 (95% CI 0.7 to 6.7) at 6 months Low Back Pain Rating Scale (0 to 100): 27.2 (95% CI 20.9 to 24.5) at 3 months, 2.7 (95% CI -0.3 t0 5.7) at 6 months SF-36 Physical Component score: 4.7 (95% CI 4.0 to 5.4) at 3 months, 0.6 (95% CI -0.2 to 1.3) at 6 months SF-36 Mental Component score: 2.1 (95% CI 1.4 to 2.8) at 3 months, 0.2 (95% CI -0.6 to 1.0) at 6 months
Quality score* 8/10
7/10
8/10
Efficacy of acupuncture versus other interventions

For acute low back pain, both systematic reviews included one higher-quality trial that found no differences between acupuncture and NSAIDs for pain relief68-70. In mixed populations of patients with acute or longer duration low back pain, there was no significant difference between acupuncture and moxibustion (one lower-quality trial), but electroacupuncture was superior to TENS for pain relief (one higher-quality trial).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For chronic low back pain, both systematic reviews found acupuncture inferior to spinal manipulation for short-term pain relief, though the number and quality of trials was limited68-70. One of the systematic reviews calculated an SMD=+1.32 (95% CI +0.77 to +1.87) from two lower-quality trials68. Neither found any differences between acupuncture and other active therapies (massage, analgesic medication, or TENS, each comparison evaluated in one to four trials).
Efficacy of one acupuncture technique versus another acupuncture technique

The Cochrane review compared benefits of different acupuncture techniques (8 trials, 2 higherquality)69, 70. In one higher-quality trial, deep stimulation was superior to superficial stimulation at short-term follow-up573. In the other higher-quality trial, there was no difference between manual acupuncture and electroacupuncture574. There was insufficient evidence (single comparisons from lower-quality trials) to judge comparative efficacy of other acupuncture techniques.
Efficacy of dry needling

For acute low back pain, the Cochrane review69, 70 included one lower-quality trial that found no differences between one session of dry needling versus trigger point injection with lidocaine and steroid, trigger point injection with lidocaine only, or cooling spray over the trigger point area followed by acupuncture575. For chronic low back pain, one higher-quality trial found superficial needling of trigger points superior to sham TENS for immediate pain relief140 and one lowerquality trial found dry needling added to a regimen of physiotherapy, occupational therapy, and industrial assessments superior to the regimen without dry needling576 for short- and intermediate-term functional status. Harms The Cochrane review found only 14 of 35 trials reported any complications or side effects69, 70. Minor complications occurred in 5% (13/245) patients receiving acupuncture, 0% (0 of 156) receiving sham, and 10% (21/205) receiving other interventions. None of the complications were fatal or required hospitalization. Another systematic review of prospective (randomized and non-randomized) studies of over 250,000 acupuncture treatments for various conditions found wide variation in rates of adverse events, ranging from 1% to 45% for needle pain, and 0.03% to 38% for bleeding556. The wide ranges are probably related to multiple factors, including differences in populations, interventions, and methods for defining, identifying, and reporting adverse events. Feelings of faintness and syncope were uncommon, with an incidence of 0% to 0.3%. Serious adverse events were rare. Pneumothorax was reported in two patients, and there were no cases of infections. Costs Three trials estimated cost-effectiveness for acupuncture. One found routine offering of acupuncture associated with an incremental cost-effectiveness ratio of $8,185/QALY (95% CI $369 to $54,090) relative to usual care (converted from British pounds to U.S. dollars at January
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain 2007 exchange rate of 1 British pound = 1.93 U.S. dollars)577, and another estimated costeffectiveness of 10,526/QALY (about $13,684 U.S./QALY in January 2007) for acupuncture versus no acupuncture568. The third trial found no significant differences in back pain-related HMO costs between patients randomized to acupuncture, massage, and self-care (massage was the most effective therapy for patient outcomes)369.
Summary of evidence
For acute low back pain, there is insufficient evidence (two trials, one higher-quality) to judge efficacy of acupuncture compared to sham acupuncture, as results are inconsistent and the acupuncture intervention was suboptimal in the higher-quality trial (level of evidence: poor). For acute low back pain, one higher-quality trial found no difference between acupuncture and NSAIDs (level of evidence: poor). For mixed populations of patients with acute and longer duration low back pain, one lowerquality trial found no difference between acupuncture and moxibustion and one higher-quality trial found electroacupuncture superior to TENS (level of evidence: poor to fair). For chronic low back pain, there is consistent evidence from multiple trials that acupuncture is moderately effective for short-term pain relief compared to no treatment and sham TENS in patients with chronic low back pain, and superior to no treatment for short-term functional outcomes (level of evidence: good). For chronic low back pain, evidence on efficacy of acupuncture versus sham acupuncture is inconsistent. Although four trials (three higher-quality) found acupuncture moderately more effective than sham acupuncture for short-term pain relief, a recent, large (n=298), higherquality trial found no significant differences (level of evidence: fair). For chronic low back pain, evidence on longer-term (>6 weeks) outcomes is sparse but suggests acupuncture is more effective than sham TENs or no treatment, though benefits may become attenuated with longer follow-up. One recent, higher-quality trial found small beneficial effects on pain persist for up to 24 months (level of evidence: fair). Acupuncture was substantially inferior to spinal manipulation in two lower-quality trials (level of evidence: fair) There is no clear evidence of significant differences between acupuncture and TENS (4 trials), medications (3 trials), or massage (1 trial) (level of evidence: fair). Dry needling alone was not effective compared to trigger point injections or acupuncture for acute low back pain in one lower-quality trial (level of evidence: poor), but was more effective than placebo or when added to other interventions for chronic low back pain in two trials (one higher-quality) (level of evidence: fair). Serious adverse events with acupuncture appear rare, though rates of minor events vary widely and were often poorly reported (level of evidence: fair).
EVIDENCE REVIEW

The AHCPR guidelines recommend against invasive needle acupuncture and other dry needling techniques for patients with acute low back problems (strength of evidence: D). The VA/DoD and UK RCGP guidelines on acupuncture for acute low back pain are similar. The European COST guidelines make no recommendations on acupuncture for acute low back pain, and found insufficient evidence to recommend acupuncture for chronic low back pain.
Acupressure
Acupressure is a non-invasive method that involves manipulation of the skin and soft tissues with the fingers or other blunt devices instead of needles on acupuncture points. It is less wellstudied than acupuncture. Results of search: systematic reviews We found no systematic review evaluating efficacy of acupressure. Results of search: trials From nine potentially relevant citations, we identified two higher-quality, open-label trials of acupressure for chronic low back pain578, 579. Both were conducted in Taiwan by the same group of investigators. Duration of follow-up in both trials was six months.
Efficacy of acupressure versus physical therapy

For chronic low back pain, one higher- and one lower-quality trial found acupressure more effective than unstandardized physical therapy that consisted of multiple techniques at the discretion of the physical therapist (Table 39)578, 579. In the one trial that reported functional outcomes, acupressure was associated with moderate effects on the RDQ (mean difference 5.36, 95% CI -7.21 to -3.52) but only small effects on the ODI (-7.99, 95% CI -10.8 to -5.17) that persisted through six months579. Days off from work or school also improved more in the acupressure group (mean difference compared to baseline -2.79 days, p<0.0001). In both trials, acupressure was moderately to substantially superior to physical therapy for pain relief, with mean differences between treatments -27.2 on a 100 point VAS (p<0.0001) and -4.46 on the 0 to 45 point Short-Form Pain Questionnaire (p=0.0001) after 6 months. In one trial, effects on pain relief were about twice as high as seen with most other conventional interventions or acupuncture579.
EVIDENCE REVIEW
Table 39. Trials of acupressure versus physical therapy

Number of patients Duration of follow-up n=146 6 months n=158 6 months
Author, year Hsieh, 2004578 Hsieh, 2006579
Main results Acupressure versus physical therapy Short-form Pain questionnaire, mean change from baseline: -8.69 vs. -4.23 (p=0.0001) Acupressure versus physical therapy RDQ score, difference in mean change from baseline: -5.36, 95% CI -7.21 to -3.52 (p<0.0001) Modified ODI score, difference in mean change from baseline: -7.99, 95% CI -10.8 to -5.17 (p<0.0001) Pain (VAS, 0 to 100), difference in mean change from baseline between interventions: -27.12 (p<0.0001)
Quality score* 6/10 5/10
Harms One of the trials reported no adverse events in the acupressure group578. The other trial did not report adverse events. Costs We found no studies evaluating costs.
Summary of evidence
There is no evidence on acupressure for acute low back pain. For chronic low back pain one higher and one lower-quality trial found acupressure moderately to substantially more effective than physical therapy for pain and functional outcomes. However, it is not clear if these results can be generalized to other settings because both trials were conducted in Taiwan by the same investigators and the physical therapy interventions were not standardized (level of evidence: fair) Acupressure does not appear associated with serious adverse events, but harms were only reported by one trial (level of evidence: fair).

The other guidelines do not address acupressure.
Neuroreflexotherapy
Neuroreflexotherapy is a technique characterized by the temporary implantation of staples superficially into the skin over trigger points in the back and referred tender points in the ear. Like acupuncture, it involves the use of puncture devices in the skin. However, neuroreflexotherapy is believed to stimulate different zones of the skin. Results of search: systematic reviews We identified one recent, higher-quality Cochrane review (three trials, two rated higherquality580, 581) on effectiveness of neuroreflexotherapy for chronic low back pain582.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials The Cochrane review included three trials (two rated higher-quality) on neuroreflexotherapy for chronic low back pain580, 581. The same principal investigator conducted all three trials in Spain (total number of patients: 273). We did not search for additional trials.
Efficacy of neuroreflexotherapy versus sham neuroreflexotherapy

The two higher-quality trials580, 581 both found neuroreflexotherapy substantially superior to sham therapy for short-term (up to 45 days) pain relief. In one trial, the proportion of patients with pain relief was 96% with neuroreflexotherapy versus 2.3% with sham (p<0.0001)580. In the other trial, neuroreflexotherapy was associated with an average improvement in spontaneous pain of 3.09 (on a 10 point scale) compared to 0.34 with sham treatment that was statistically significant (p<0.001), but not clinically significant581. One580 of the two trials found neuroreflexotherapy superior to sham therapy on a variety of functional and work-related outcomes, but the other581 found no significant differences.
Efficacy of neuroreflexotherapy versus usual care

The third, lower-quality trial compared neuroreflexotherapy to usual care583. It found neuroreflexotherapy moderately superior for short-term (60 days) pain relief (average improvement 5.50 on a 10 point scale versus 1.92, p<0.0005) and substantially superior for functional status (average improvement 8.67 on RDQ scale versus 2.05, p=0.007). Number of days on sick leave and duration of sick leave (average 3.2 vs. 105.2 days, p=0.001) and use of health care services were also lower in the neuroreflexotherapy group after one year. There were no differences in quality of life.
Harms
One trial found a higher incidence of adverse effects in the sham therapy group (65% vs. 9%), primarily due to gastric discomfort probably associated with increased NSAID use580. Skin tightness was associated with implantation of staples, but did not require early extraction in any patient. Scarring was not specifically reported in any trial, but is not believed to be an important problem because of the superficial nature of the staple implantations. Costs One trial included a cost-effectiveness analysis that found neuroreflexotherapy dominated usual care (total costs lower and clinical outcomes superior)583. Neuroreflexotherapy was associated with median costs of $800 compared to $3,800 with usual care, and superior by an average of 5.5 points on the RDQ Scale (0 to 24).
Summary of evidence
There is no evidence on efficacy of neuroreflexotherapy for acute low back pain. For chronic low back pain, three trials (two higher-quality) found neuroreflexotherapy substantially superior to sham therapy or usual care for short-term pain relief. All of the trials were conducted in Spain by the same principal investigator at a specialized center, potentially limiting applicability of results to other settings (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Evidence on beneficial effects of neuroreflexotherapy relative to sham treatment on functional outcomes is mixed (level of evidence: fair). The single lower-quality trial assessing one-year outcomes found lower self-reported sick leave and consumption of health care resources following neuroreflexotherapy compared to usual care (level of evidence: poor).

The European COST guidelines recommend consideration of neuroreflexotherapy for patients with moderate or severe (>3/10 on VAS) chronic low back pain.
Educational interventions Back schools

The original Swedish back school was introduced in 1969584, 585. The basic elements of back schools consist of an educational and skills program, including exercises, in which all lessons are given to groups of patients and supervised by a therapist or medical specialist. However, the content and intensity of back schools meeting this basic definition can vary widely. Results of search: systematic reviews We identified one recent, higher-quality Cochrane review of 19 trials (6 rated higher-quality) on back schools for acute or chronic low back pain586, 587. We also included two other recent, lower-quality systematic reviews588, 589. Another recent, higher-quality systematic review evaluated factors that could predict better outcomes from back schools and multidisciplinary rehabilitation (results not clearly separated for the two interventions)590. We excluded an outdated Cochrane review591 and ten other outdated systematic reviews193, 345, 346, 386, 592-597. Results of search: trials Thirty-one unique trials were included in three systematic reviews of back schools586-589. We did not search for additional trials.
Efficacy of back schools versus placebo or wait list control

For acute or subacute low back pain, the Cochrane review included one lower-quality trial598 that found back school superior to sham treatment (shortwave therapy at the lowest intensity) for short-term recovery and return to work, but not for short-term pain or long-term recurrences586, 587 . For chronic low back pain, the Cochrane review found conflicting evidence from eight trials (2 higher-quality599, 600) on effectiveness of back schools versus placebo or wait list controls. For short-term outcomes, seven RCTs found no benefit from back schools. For long-term outcomes, one higher-quality trial600 found beneficial effects on functional status and return to work, though two lower-quality trials601, 602 found no long-term benefits. Results of back schools were generally more promising in trials conducted in an occupational setting (moderate evidence for improved short- and intermediate-term pain and return to work) and for more intensive (three to five-week stays in specialized centers) programs consisting of modifications of the original Swedish back school. In general, however, any benefits associated with back schools were small.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Conclusions of two other (lower-quality) systematic reviews were generally consistent with the Cochrane review588, 589. The three systematic reviews included a total of 13 trials of back schools versus placebo or wait list controls for chronic or subacute low back pain586-589. No trial evaluated efficacy of back school versus placebo or wait list control in patients with exclusively acute (<4 weeks) low back pain).
Efficacy of back schools versus other interventions

For acute or subacute low back pain, the Cochrane review586, 587 included four trials (two higher quality603-605) on effectiveness of back school versus other treatments (physical therapy, usual care, or advice). Although one higher-quality trial found back school associated with decreased sick leave compared to usual care after 200 days (30% vs. 60%) and 5 years (19% vs. 34%)603, 604 , the other three trials reported no significant differences between back school and other treatments598, 605, 606. Only one of these trials (rated lower-quality) evaluated patients with exclusively acute (<4 weeks) low back pain606. It found no differences between back school and advice. For chronic low back pain, the Cochrane review included six trials (one higher-quality, four in occupational settings) that found back schools slightly superior to other treatments (exercises, spinal manipulation, myofascial therapy, or some kind of advice) for short and intermediate-term pain relief and improvement in functional status, but not for long-term outcomes586, 587. Altogether, the three systematic reviews included a total of 13 unique trials of back schools versus other interventions for chronic or subacute low back pain586-589. None of the systematic reviews found sufficient evidence to conclude that back schools are clearly effective. The systematic review on factors associated with better outcomes after back school or multidisciplinary rehabilitation found consistent evidence that higher baseline pain is associated with worse outcomes.590 It also found several work-related parameters (such as high satisfaction) and low levels of active coping skills at baseline associated with better outcomes. Many predictors were evaluated in only one study or lacked consistent predictive value, in part due to flaws in the studies. Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
For acute low back pain, back schools were no better than advice in a single lower-quality trial (level of evidence: poor). For acute or subacute low back pain, back schools were superior to placebo in a single lowerquality trial for short-term recovery and return to work, but not for pain or long-term recurrences (level of evidence: poor).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For acute or subacute low back pain, evidence on efficacy of back schools versus physical therapy, usual care, or advice was inconsistent, though most studies found no differences (four trials, two higher-quality) (level of evidence: fair). For chronic low back pain, evidence on effects of back schools versus placebo or wait list controls is inconsistent, though most trials found no beneficial effects (level of evidence: fair). For chronic low back pain, back schools are slightly superior to exercises, spinal manipulation, myofascial therapy, or advice for short-term pain and functional status, but not for long-term outcomes (level of evidence: fair). More intensive back school programs based on the original Swedish program and back school programs in occupational settings appear to be the most effective (level of evidence: fair).

The AHCPR guidelines found back schools in the workplace that include worksite-specific education may be effective adjuncts to individual education efforts by the clinician for the treatment of patients with acute low back problems (strength of evidence: C). The AHCPR guidelines found that efficacy of back schools in nonoccupational settings had not been proven (strength of evidence: C).
The European COST guidelines recommend considering back schools where information
given is consistent with evidence-based recommendations for short-term (<6 weeks) pain relief and improvements in functional status. They do not recommend back schools as a treatment for chronic low back pain when aiming at long-term effects (>12 months).
Brief educational interventions

We defined brief interventions as a detailed clinical examination by a physician and/or physiotherapist followed by individualized back education and advice. As we defined them, brief educational interventions typically require several hours and are usually completed in one or two sessions. Brief interventions differ from back schools because they dont involve group education and exercises. They also are distinct from multidisciplinary rehabilitation, which generally includes a coordinated cognitive-behavioral or other psychological therapy component as well as a supervised rehabilitation program. Results of search: systematic reviews We found no systematic reviews of brief educational interventions. Results of search: trials We identified three trials (all in workers with low back pain for less than three months) that evaluated brief educational interventions in workers with subacute low back pain (Table 40)603, 604, 607-610 . Two were rated higher-quality603, 604, 609, 610. A fourth, higher-quality trial evaluated a brief educational intervention for chronic low back pain611, 612.
EVIDENCE REVIEW
Efficacy of brief educational interventions versus usual care

For workers on sick leave for 8 to 12 weeks due to low back pain, one lower-quality trial found a brief educational intervention (a single visit to a spine clinic with a detailed examination by a physiatrist and physical therapist and advice to remain active) associated with no differences compared to usual care in the proportion who continued to report low back pain at 6 months or 1 year or the proportion off sick leave at 3 years, though patients randomized to the intervention were more likely to be off sick leave at 1 year (OR=1.60, 95% CI 1.08 to 2.39)607, 608. In workers with bothersome low back pain for up to 3 months, a higher-quality trial found a brief intervention associated with fewer sick days after 1 year (19 versus 41 days, p=0.02)609 and 2 years (30 versus 62 days, p=0.03) compared to usual care 610. There were no differences in pain or ODI scores at any follow-up period. A smaller proportion of patients reported severe symptoms at 3 months, but not with longer duration of follow-up. In workers with back pain for four to twelve weeks, another higher-quality trial found a brief educational intervention (detailed examination plus three hours of advice for light duty) associated with a lower likelihood for sick leave (19% versus 34%, p<0.001) or permanent disability (49% vs. 69%, p<0.03) after five years compared to usual care603, 604.
Table 40. Trials of brief educational interventions versus usual care
Author, year Indahl, 1995 and 1998603, 604
Karjalainen, 2003 609, 610 and 2004
n=170 2 years
608
Molde Hagen, 2000 and 2003607,
n=510 3 years
Main results Brief educational intervention versus usual care On sick leave: 30% vs. 60% at 200 days, 19% vs. 34% at 5 years (p<0.001) Long term or permanent disability status after 5 years: 19% vs. 34% (p<0.001) Sick listed > 2 x: 49% vs. 69% (p<0.03) Brief educational intervention versus mini intervention plus work site visit versus usual care Pain intensity: 3.5 vs. 3.2 vs. 3.4 at 24 months (NS) Very or extremely bothersome symptoms during the past week: 29% vs. 35% vs. 48% at 3 months, 23% vs. 20% vs. 29% at 24 months (p=0.048 for A vs. C at 3 months, NS for B vs. C) ODI: 19 vs. 18 vs. 18 at 24 months (NS) Days on sick leave: 30 vs. 45 vs. 62 (p=0.030 for A vs. C, NS for B vs. C) Brief educational intervention versus usual care LBP still present at 1 year: 47% vs. 52% (NS) Off sick leave at 1 year: 69% vs. 57% (p<0.05) Off sick leave at 3 years: 64% vs. 62% (NS) New episodes of sick leave due to LBP (through 3 years): 62% (147/237) vs. 61% (135/220) (NS)
Quality score* 7/9
7/9
4/9
Efficacy of brief educational interventions versus brief educational interventions plus manipulation and exercise
For chronic low back pain, one higher-quality trial found a brief intervention (consisting of a physician consultation and individualized reassurance, education, and back advice with a repeat visit at 5 months) slightly inferior to the brief intervention plus manipulation (using a muscle energy technique involving contraction of muscles against an applied counterforce) and
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain exercise for pain relief at 12 and 24 months (difference of about 6 points on a 100 point pain scale at 12 months and about 3 points at 24 months) (Table 41)611, 612. Effects on disability, health-related quality of life and number of days of sick leave through 1 year (20 vs. 14 days) were similar.
Table 41. Trial of brief educational intervention versus brief educational intervention plus exercise and manipulation
Author, year Niemisto, 2003 and 2005611, 612
Main results Brief educational intervention versus brief intervention plus manipulation (using a muscle energy technique) plus exercise Pain (0 to 100): 32.2 vs. 25.7 at 12 months (p=0.01), 33.1 vs. 30.7 at 24 months ODI: 16.5 vs. 13.7 at 12 months (p=0.20), 14.0 vs. 12.0 at 24 months Health-related Quality of Life (15D): No differences Number of days of work absence through 1 year: 20 vs. 14
Quality score* 8/9
Harms No trial reported adverse events. Costs Cost-benefit analyses of two trials of workers with subacute low back pain found the brief educational intervention superior to usual care by an average of $3,497608 and 4,839 (about $6,290 U.S.)610, largely due to decreased sick leave in the first year after the intervention. For chronic low back pain, a third trial estimated an incremental cost-effectiveness of $512 per additional point of improvement on a 100-point pain scale for combined manipulation and exercise plus a brief intervention, versus the brief intervention alone612.
Summary of evidence
In workers with subacute low back pain, three trials (two higher-quality) found a brief educational intervention associated with beneficial effects on sick leave compared to usual care, with most benefits observed in the first year after the intervention. There were no clear effects on pain or functional status (level of evidence: good). For chronic low back pain, a brief intervention was only slightly inferior to the brief intervention plus exercise and manipulation (one higher-quality trial) (level of evidence: fair).

The European COST guidelines recommend brief educational interventions that encourage a return to normal activity to reduce sickness absence and disability associated with chronic low back pain.
EVIDENCE REVIEW
Exercise and related interventions Exercise

We defined exercise therapy as either a supervised exercise program or formal home exercise regimen. Exercise therapy can range from programs aimed at general physical fitness or aerobic exercise to programs more specifically aimed at muscle strengthening, flexibility, or stretching, or different combinations of these elements. Results of search: systematic reviews We identified a recent, higher-quality Cochrane review (61 trials) on efficacy of exercise therapy for nonspecific low back pain613, 614. A separate article based on the Cochrane review reported results of meta-regression and modeling to identify features of more effective exercise regimens615. We included five other recent systematic reviews of exercise therapy, all with a less comprehensive scope than the Cochrane review616-620. Two were rated higher-quality617, 619 . We excluded an outdated Cochrane review621, one systematic review that only evaluated trials included in the outdated Cochrane review622; nine other outdated systematic reviews193, 344346, 386, 623-626 , one review that didnt use systematic methods627, and one systematic review focusing on rehabilitation following lumbar disc surgery628. Results of search: trials Seventy-nine unique trials of exercise therapy were included in six systematic reviews613, 614, 616620 . Most of the trials evaluated patients with chronic low back pain and had methodological shortcomings. For example, 43 of 61 trials in the Cochrane review evaluated patients with chronic low back pain613, 614. Only eight of the 61 trials met all four quality rating criteria used by this review. We also identified a recent, large (n=1334), lower-quality trial not included in the systematic reviews that compared manipulation, exercise, or both to usual care629. Another recent trial that compared different physical therapy regimens for spinal stenosis (consisting of different types of exercise plus manipulation/mobilization) is reviewed in Key Question 10 (combination therapies)630.
Efficacy of exercise therapy versus placebo or usual care

For acute low back pain, the Cochrane review613, 614 found exercise therapy superior to usual care or no treatment in only two (one higher quality349) of nine trials. Among trials with data that could be pooled, there was no difference between exercise therapy usual care for pain relief (3 trials) or functional outcomes (3 trials) at any time period. The Cochrane review also included five trials that compared exercise to usual care or no treatment in patients with subacute low back pain. Although two trials631, 632 (one higher-quality632) found a graded-activity intervention in the workplace associated with reduced absenteeism compared to usual care and one lowerquality trial633 found an exercise program combined with behavioral therapy associated with improved functioning compared to usual care, pooled results showed no significant differences in pain scores (5 trials, WMD=1.89 on a 100 point scale, 95% CI -1.13 to 4.91) or functional outcomes (4 trials, WMD=1.07, 95% CI -3.18 to 5.32).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For chronic low back pain, the Cochrane review found exercise slightly to moderately superior to no treatment for pain relief at the earliest follow-up period (19 trials, WMD=10 points on a 0 to 100 scale, 95% CI 1.31 to 19.09), though not for functional outcomes (17 trials, WMD=3.00 on a 0 to 100 scale, 95% CI -0.53 to 6.48)613, 614. Results were similar at later follow-up. The differences were somewhat greater in health care settings (WMD=13.3 points for pain, 95% CI 5.5 to 21.1 and WMD=6.9 for function, 95% CI 2.2 to 11.77) than in occupational or general population settings. Three other systematic reviews were less comprehensive than the Cochrane review, but reached consistent conclusions617, 618, 620. A higher-quality systematic review that focused on work outcomes (14 trials) found exercise (including exercise as part of a multidisciplinary intervention) slightly reduced sick leave during the first year (SMD=-0.24, 95% CI -0.36 to -0.11) and improved the proportion returned to work (RR=1.37 at 1 year, 95% CI 1.05 to 1.78), though no benefit was observed in the severely disabled subgroup (>90 days sick leave under usual care) or in patients receiving disability payments617. One lower-quality, qualitative systematic review found positive results on at least one outcome (pain or back specific function) for all six included trials that compared exercise therapy to wait list, advice, or TENS618. Another lower-quality systematic review620 that focused on exercise for spondylolysis and spondylolisthesis included only two trials (one higher-quality634), both of which found exercise superior to usual care634 or sham exercise635. The recently published, large (n=1334) UK BEAM Trial also reported results consistent with the Cochrane review (Table 42)629. In patients with low back pain for at least 28 days, exercise was only marginally superior to usual care for pain and disability.
Table 42. Results of the UK BEAM trial
Author, year UK BEAM Trial, 2004629
Main results Manipulation + exercise versus manipulation alone versus exercise alone (all results are absolute net benefit relative to usual care at 12 months) RDQ Questionnaire (0 to 24 scale): 1.30 (95% CI 0.54 to 2.07) vs. 1.01 (95% CI 0.22 to 1.81) vs. 0.39 (95% CI -0.41 to 1.19) Modified Von Korff pain score (0 to 100 scale): 6.71 (95% CI 2.47 to 10.95) vs. 5.87 (95% CI 1.58 to 10.17) vs. 4.90 (95% CI 0.30 to 9.50) Modified Von Korff disability score (0 to 100 scale): 6.71 (95% CI 2.62 to 10.80) vs. 5.65 (95% CI 1.57 to 9.72) vs. 4.56 (95% CI 0.34 to 8.78)
Quality score* 2/9
Efficacy of exercise therapy versus other interventions

For acute low back pain, the Cochrane review included seven trials that found no difference between exercise therapy and other non-invasive treatments for pain (WMD=0.31 point, 95% CI
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain -0.10 to 0.72) and function613, 614. For chronic low back pain, exercise was associated with statistically significant but only small benefits compared to other non-invasive treatments for pain (WMD=5.93 points, 95% CI 2.21 to 9.65) and function (WMD=2.37 points, 95% CI 0.74 to 4.0). One higher-quality619 and one lower-quality trial616 focused on efficacy of the McKenzie method. The McKenzie method is an exercise-based intervention which places patients in one of three broad categories (derangement, dysfunction, and postural syndrome) to guide therapy. Patients are taught to perform exercises that centralize their symptoms and to avoid movements that peripheralize them, using techniques that primarily rely on patient-generated forces and emphasizing self-care. The higher-quality systematic review (11 trials) found conflicting evidence on effectiveness of McKenzie therapy versus other interventions619. For acute low back pain (9 trials), the McKenzie method was slightly superior (mean differences <5 points on 100 point pain and disability scales) to passive therapies (educational booklets, bed rest, ice packs, and massage), but slightly inferior to advice to stay active, with inconsistent results compared to spinal manipulation. For back pain of mixed duration, a lower-quality and less comprehensive (5 trials) systematic review found the McKenzie method associated with small short-term improvements in short-term pain and disability compared to other non-invasive interventions (WMD=-8.6, 95% CI -13.7 to -3.5 for pain and WMD=-5.4, 95% CI -8.4 to -2.4 for function), but no better for intermediate term disability or work absence616. The recent UK BEAM Trial629 found no clear differences between exercise therapy and manipulation (see Table 42 above).
Efficacy of one type of exercise therapy versus another

A lower-quality, qualitative systematic review found no clear differences between different exercise regimens, including no differences between supervised exercise and home exercise programs in three trials618. A more detailed meta-regression that analyzed potential predictors of greater effectiveness was conducted by the authors of the Cochrane review615. Compared to home exercises only, it found improved pain scores with individually designed programs (5.4 point improvement in pain scores, 95% credible interval 1.3 to 9.5), supervised home exercise (6.1 points, credible interval -0.2 to 12.4), group exercise (4.8 points, 95% credible interval 0.2 to 9.4 points), and individually supervised programs (5.9 points, 95% credible interval 2.1 to 9.8 points). High-dose exercise programs (20 or more hours of intervention time) were not superior to low-dose programs. Interventions that included additional non-invasive therapy were superior (5.1 points, 95% credible interval 1.8 to 8.4 points) to those without additional non-invasive therapy. The exercise regimens that were most effective used stretching and strengthening, though there was some overlap with other types of exercise (aerobic, mobilizing, or other specific exercise methods). The meta-regression suggested that an intervention incorporating all of the features of an effective exercise regimen would improve pain scores by 18.1 points (95% credible interval 11.1 to 25.0 points) compared to no treatment and by 13.0 points (95% credible interval 6.0 to 19.9 points) compared to other non-invasive treatment. Function would improve by 5.5 points (95% credible interval 0.5 to 10.5) compared to no treatment and by 2.7
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain points (95% credible interval -1.7 to 7.1) compared to other non-invasive treatment. No trials of such an intervention are available to confirm these estimates. For acute low back pain, a higher-quality systematic review included one higher-quality trial that found marginal differences between the McKenzie method and flexion exercises (mean differences=2 points on a 0 to 100 scale) for acute pain, though a second, lower-quality trial found the McKenzie method associated with large benefits on short-term (5 days) disability (mean difference=-22 points on a 0 to 100 scale, 95% CI -26 to -18)619. For chronic low back pain, there were no clear differences between the McKenzie method and either flexion exercise or strengthening exercises (one trial for each comparison). Harms One systematic review attempted to evaluate adverse events associated with exercise therapy, but found insufficient evidence to generate reliable estimates618. It found 29 of 51 trials did not report adverse events at all and nine others gave insufficient information on adverse events. Reported adverse events include two myocardial infarctions (neither thought related to exercise) and increased pain. Costs Two trials calculated cost-effectiveness ratios for exercise therapies. The UK BEAM trial found the addition of exercise associated with an incremental cost-effectiveness of 8300/QALY (about $16019/QALY) relative to best care, though exercise was dominated by the combination of exercise and manipulation (more costly and less effective)629. Another British trial estimated an incremental cost-effectiveness of 3,010/QALY (about $5,809 U.S./QALY) for physiotherapy relative to physiotherapy advice alone, but a high likelihood of no significant differences between interventions636. Two trials compared costs between exercise programs and usual care. One found no significant cost differences related to health services, equipment, and days off work between a progressive exercise program and usual primary care633. A cost-minimization analysis from another trial found no differences in total costs (direct and indirect) between both standard or intensive physical therapy (including exercise) and usual care637. Three other trials included cost-benefit analyses of exercise therapy versus other interventions. For acute low back pain, one trial found no significant cost difference between exercise and either bed rest or usual activities (usual activities associated with more rapid recovery in this trial)349. Another trial found exercise associated with greater costs compared to providing a selfcare education book ($437 versus $153), and only marginally better outcomes367. Studies that compared costs between exercise therapy and spinal manipulation are discussed in the spinal manipulation section.
EVIDENCE REVIEW
Summary of evidence
For acute low back pain, evidence on efficacy of exercise relative to placebo or no treatment is somewhat inconsistent, though most trials found no benefit (level of evidence: fair). For chronic low back pain, numerous trials found exercise moderately superior to placebo for pain relief and work-related outcomes, though exercise was not associated with beneficial effects on functional outcomes (level of evidence: good). For either acute or chronic low back pain, numerous trials found no consistent, clinically significant differences between exercise therapy and other non-invasive interventions (level of evidence: good). Exercise regimens incorporating features such as individual tailoring, supervision, stretching, and strengthening were associated with the best outcomes in a meta-regression analysis (level of evidence: fair).
There are no clear differences in four trials (two higher-quality) between the McKenzie method
and flexion or strengthening exercises, with only one lower-quality trial finding the McKenzie method superior (level of evidence: fair).

The AHCPR guidelines found that low-stress aerobic exercise can prevent debilitation due to inactivity during the first month of symptoms and help patients with acute low back problems return to usual functioning (strength of evidence: C). The AHCPR guidelines suggest that low-stress aerobic exercise programs can be started during the first 2 weeks for most patients with acute low back problems (strength of evidence: D). The AHCPR guidelines suggest that conditioning exercises for trunk muscles are helpful for patients with acute low back problems, particularly if symptoms persist, but may aggravate symptoms more than aerobic exercise in the first 2 weeks (strength of evidence: C). The AHCPR guidelines found no evidence that back-specific exercise machines provide benefit over traditional exercise (strength of evidence: D). The AHCPR guidelines found no evidence to support stretching of the back muscles for acute low back problems (strength of evidence: D). The AHCPR guidelines suggest that gradually increasing exercise quotas result in better outcomes than telling patients to stop exercising if pain occurs (strength of evidence: C). The VA/DoD guideline recommendations for exercise are similar to the AHCPR recommendations. The UK RCGP guidelines concluded that it is doubtful that specific back exercises produce significant improvement in acute low back pain, or that it is possible to select which patients will respond to which exercises (strength of evidence: ***).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The UK RCGP guidelines found some evidence that exercise programs and physical reconditioning can improve pain and function in patients with chronic low back pain (strength of evidence: **). The UK RCGP guidelines found theoretical arguments for starting exercise programs at around 6 weeks after start of symptoms (strength of evidence: *). The European COST guidelines recommend against advising specific exercises for acute low back pain. The European COST guidelines recommend supervised exercise as a first-line treatment for chronic low back pain. They suggest exercise programs that dont require expensive training machines, the use of a cognitive-behavioral approach with graded exercises, and quotas. Group exercises are suggested as a low-cost option. The guidelines provide no recommendations on specific types of exercise, and suggest the patient and therapist could best determine that.
Hydrotherapy
For this review, we defined hydrotherapy as exercises performed in a pool or other water-based setting. In contrast to spa therapy and balneotherapy, which involve immersion in thermal mineral water, hydrotherapy generally employs normal (or chlorinated) tap water. Results of search: systematic reviews We found no systematic reviews evaluating efficacy of hydrotherapy for low back pain. Results of search: trials From 88 potentially relevant citations, we identified three lower-quality trials of hydrotherapy for chronic low back pain638-640.
Efficacy of hydrotherapy versus delayed hydrotherapy

For chronic low back pain, one lower-quality trial (n=109) found hydrotherapy superior to delayed hydrotherapy for back-specific functional status, but not for pain (Table 43)638. Incomplete and inconsistent reporting of results data makes this trial difficult to interpret.
Table 43. Trial of hydrotherapy versus delayed hydrotherapy
Author, year McIlveen, 1998638
Main results Hydrotherapy versus delayed hydrotherapy ODI, percent improved: 27% vs. 8% (p=0.05) Pain rating index of McGill Pain Questionnaire, percent improved >10 points: 11% vs. 8% (NS) Present pain intensity of McGill Pain Questionnaire, percent improved by >1 point: 33% vs. 22% (NS)
Quality score* 3/9
EVIDENCE REVIEW
Efficacy of hydrotherapy versus land-based therapy

For chronic low back pain, two lower-quality trials (n=60 and n=30) each found no differences between hydrotherapy and land-based therapy for short-term pain or functional status (Table 44)639, 640.
Table 44. Trials of hydrotherapy versus land-based therapy
Number of patients Duration of follow-up n=60 4 weeks n=30 4 weeks
Author, year Sjogren, 1997639 Yozbatiran, 2004640
Main results Hydrotherapy vs. land-based therapy Pain, mean improvement in VAS (0-10 scale): 1.35 vs. 0.79 (NS) ODI, mean improvement: 3.25 vs. 2.40 (NS) Hydrotherapy vs. land-based therapy Pain, mean improvement in VAS (0-10 scale): 3.53 vs. 2.53 (NS) ODI, mean improvement: 19.34 vs. 17.34 (NS)
Harms No trial reported adverse events. Costs We found no studies on costs.
Summary of evidence
There is no evidence on effects of hydrotherapy for acute low back pain. For chronic low back pain, there is insufficient evidence (one lower-quality trial) to judge efficacy of hydrotherapy versus delayed hydrotherapy (level of evidence: poor). For chronic low back pain, there is consistent evidence from two lower-quality trials that hydrotherapy and land-based therapy are associated with similar outcomes (level of evidence: fair).

The other guidelines do not address hydrotherapy.
Yoga
Yoga can typically be distinguished from traditional exercise by its emphasis on achieving specific body positions and movement, breathing techniques, and emphasis on mental focus. One challenge in evaluation of yoga is that many styles are practiced, each associated with different postures and techniques as well as different degrees of physical difficulty and intensity. Results of search: systematic reviews We identified no systematic reviews evaluating efficacy of yoga for low back pain.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials From 27 potentially relevant trials, we identified three trials (two higher-quality371, 641) on efficacy of yoga for chronic low back pain (Table 45)371, 641, 642.
Efficacy of yoga versus other interventions

For chronic low back pain, one higher-quality trial (n=101) found six weeks of viniyoga (a therapeutically oriented style) slightly superior to both conventional exercise and a self-care education book on the RDQ at twelve weeks (mean difference versus exercise=-1.8, 95% CI -3.5 to -0.1 and mean difference versus self-care education book=-3.4, 95% CI -5.1 to -1.6), but only superior to the self-care book at 26 weeks (mean difference=-3.6, 95% CI -5.4 to -1.8)371. Effects on symptom bothersomeness scores were similar at 12 weeks for all three interventions, though yoga was superior to the self-care book at 26 weeks (mean difference=-2.2, 95% CI -3.2 to -1.2). Yoga was also associated with decreased medication use at week 26 (21% vs. 50% vs. exercise and 21% vs. 59% vs. self-care book, p<0.05 for both comparisons), though there was no significant difference in the proportion of patients that visited health care providers for low back pain. Two smaller (n=60 and 22), trials evaluated Iyengar yoga, a commonly practiced style of Hatha yoga that makes frequent use of props. The larger trial (higher-quality) found yoga slightly more effective than exercise instruction (from a weekly newsletter) for reducing disability641. Benefits were present 3 months after the end of a 16-week course of treatment (-10.4 vs. -8.5 improvement on a 70 point disability scale, p=0.009). Differences in pain outcomes were small and only significant when adjusted for baseline differences in the intervention groups. In addition, interpretation of results is difficult because of differences in baseline disability scores (14.3 vs. 21.2) and because nearly a third of the patients did not complete the study or were lost to follow-up. The other trial (lower-quality) found no significant differences between Iyengar yoga and usual activities on measures of back-specific function or depression642. Pain outcomes were not assessed.
EVIDENCE REVIEW
Table 45. Trials of yoga versus exercise

Number of patients Duration of follow-up n=22 6 weeks n=101 26 weeks
Author, year Galantino, 2004642 Sherman, 2005371
Williams, 2005641
n=60 7 months
Main results Iyengar yoga versus usual activities Oswestry Disability Index (change from baseline): 3.83 vs. 2.18 Proportion with lower scores on Oswestry: 46% vs. 40% Viniyoga versus exercise RDQ Score (0 to 24 scale), mean difference between groups relative to baseline: -1.8 (95% CI -3.5 to -0.1) at 12 weeks (p=0.034) and -1.5 (95% CI -3.2 to 0.2) at 26 weeks (p=0.092) Viniyoga versus self-care book RDQ Score, mean difference between groups relative to baseline: -3.4 (95% CI -5.1 to -1.6) at 12 weeks (p=0.0002) and -.6 (95% CI -5.4 to -1.8) at 26 weeks (p<0.001) Iyengar yoga versus exercise education Present Pain Index, mean change at 7 months (0 to 5 scale): -0.5 vs. -0.9, p=0.140 Pain Disability Index, mean change at 7 months (7 to 70 scale): -8.5 vs. -10.4, p=0.009 Pain, VAS, mean change at 7 months (0 to 10 scale): 1.2 vs. -1.6, p=0.398
5/9
Summary of evidence
There is no evidence on efficacy of yoga for acute low back pain. For chronic low back pain, viniyoga was slightly superior to traditional exercises and moderately superior to a self-care education book for back-specific functional status and use of medications in one higher-quality trial (level of evidence: fair). There is insufficient evidence to judge effectiveness of other types of yoga (two smaller trials of Hatha yoga, one rated higher-quality, but both with significant methodological shortcomings) (level of evidence: poor).

The other guidelines do not address yoga.
Interdisciplinary interventions Interdisciplinary rehabilitation (multidisciplinary rehabilitation)

Interdisciplinary or multidisciplinary rehabilitation combines and coordinates physical, vocational, or psychological components and is provided by at least two health care professionals with different clinical backgrounds. The intensity and content of interdisciplinary
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain therapy varies widely, but most involve an exercise program and some type of psychological therapy. Results of search: systematic reviews We identified two higher-quality Cochrane reviews of interdisciplinary rehabilitation. One included trials of patients with chronic (>3 months) low back pain (ten trials, three higherquality)643, 644 and the other included trials of subacute (defined as >4 weeks and <3 months in duration) low back pain (two lower-quality trials)299, 300. No systematic review evaluated effectiveness of interdisciplinary rehabilitation for acute low back pain. We included one other systematic review on effects of multiple interventions that included five trials of interdisciplinary rehabilitation645. We excluded one outdated systematic review593. Results of search: trials Twenty unique trials of interdisciplinary rehabilitation were included in the three systematic reviews299, 300, 643-645. We also identified one recent, higher-quality trial of intensive interdisciplinary rehabilitation for high-risk patients with low back pain of less than eight weeks duration that wasnt included in the systematic reviews306. This trial is also discussed in Key Question 1c (identification and treatment of yellow flags and subsequent outcomes).
Efficacy of interdisciplinary rehabilitation versus usual care or noninterdisciplinary rehabilitation

For subacute low back pain, one of the Cochrane reviews found interdisciplinary rehabilitation (defined by the review as an intervention consisting of a physicians consultation plus a psychological, social, or vocational intervention, or a combination of these) with a workplace visit more effective than usual care, but only included two lower-quality trials (n=103 and 104)299, 300 . In one of the trials, return to work averaged 10 weeks (SD=12.7) with interdisciplinary rehabilitation (which consisted of measurement of functional capacity, a work-place visit, back school, and graded exercise with an operant-conditioning approach) versus 15 weeks (SD=15.6) with traditional care (p=0.03 for difference), and there was less sick leave in the interdisciplinary rehabilitation group in the following year (mean difference=-7.5 days, 95% CI=15.06 to 0.06)632. Subjective disability was also slightly superior in the intervention group. In the second trial, the duration of absence from work was lower with a combined occupational (occupational physician consultation and work place visit) and clinical intervention (back school, visit to back specialist, and multidisciplinary work rehabilitation including functional rehabilitation if needed) compared to the occupational or clinical interventions alone or to usual care (median days off work=60 vs. 67 vs. 131 vs. 120 days, p<0.05)646. Return to work was 2.4 times faster (95% CI 1.19 to 4.89) in the combined intervention group compared to the usual care group and 1.91 times faster (95% CI 1.18 to 3.1) with any occupational intervention compared to the two groups without the occupational intervention. The combined intervention group also was associated with greater improvements in ODI scores after one year compared to usual care (mean difference=10.7, p=0.02). For chronic low back pain, the other Cochrane review (10 trials) included three trials647-649 (one higher-quality648) that found intensive (>100 hours), daily interdisciplinary rehabilitation (defined
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain as an intervention with a physical component plus a psychological and/or social/occupational component meeting pre-defined criteria) with functional restoration moderately superior to noninterdisciplinary rehabilitation or usual care for improving short- and long-term functional status (SMD=-0.40 to -0.90 at 3-4 months and SMD=-0.56 to -1.07 at 60 months)643, 644. Two trials647, 648 (one higher-quality648) found interdisciplinary rehabilitation moderately superior for pain outcomes at 3-4 months (SMD=-0.56 and SMD=-0.74), though long-term effects were inconsistent (SMD=-0.51 and SMD=0.00 at 60 months)643, 644. There was also inconsistent evidence regarding vocational outcomes, with one higher-quality trial648 showing improvements in work-readiness but two other trials649, 650 (one higher-quality650 found no effects on sick leave. In contrast to the intensive interventions, less intensive interdisciplinary rehabilitation was not associated with improvements in pain, function, or vocational outcomes compared to non-interdisciplinary outpatient rehabilitation or usual care (five trials, two higher-quality). A smaller (five trials) systematic review reported results consistent with the Cochrane review645. For patients with low back pain for less than 8 weeks identified as being at higher risk for development of chronic disabling symptoms, one recent, small (n=70), higher-quality trial found an intensive interdisciplinary intervention (including 3 physician evaluations and up to 45 physical therapy, biofeedback/pain management, group didactic, and case manager/occupational therapy sessions) associated with improved pain, decreased disability, and decreased costs (mainly related to lost wages) compared to usual care (Table 46)306.
Table 46. Trial of intensive interdisciplinary rehabilitation in patients with low back pain for <8 weeks
Author, year Gatchel, 2003306
Main results Intensive interdisciplinary functional restoration vs. usual care Return to work at 12 months: 91% vs. 69% (p=0.027) Average number of disability days due to back pain: 38 vs. 102, p=0.001 Average self-rated pain over last 3 months: 27 vs. 43, p=0.001 Taking opioid analgesics: 27% vs. 44%, p=0.020
Quality score* 6/9
Harms No trial reported adverse events. Costs In one trial of workers disabled due to low back pain, interdisciplinary rehabilitation with physical conditioning was associated with an average cost-benefit of $18,585 after 6.4 years of followup, though the difference was not statistically significant, in part because of highly skewed distributions651. In workers with chronic low back pain, another trial found a light interdisciplinary intervention associated with an average cost-benefit of about $15,000 after 2 years relative to usual care652. For patients with acute or subacute low back pain identified as being at higher risk for developing chronic disabling symptoms, a cost-benefit analysis of a trial that compared
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain an intensive, early interdisciplinary intervention to usual care estimated a net gain of $9,122, mostly related to fewer lost wages in the interdisciplinary intervention group306.
Summary of evidence
For subacute low back pain, interdisciplinary rehabilitation (particularly with a work site visit) was associated with quicker return to work, reduced sick leave, and moderately improved disability relative to usual care in two lower-quality trials (level of evidence: fair). In higher-risk patients with acute or subacute low back pain, one higher-quality trial found interdisciplinary rehabilitation moderately more effective than usual care for pain relief, use of analgesic medications, and return to work (level of evidence: poor). For chronic low back pain, intensive interdisciplinary rehabilitation with functional restoration is moderately more effective than usual care or non-interdisciplinary rehabilitation for reducing pain and improving function, though effects on work-related outcomes are inconsistent (four trials, two higher-quality) (level of evidence good). Less intensive (<100 hours) interdisciplinary rehabilitation was not more effective than usual care or non-interdisciplinary rehabilitation (five trials) (level of evidence: good).

The European COST guidelines recommend consideration of interdisciplinary treatment programs in occupational settings for workers on sick leave for more than 4-8 weeks and interdisciplinary intervention with functional restoration in patients with chronic low back pain who have failed monodisciplinary treatment options.
Functional restoration (physical conditioning, work conditioning, or work hardening)

Functional restoration programs (variously referred to as physical conditioning, work conditioning, or work hardening programs) involve simulated or actual work tasks in a supervised environment in order to enhance job performance skills and improve strength, endurance, flexibility, and cardiovascular fitness in injured workers653. The goal of such programs is to improve functional and work outcomes. A challenge in assessing the efficacy of functional restoration is the wide variation in the content (such as the use of behavioral therapy or the type of exercise) and intensity of treatment programs. Results of search: systematic reviews We identified one higher-quality Cochrane review on efficacy of functional restoration programs for acute or chronic low back pain302, 303. Several trials evaluated in this Cochrane review were also included in Cochrane reviews of interdisciplinary rehabilitation for subacute (2 of 2 trials)299, 300 and chronic (3 of 10 trials)643, 644 low back pain. Results of search: trials Eighteen trials of functional restoration (9 rated higher-quality) were included in the Cochrane review302, 303. We identified one additional higher-quality trial of an intensive interdisciplinary
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain functional restoration intervention in patients with low back pain for less than eight weeks (see discussion in interdisciplinary rehabilitation section)306.
Efficacy of functional restoration versus usual care

For acute low back pain, the Cochrane review302, 303 included four trials, three of which found functional restoration no better than usual care, normal activities, or standard exercise therapy349, 654, 655 (two trials rated higher-quality349, 654). In the only trial that found a beneficial effect (rated higher-quality656), functional restoration was compared to an intervention consisting of lying prone and using ice packs. A recent trial not included in the Cochrane review found intensive, interdisciplinary functional restoration superior to usual care for several outcomes in high-risk patients with low back pain for less than eight weeks (see section on interdisciplinary rehabilitation)306. For chronic low back pain (14 trials), functional restoration programs with a cognitive-behavioral approach generally appeared effective for reducing time off work. In two relatively homogeneous trials632, 646 (one higher-quality646) of functional restoration versus usual care, the decrease in number of sick days after 12 months follow-up averaged 45 days (95% CI 3 to 88). There was little evidence for or against effectiveness of functional restoration not accompanied by a cognitive behavioral approach.
Efficacy of functional restoration versus other interventions

For chronic low back pain, the Cochrane review302, 303 included two higher-quality trials that found functional restoration associated with an average of 112 and 243 fewer lost work days compared to traditional physical therapy (about two-thirds of patients received physical modalities and one-third manipulation)657 or traditional exercise therapy plus behavioral therapy658. Harms No trial reported adverse events. Costs See section on interdisciplinary rehabilitation.
Summary of evidence
For acute low back pain, evidence from six heterogeneous trials on efficacy of functional restoration is inconsistent, with the majority of studies showing no benefit (level of evidence: fair). For chronic low back pain, functional restoration with a cognitive-behavioral approach was moderately effective for reducing time off work (14 trials) (one higher quality) (level of evidence: fair). For chronic low back pain, functional restoration was more effective than traditional physical therapy (including physical modalities or manipulation) and traditional exercise therapy plus
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain behavioral therapy for reducing days lost from work (two higher-quality trials) (level of evidence: good). There is insufficient evidence to evaluate benefits of functional restoration without a cognitivebehavioral approach.

The European COST guidelines recommend consideration of interdisciplinary treatment programs in occupational settings for workers on sick leave for more than 4-8 weeks and interdisciplinary intervention with functional restoration in patients with chronic low back pain who have failed monodisciplinary treatment options.
Physical modalities Interferential therapy

Interferential therapy involves the application of a medium frequency alternating current modulated to produce low frequencies up to 150 Hz. It is thought to provide pain relief in part by increasing blood flow to tissues, and is considered more comfortable for patients than transcutaneous electrical nerve stimulation. Results of search: systematic reviews We found no systematic reviews of interferential therapy. Results of search: trials From eight potentially relevant citations, we identified three trials (one higher-quality659) of interferential therapy659-661. Two trials evaluated patients with subacute (>4 weeks) back pain and the other evaluated patients with back pain of mixed duration (mainly chronic). Interferential therapy was compared to spinal manipulation, traction, and a back self-care book in one trial each.
Efficacy of interferential therapy versus spinal manipulation or traction

For subacute (>4 weeks) low back pain, one higher-quality trial (n=240) found no difference between an 8-week course of interferential therapy and spinal manipulation on pain, functional disability, quality of life, work status, or other outcomes after 6 to 12 months (Table 47)659. For back pain of unspecified duration (primarily >5 years), a lower-quality trial (n=152) also found no differences between interferential therapy and traction on pain or the ODI after 3 months660.
EVIDENCE REVIEW
Table 47. Trials of interferential therapy versus other interventions

Author, year Hurley, 2004659
Werners, 1999660
n=152 3 months
Main results Interferential therapy versus manipulative therapy versus combination (mean improvement at 12 months) Pain (0 to 100 VAS): -26.5 vs. -18.2 vs. -25.7 (NS) McGill Pain Questionnaire Pain Rating Index (0 to 78): -8.3 vs. -6.4 vs. -9.2 (NS) RDQ score (0 to 24): -4.9 vs. -4.7 vs. -6.5 (NS) SF-36: No differences Recurrent low back pain: 69% vs. 77% vs. 64% (NS) Absent from work >30 days: 8% vs. 12% vs. 12% Interferential therapy versus traction (mean difference from baseline to 3 months) Pain (0 to 100): -9.8 vs. -14.6 (NS) Oswestry (0 to 100): -7.7 vs. -7.4 (NS)
Quality score* 7/10
4/10
*Excludes criterion involving blinding of care providers, for maximum score of 10
Efficacy of interferential therapy plus a back self-care book versus a back selfcare book alone
For subacute low back pain (>4 weeks), one small (n=60), higher-quality trial found interferential therapy applied to the paraspinal area (near the target spinal nerve) plus a back self-care book superior to the back self-care book alone on the RDQ after 3 months, but not on the Pain Rating Index or EQ-5D (Table 48)661. Interpretation of effects on functional status are difficult because baseline RDQ scores were higher in the interferential therapy group (median 9.0 vs. 5.0), and median RDQ scores were identical at 3 months in the two groups (1.0 vs. 1.0). This trial also found no differences between interferential therapy applied to the painful area plus a self-care book versus the self-care book alone.
Table 48. Trial of interferential therapy + self-care book versus a self-care book alone
Main results Interferential therapy applied to painful area + selfcare book versus interferential therapy applied to area of spinal nerve + self-care book versus self-care book alone (difference in median scores from baseline to 3 months) McGill Pain Questionnaire Pain Rating Index (0 to 78): +2.2 vs. -2.5 vs. -9.7 RDQ Score (0 to 24): -3.5 vs. -8.0 vs. -4.0 EQ-5D: No difference RDQ Score, median score at 3 months: 2.0 vs. 1.0 vs. 1.0
Quality score* 5/9
Harms One trial reported no adverse events with interferential therapy or manipulation659. The other two trials reported no information on adverse events.
EVIDENCE REVIEW
Summary of evidence
For subacute low back pain, one higher-quality trial found no difference between interferential therapy and spinal manipulation (level of evidence: fair). For subacute low back pain, one higher-quality trial found interferential therapy plus a selfcare book superior to the self-care book alone, but differences could be due to baseline differences between groups (level of evidence: fair). For primarily chronic low back pain, one lower-quality trial found no differences between interferential therapy and traction (level of evidence: poor).

The AHCPR guidelines found physical agents and modalities (including electrical stimulation) of insufficiently proven benefit to justify their cost in patients with acute low back pain (strength of evidence: C). The VA/DoD and UK RCGP guidelines reached similar conclusions. The European COST guideline made no recommendation for interferential therapy in acute low back pain, and found insufficient evidence to recommend interferential therapy for chronic low back pain.
Low-level laser therapy

Low-level laser therapy involves application of laser at wavelengths varying from 632 to 904 nm to the skin in order to apply electromagnetic energy to soft tissues. Optimal treatment parameters (wavelength, dosage, dose-intensity, type of laser) are uncertain. Results of search: systematic reviews We identified no systematic reviews on efficacy of low-level laser therapy specifically for low back pain. We excluded five systematic reviews on low-level laser therapy for various musculoskeletal conditions623, 662-665. Results of search: trials We identified seven trials (four higher-quality666-669) of low-level laser therapy for low back pain666-672. Four trials evaluated patients with chronic low back pain, one evaluated patients with acute low back pain, and two did not specify duration of back pain symptoms. Although lowlevel laser therapy is frequently used in Russia and Asia, we found no non-English language trials. However, studies in Russian and Asian languages are frequently not indexed in Englishlanguage electronic databases.
Efficacy of low-level laser therapy versus sham therapy or placebo

For chronic low back pain or low back pain of unspecified duration, results of six trials of lowlevel laser therapy are difficult to interpret because they evaluated heterogeneous outcome
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain measures and different types of lasers at varying doses. Two666, 668 of the three667 higher-quality trials found laser therapy slightly superior to placebo or sham laser at the end of treatment for back-specific function (about 4 point difference on the ODI score)666 and moderately superior for the proportion of patients with >60% pain relief (71% vs. 36%, p<0.007)668 (Table 49). In one trial, benefits persisted for one month following treatment666, and in the other, relapse of back pain was less likely 6 months following the end of treatment668. One other higher-quality trial found laser more effective than sham, but used a poorly described and unvalidated outcome measure669. One lower-quality trial of patients with back pain of unspecified duration reported similar findings, with decreased relapse through one year following treatment671. In the one higher quality trial that found no difference between laser and sham laser, each group also received a standardized home exercise regimen667.
Table 49. Trials of low-level laser therapy versus sham laser
Number of patients Duration of follow-up n=61 1 month after end of treatment Klein, 1990667 n=20 1 month after treatment n=120 1 year after treatment n=85 6 months after end of treatment n=41 1 day after treatment
Author, year Basford, 1999666
Main results Nd:YAG laser versus sham (mean change from baseline) ODI score: -6.3 vs. -2.1 Maximal pain in the last 24 hours (0 to 100 VAS): -16.1 vs. -2.3 GaAS laser + exercise versus sham laser + exercise (mean change from baseline) Pain (0 to 7.5 VAS): -1.3 vs. -1.2 RDQ Disability score: -1.8 vs. -3.0 904 nm laser vs. 10600 nm laser vs. sham Complete disappearance of pain 1 month after treatment: 95% vs. 82.5% vs. 2.5% Relapse 1 year after treatment: 65% vs. 70% vs. 95% GaAS laser versus sham Proportion with >60% pain relief at end of treatment: 71% (27/38) vs. 36% (12/33), p<0.007 GaAS laser versus sham Treatment effective: 94% (15/16) vs. 48% (12/25)
Quality score 8/11
6/11
Longo, 1988671
5/11
Soriano, 1998668
6/11
Toya, 1994669
10/11
One systematic review found low level laser effective for a variety of musculoskeletal conditions when the subgroup of trials that evaluated higher laser doses were analyzed663. The criteria for adequate doses were defined for various locations in an a priori matter. There were too few trials (four) to assess effects of dose in patients specifically with low back pain.
Efficacy of low-level laser therapy versus other interventions

For acute low back pain, one trial of low-level laser therapy was uninterruptable because of poor methodologic quality, unclear reporting of outcomes, and comparison to mesotherapy (an unproven technique involving injections of various substances into fat) (Table 50)672. For chronic low back pain, another lower-quality trial found no differences between laser, exercise, and the combination of laser plus exercise for pain and back-specific functional status670.
EVIDENCE REVIEW
Table 50. Trials of low-level laser therapy versus other interventions

Number of patients Duration of follow-up n=75 1 month after treatment Monticone, 2004672 n=22 Up to 12 months after treatment
Author, year Gur, 2003670
Main results Laser versus exercise versus laser + exercise (mean change from baseline) Pain (0 to 10 VAS): -4.2 vs. -3.6 vs. -4.4 (p>0.05) RDQ Score: -9.7 vs. -9.6 vs. -11.5 (p>0.05) Modified ODI: -16.4 vs. -16.9 vs. -17.6 (p>0.05) Laser versus stabilization (exercise, lumbar therapy, and mesotherapy) Pain at rest (VAS 0 to 10), mean change from baseline and 12 months following end of treatment: 0 vs. -5; -1 vs. -6 Pain with movement (VAS 0 to 10), mean change from baseline and 12 months following end of treatment: -4 vs. -7, -2 vs. -8
Quality score 3/11
1/11
Harms In a systematic review of low-level laser therapy for various musculoskeletal conditions, six of the 11 trials evaluating higher dose regimens reported no adverse events663. One other trial reported one transient adverse event in both laser and sham groups666. Costs We found no studies evaluating costs.
Summary of evidence
For acute low back pain, there is no reliable evidence (one lower-quality trial) on efficacy of low-level laser therapy (level of evidence: poor). For chronic low back pain, there is conflicting evidence from five trials (four higher-quality) on efficacy of low-level laser compared to placebo or sham laser. Four trials (three higherquality) found laser therapy superior to sham for pain or functional status up to one year following treatment (estimates of effects ranged from small to large), but one higher-quality trial found no difference between laser and sham in patients also receiving exercise. In addition, interpretation of results is compromised by the use of heterogeneous and nonstandardized outcome measures in some studies (level of evidence: fair). For chronic low back pain, there was no difference between low-level laser therapy, exercise, or the combination of laser plus exercise in one lower-quality trial (level of evidence: poor). Additional research is needed on optimal doses of low-level laser therapy, number of sessions, and type of laser. Publication bias from non-English language studies could affect these conclusions.

The AHCPR guidelines found physical agents and modalities (including low-level laser) of insufficiently proven benefit to justify their cost for acute low back pain (strength of evidence: C).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The VA/DoD guidelines reached similar conclusions. The UK RCGP guidelines dont address low-level laser therapy. The European COST guidelines found insufficient evidence to recommend low-level laser for chronic low back pain.
Shortwave diathermy
Shortwave diathermy involves application of shortwave electromagnetic radiation with a frequency range from 10 to 100 MHz in order to elevate the temperature of deep tissues. Results of search: systematic reviews We identified no systematic review of shortwave diathermy. Results of search: trials From 14 potentially relevant citations, we identified three lower-quality trials of shortwave diathermy673-675.
Efficacy of shortwave diathermy versus sham diathermy

For low back pain of at least two months duration, one lower-quality trial found no significant differences between two weeks of short-wave diathermy and sham diathermy in median pain scores and the proportion of patients free of pain through 12 weeks, following a two-week course of therapy673. For back pain present for longer than one week (widely varying durations), another lower-quality trial found no differences in global response (other outcomes not reported) between short-wave diathermy and sham diathermy after 2 weeks (Table 51)675.
EVIDENCE REVIEW
Table 51. Trials of shortwave diathermy for low back pain

Author, year Duration of low back pain Gibson, 1985673 Low back pain >2 months
Rasmussen, 1979674 Low back pain <3 weeks Sweetman, 1993675 Low back pain >1 week
n=24 2 weeks n=400 2 weeks
Main results Shortwave diathermy vs. osteopathic manipulation vs. detuned (sham) diathermy Median daytime pain score (0 to 100) at 2 weeks: 35 vs. 25 vs. 28 Median daytime pain score (0 to 100) at 12 weeks: 25 vs. 13 vs. 6 Proportion free of pain at 2 weeks: 35% vs. 25% vs. 28% Proportion free of pain at 12 weeks: 37% vs. 42% vs. 44% Proportion needing analgesics at 2 weeks: 22% vs. 18% vs. 32% Proportion needing analgesics at 12 weeks: 7% vs. 18% vs. 22% Proportion unable to work or with modified activities at 2 weeks: 31% vs. 13% vs. 38% Proportion unable to work or with modified activities at 12 weeks: 7% vs. 5% vs. 19% Shortwave diathermy vs. spinal manipulation Proportion 'fully restored" by 14 days: 25% (3/12) vs. 92% (11/12) Shortwave diathermy versus extension exercises versus traction versus sham diathermy Global effect "better" at 2 weeks: 39% (39/100) vs. 45% (45/100) vs. 49% (49/100) vs. 37% (37/100) (NS)
Quality score 4/11
3/11 5/11
Efficacy of shortwave diathermy versus other interventions

For low back pain present for at least two months, one lower-quality trial found no significant differences between shortwave diathermy and spinal manipulation on median pain scores, proportion free of pain, or requirement for analgesics through 12 weeks, following a two-week course of treatment673. For acute low back pain, a second lower-quality trial found a lower rate of symptom resolution two weeks following a course of shortwave diathermy (3 of 12) compared to a course of spinal manipulation (11 of 12)674. However, no details about the shortwave diathermy intervention were provided. For low back pain present longer than one week, a lower-quality trial found no difference between shortwave diathermy and either extension exercises or traction after two weeks675. Harms No trial reported adverse events. Cost We found no studies evaluating costs.
Summary of evidence
For acute low back pain, one small, lower-quality trial found shortwave diathermy inferior to spinal manipulation for the proportion of patients reporting resolution of symptoms after 2 weeks (level of evidence: poor).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For subacute or chronic low back pain, one lower-quality trial found no difference between shortwave diathermy and sham diathermy in pain relief through 12 weeks (level of evidence: poor). For subacute or chronic low back pain, one lower-quality trial found no difference between shortwave diathermy and osteopathic spinal manipulation in pain relief through 12 weeks (level of evidence: poor). For back pain of varying duration, one lower-quality trial found no difference between shortwave diathermy, sham diathermy, exercise, or traction using an unvalidated measure of global effect after 2 weeks (level of evidence: poor).

The AHCPR guidelines found physical agents and modalities (including shortwave diathermy) of insufficiently proven benefit to justify their cost in patients with acute low back pain (strength of evidence: C). The VA/DoD and UK RCGP guidelines reached similar conclusions. The European COST guidelines found insufficient evidence to recommend shortwave diathermy for chronic low back pain.
Traction
Traction involves drawing or pulling of the body in order to stretch the lumbar spine. A variety of methods are used and usually involve a harness around the lower rib cage and around the iliac crest, with the pulling motion performed using free weights and a pulley, motorized equipment, inversion techniques, or an overhead harness. Results of search: systematic reviews We identified a higher-quality Cochrane review (23 RCTs, 5 rated high-quality) of traction for low back pain676, 677. All included trials enrolled patients with low back pain and sciatica, though seven also included patients without sciatica. We included three other higher-quality systematic reviews that each included between 8 and 14 trials of traction100, 399, 678. We excluded three older systematic reviews193, 623, 679. Results of search: trials Twenty-four unique trials of traction were included in four systematic reviews100, 399, 676-678. Sixteen trials only included patients with sciatica. The remaining trials evaluated mixed populations of patients with and without sciatica. We did not search for additional trials.
Efficacy of traction versus placebo, sham, or no treatment

For low back pain of varying duration (with or without sciatica), the Cochrane review676, 677 included two higher-quality trials680-682 that found traction no more effective than placebo, sham, or no treatment for pain, functional status, overall improvement, or work absenteeism.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For low back pain specifically with sciatica (varying duration), the Cochrane review676, 677 included two lower-quality trials683, 684 that found autotraction more effective than placebo, sham, or no treatment for pain, global improvement, or work absenteeism, but other forms of traction (continuous or intermittent traction) were not associated with beneficial effects in eight other trials (one higher-quality685). Three other systematic reviews did not include any trials not in the Cochrane review and found either no evidence that traction is effective for low back pain with or without sciatica100, 399, or insufficient evidence to draw reliable conclusions678.
Efficacy of traction versus other interventions

For sciatica of varying duration, six RCTs (five rated lower-quality) included in the Cochrane review compared various types of traction to other non-invasive interventions. In the lone higher-quality trial, autotraction was superior to abdominal and pelvic floor muscle isometric exercises at the end of treatment686. However, benefits were no longer present after one month. In a lower-quality trial, intermittent traction was superior to physiotherapy for global well-being after three to five weeks, though no better than superficial application of hot packs687. In the other four lower-quality trials, no statistically significant differences were seen between traction and spinal manipulation and a corset389, an infra-red lamp151, 688, exercise and shortwave diathermy675, or strengthening and range of motion exercises689. For chronic low back pain with sciatica, traction was no more effective than isometric exercise in two trials689, 690, and superior to TENS in the third691 (none rated higher- quality). For low back pain of varying duration without sciatica, one higher-quality trial found no differences between intermittent traction and interferential treatment in pain or function three months after treatment660.
Efficacy of one type of traction versus another

For chronic low back pain with or without sciatica, one small (n=44) trial found autotraction more effective than mechanical traction for global improvement (but not pain or function)692. In two other small trials, there were no differences between static and intermittent traction693 or between autotraction and manual traction694. One trial found no differences between intermittent or continuous traction using different levels of force695. Harms Adverse events were generally reported inconsistently and poorly in the 23 trials included in the Cochrane review676, 677. Two trials reported no adverse events685, 696. Six other trials reported adverse events including increased pain, increased rate of subsequent surgery, aggravation of neurological signs, aggravation of symptoms676, 677. The other sixteen trials did not mention adverse events. Costs We found no studies evaluating costs.
EVIDENCE REVIEW
Summary of evidence
For low back pain of varying duration (with or without sciatica), there is consistent evidence from two higher-quality trials that continuous traction is not associated with superior outcomes compared to placebo, sham, or other treatments (level of evidence: good). For low back pain of varying duration with sciatica, eight trials (one higher-quality) consistently found no differences between continuous or intermittent traction and placebo, sham, or other treatments (level of evidence: good). For low back pain of varying duration with sciatica, two lower-quality trials found autotraction superior to placebo or sham therapies and one lower-quality trial found autotraction superior to mechanical traction (level of evidence: fair). For chronic low back pain with sciatica, traction was no better than isometric exercises in two lower-quality trials and inferior to TENS in a third lower-quality trial (level of evidence: poor). Adverse events associated with traction may include aggravation of signs and symptoms or subsequent surgery, but were inconsistently and poorly reported in the trials (level of evidence: poor).

The AHCPR guidelines recommend against traction for treatment of patients with acute low back problems (strength of evidence: B). The VA/DoD and UK RCGP also recommend against traction, but rate the strength of evidence differently (strength of evidence: C and ***, respectively). The European COST guidelines recommend against traction for acute low back pain and found insufficient evidence to recommend traction for chronic low back pain.
Transcutaneous electrical nerve stimulation (TENS)

Transcutaneous electrical nerve stimulation involves the use of a small battery-operated device to provide continuous electrical impulses via surface electrodes, with a goal of providing symptomatic relief by modifying pain perception. Result of search: systematic reviews We identified one recent, higher-quality Cochrane review on efficacy of TENS versus sham TENS (two trials, one rated higher-quality697) for low back pain698, 699. In addition, higher-quality systematic reviews of acupuncture68, massage700, 701, spinal manipulation66, 67; traction676, 677, and superficial heat or cold398 each included one to four trials comparing the target intervention to TENS. We excluded three outdated Cochrane reviews702-704 and three other outdated systematic reviews193, 623, 705. Results of search: trials Eleven unique trials of TENS were included in the systematic reviews68, 398, 676, 698, 699, 701, 706. We did not search for additional trials.
EVIDENCE REVIEW
Efficacy of TENS versus sham TENS

For chronic low back pain, the Cochrane review included one higher-quality trial (n=145)697 that found no differences between TENS and sham TENS for any measured outcome (including pain and functional status) after 4 weeks698, 699. A smaller (n=30), lower-quality trial found active TENS associated with greater reduction in pain over the 60-minute treatment session compared to sham TENS (WMD=-33.62, 95% CI -52.27 to -13.97)707. Longer-term results and adverse events were not reported.
Efficacy of TENS versus other interventions

For chronic low back pain, a systematic review of acupuncture included five trials (none higherquality) that found no differences between acupuncture and TENS for short- (four trials pooled, SMD=0.15, 95% CI -0.33 to 0.63) or long-term pain (two trials, SMD=0.32, 95% CI -0.33 to 0.96)68. Results of studies that compared TENS to other interventions for chronic low back pain are mixed: one lower-quality trial found TENS inferior to traction691, one higher-quality trial found TENS superior to minimal massage708, and one lower-quality trial found no differences between TENS and gentle ice massage405. For acute low back pain, a systematic review of acupuncture included one lower-quality trial709 that found TENS inferior to acupuncture for pain relief68. For subacute low back pain, a systematic review of spinal manipulation66, 67 included one higher-quality trial390, 391 that found TENS moderately inferior to spinal manipulation for pain (SMD 0.5, 95% CI 0.1 to 1.0) and substantially inferior for disability (SMD 1.3, 95% CI 0.5 to 2.0), though there were no differences between TENS and gentle massage. Harms In trials of TENS, one third of patients with either active or sham TENS had minor skin irritation, with one patient (sham) discontinuing due to severe dermatitis698, 699. The proportion of patients with skin irritation was similar in patients who received active or sham TENS. Costs We found no studies evaluating costs.
Summary of evidence
For acute low back pain, TENS was inferior to acupuncture in one lower-quality trial (level of evidence: poor) For subacute low back pain, TENS was inferior to spinal manipulation in one higher-quality trial (level of evidence: fair). For chronic low back pain, the only higher-quality trial found no differences between TENS and sham TENS (level of evidence: fair). For chronic low back pain, five lower-quality trials found consistent evidence of no differences between TENS and acupuncture (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For chronic low back pain, evidence on efficacy of TENS compared to other interventions is limited to single trials of traction (traction superior), minimal massage (TENS superior), and gentle ice massage (no differences) (level of evidence for each comparison: poor). TENS is associated with skin irritation that is usually minor (level of evidence: fair).

The AHCPR guidelines recommend against TENS in patients with acute low back problems (strength of evidence: C). The VA/DoD guidelines are identical. The UK RCGP found inconclusive evidence on the efficacy of TENS in patients with acute low back problems (strength of evidence: **). The European COST guidelines recommend against TENS for chronic low back pain.
Percutaneous electrical nerve stimulation (PENS)

Percutaneous electrical nerve stimulation (PENS) involves the insertion of acupuncture-like needles and applying low-level electrical stimulation. It differs from electroacupuncture in that the insertion points target dermatomal levels for local pathology, rather than acupuncture points. However, there is some uncertainty over whether PENS should be considered a novel therapy or a form of electroacupuncture710. Results of search: systematic review We identified no systematic reviews of PENS. Results of search: trials We identified three trials of PENS for chronic low back pain711-713 and one trial of PENS for sciatica714. All were rated lower quality.
Efficacy of PENS versus sham PENS

For chronic low back pain, two trials compared PENS to sham PENS (Table 52)712, 713. Both found PENS moderately superior to sham PENS for pain outcomes, either at the end of treatment712 or three months after a course of treatment713. One trial also found moderate to substantial improvements in functional outcomes and quality of sleep at the end of treatment712. The other trial found no benefits on measures of depression or functional status three months after treatment713. In both trials, success of blinding was not assessed. For sciatica of at least six weeks duration, a third trial found PENS moderately to substantially superior to sham PENS immediately after a two-week course of treatment for pain, functional status, and measures of sleep quality714.
EVIDENCE REVIEW
Table 52. Trials of PENS versus sham PENS

Number of patients Duration of follow-up n=60 At end of 2-week course of treatment n=64 At end of 2-week course of treatment n=34 3 months after treatment
Author, year Ghoname, 1999712 (non-sciatic low back pain) Ghoname, 1999714 (sciatica) Weiner, 2003713 (non-sciatic low back pain)
Main results PENS vs. sham PENS (mean improvement from baseline) Pain (VAS 0 to 10): -2.9 vs. -0.2 (p<0.02 for PENS) Level of activity (0 to 10): -2.3 vs. -0.2 (p<0.02 for PENS) Quality of sleep (0 to 10): -2.4 vs. 0 (p<0.02 for PENS) PENS vs. sham PENS Pain (VAS 0 to 10): -3.1 vs. -0.5 (p<0.01) Level of activity (0 to 10): -2.4 vs. -0.5 (p<0.01) Quality of sleep (0 to 10): -2.4 vs. -0.3 (p<0.01) PENS + physical therapy versus sham PENS + physical therapy (mean scores 3 months after treatment) McGill Pain Questionnaire: 6.19 vs. 11.82 (p=0.04) Multidimensional Pain Inventory Pain Inventory score: 2.16 vs. 3.10 (p=0.003) RDQ scale: 9.25 vs. 12.18 (p=0.26)
Quality score 2/11
1/11
4/11
Efficacy of PENS versus other interventions

For chronic low back pain, two trials compared PENS to TENS711, 712 (Table 53). Both found PENS moderately superior to TENS at the end of treatment for measures of pain and functional status, but the only trial that followed patients after the end of treatment found that benefits were no longer present after 1 to 2 months711. One of these trials also compared PENS to a minimal exercise intervention (seated flexion and extension)712. PENS was substantially superior to exercise on measures of pain and functional status at the end of a two-week course of treatment. For patients with sciatica, one lower-quality trial found PENS slightly superior to TENS on measures of pain and moderately superior for functional status at the end of a two-week course of treatment714.
EVIDENCE REVIEW
Table 53. Trials of PENS versus other interventions

Number of patients Duration of follow-up n=60 At end of 2-week course of treatment
Author, year Ghoname, 1999712 (non-sciatic low back pain)
Ghoname, 1999714 (sciatica) Yokoyama, 2004711 (low back pain, presence or absence of sciatica not specified)
n=64 At end of 2-week course of treatment n=60 2 months after treatment
Main results PENS vs. TENS vs. exercise, mean improvement from baseline Pain (VAS 0 to 10): -2.9 vs. -0.6 vs. -0.1 (p<0.02 for PENS vs. other interventions) Level of activity (0 to 10): -2.3 vs. -0.8 vs. 0 (p<0.02 for PENS vs. other interventions) Quality of sleep (0 to 10): -2.4 vs. -0.3 vs. -0.3 (p<0.02 for PENS vs. other interventions) PENS vs. TENS, mean improvement from baseline Pain (VAS 0 to 10): -3.1 vs. -2.6 (p<0.01) Level of activity (0 to 10): -2.4 vs. -1.3 (p<0.01) Quality of sleep (0 to 10): -2.4 vs. -1.0 (p<0.01) PENS vs. TENS Pain (VAS pain scores): 32 vs. 48 at end of treatment (p<0.01), no differences 2 months after treatment Physical impairment (0 to 4 scale): difference between PENS and TENS significant at end of treatment but not 1 month after treatment (data not reported)
Quality score 2/11
1/11
3/11
Summary of evidence
There is no evidence on efficacy of PENS for acute low back pain. For chronic low back pain, PENS was moderately superior to sham PENS for short-term pain outcomes in two lower-quality trials. In the only trial that assessed longer-term (not immediately after a course of treatment) outcomes, benefits on pain were present through two months, but there was no effect on functional outcomes (level of evidence: fair). For chronic low back pain, PENS was moderately superior to TENS and a minimal exercise intervention for pain and functional outcomes in one lower-quality trial immediately after a course of treatment, but in the only trial that evaluated longer-term outcomes, no benefits were present after two months (level of evidence: poor). For sciatica, PENS was moderately to substantially superior to sham PENS and slightly to moderately superior to TENS for pain and functional outcomes in one lower-quality trial, but outcomes were only assessed immediately after a two-week course of treatment (level of evidence: poor). There is insufficient evidence to accurately judge safety of PENS.
EVIDENCE REVIEW

The European COST guidelines recommend considering PENS for patients with chronic nonspecific low back pain.
Ultrasound
Ultrasound involves the therapeutic application of high-frequency sound waves up to 3 MHz to the body surface. Results of search: systematic reviews We identified four systematic reviews of ultrasound therapy for patients with a variety of musculoskeletal conditions, but none specifically evaluated efficacy of ultrasound for low back pain623, 715-717. Results of search: trials From 265 potentially relevant citations, we identified three small (n=15 to 73) lower-quality trials of therapeutic ultrasound for low back pain718-720.
Efficacy of ultrasound versus sham or placebo

For acute low back pain with prolapsed lumbar intervertebral disc and sciatica, one nonrandomized trial found ultrasound superior to sham ultrasound or analgesics for the proportion of pain-free patients (41% vs. 12% vs. 6.8%)719. Patients in all groups were also prescribed bed rest. For chronic low back pain, one small (n=15) trial found ultrasound moderately superior to sham ultrasound for functional status after ten treatment sessions, but had a number of methodological shortcomings, including high loss to follow-up (one-third of enrollees) and lack of intention-to-treat analysis718. For low back pain of unspecified duration, a second small (n=36) randomized trial found no difference between ultrasound and sham ultrasound in pain improvement after one month of therapy720. Functional status and other outcomes were not reported. Three systematic reviews found little evidence of beneficial effects with ultrasound relative to placebo for other musculoskeletal conditions, with the possible exceptions of single trials of lateral epicondylitis, carpal tunnel syndrome, and calcific tendonitis of the shoulder623, 715, 717 Harms Adverse events were not reported in the two studies. None of the systematic reviews of therapeutic ultrasound for various musculoskeletal conditions assessed adverse events. There is one report of two patients with a herniated disc who had transiently increased radicular pain after application of therapeutic ultrasound721. Costs We found no studies evaluating costs.
EVIDENCE REVIEW
Summary of evidence
There is insufficient evidence (one non-randomized trial) to judge benefits or harms of ultrasound for low back pain with sciatica (level of evidence: poor). There is insufficient evidence (two lower-quality, small randomized trials with inconsistent results) to judge benefits or harms of ultrasound for chronic low back pain or back pain of unspecified duration (level of evidence: poor).

The AHCPR guidelines found physical agents and modalities (including ultrasound) of insufficiently proven benefit to justify their cost in acute low back pain (strength of evidence: C). The VA/DoD and UK RCGP guidelines reached similar conclusions. The European COST guidelines found insufficient evidence to recommend ultrasound therapy for chronic low back pain.
Other non-invasive interventions Psychological therapies

Psychological therapies include standard cognitive-behavioral or operant therapy as well as other interventions such as biofeedback (use of auditory and visual signals reflecting muscle tension or activity to train patients to inhibit or reduce the muscle activity), progressive relaxation (deliberate tensing and relaxation of muscles to facilitate recognition and release of muscle tension) and self-regulatory therapy (biofeedback, relaxation training or hypnosis). Results of search: systematic reviews We identified two higher-quality systematic reviews on efficacy of psychological therapies for chronic low back pain301, 722. One was a Cochrane review (21 trials, 7 higher-quality) that only included trials evaluating psychological therapies as a separate treatment301. The other systematic review (22 trials, 6 higher-quality) also included trials of psychological therapies as part of interdisciplinary interventions722. We excluded an outdated Cochrane review723 and five other outdated systematic reviews193, 346, 386, 597, 724. Results of search: trials Thirty-five unique trials of psychological therapies for chronic low back pain were included in the systematic reviews301, 722. We did not search for additional trials.
Efficacy of psychological therapies versus wait list control

For chronic low back pain, the Cochrane review301 included four trials (one higher-quality725) that found combined cognitive-behavioral therapy moderately superior to wait list control for shortterm pain intensity (SMD=0.59, 95% CI 0.10 to 1.09), but not for functional status (SMD=0.31, 95% CI -0.20 to 0.82). It also included two lower-quality trials that found progressive relaxation associated with large effects on short-term pain (SMD=1.16, 95% CI 0.47 to 1.85) and behavioral outcomes (SMD=1.31, 95% CI 0.61 to 2.01). Evidence regarding effects of
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain electromyography (EMG) biofeedback versus wait list control was mixed from four trials (one higher-quality726). Although three trials (one higher-quality) found a moderate positive effect on pain intensity (SMD=0.84, 95% CI 0.32 to 1.35), a fourth trial found no differences. In addition, there were no differences between EMG biofeedback and wait list control for behavioral outcomes. Three trials (one higher-quality725) of operant treatment versus wait list controls found inconsistent effects on pain intensity and no benefits for general functional status or behavioral outcomes. The second systematic review (22 trials) also found cognitive-behavioral and self-regulatory treatments (such as relaxation therapy) moderately superior to wait list control for pain intensity (SMD=0.62, 95% CI 0.25 to 0.98 and SMD=0.75, 95% CI 0.35 to 1.15, respectively)722. Selfregulatory therapy was also moderately superior to wait list controls for measures of depression (SMD=0.81, 95% CI 0.11 to 1.52).
Efficacy of psychological therapies versus other active interventions

For workers with subacute low back pain, the Cochrane review301 included one higher-quality trial632 that found operant treatment in combination with a graded activity program associated with earlier return to work and reduced long-term sick leave compared to usual care632. For chronic low back pain, one lower-quality trial included in the Cochrane review301 found no difference between behavioral therapy and exercise on pain intensity, functional status, and behavioral outcomes through 12 months727. The other systematic review found no differences between psychological therapies (either alone or as part of multidisciplinary treatment) and other active interventions (including physical therapy interventions and usual care) for pain intensity, pain interference, health care visits, or medication use722. However, psychological therapies were slightly to moderately superior to other interventions for short- and long-term disability (3 trials, SMD=0.36, 95% CI 0.06 to 0.65 and 4 trials, SMD=0.53, 95% CI 0.19 to 0.86, respectively).
Efficacy of one psychological therapy intervention versus another

In head-to-head comparisons, neither systematic review found clear differences between different types of psychological therapies301, 722. In the Cochrane review301, the best-studied comparisons were cognitive-behavioral versus operant therapy (three higher-quality trials725, 728, 729 ) and cognitive versus respondent therapy (three lower-quality trials730-732). Harms Safety was not assessed in any of the systematic reviews. Costs One trial that compared different operant interventions found no significant differences in costs or utilities733.
EVIDENCE REVIEW
Summary of evidence
For chronic low back pain, there is consistent evidence from four trials (one higher-quality) that cognitive-behavioral therapy is moderately more effective than wait list control for shortterm pain intensity, though there were no significant differences in functional status and other outcomes (level of evidence: good). For chronic low back pain, two lower-quality trials found progressive relaxation associated with large beneficial effects on pain intensity and behavioral outcomes compared to wait list control (level of evidence: fair). For chronic low back pain, evidence on EMG biofeedback versus wait list control is mixed, though moderate benefits on pain intensity were reported in three out of four trials (one higher-quality) (level of evidence: fair). For chronic low back pain, operant therapy was not associated with any clear benefits relative to wait list controls in three trials (one higher-quality) (level of evidence: good). For chronic low back pain, psychological therapies have not clearly been shown to be superior to other non-invasive interventions for most outcomes, though one systematic review found psychological therapies associated with moderate beneficial effects on short- and long-term disability (level of evidence: fair). There is no clear evidence from head-to-head trials that one psychological therapy intervention is superior to any other (level of evidence: fair to good).

The AHCPR guidelines recommend against biofeedback in patients with acute low back problems (strength of evidence: C). The VA/DoD guideline recommendations are similar. The UK RCGP guidelines found conflicting evidence on effectiveness of biofeedback for chronic low back problems, and no evidence on effectiveness for acute low back problems (strength of evidence: *). The European COST guidelines recommend against psychological therapy for acute low back pain, but recommend it in patients with chronic low back pain.
Massage
Massage involves soft tissue manipulation using the hands or a mechanical device. It is administered using a variety of techniques, which vary in intensity and in the amount of pressure that is applied. Results of search: systematic reviews We identified one higher-quality Cochrane review700, 701 (8 trials, 5 higher-quality) and one lowerquality systematic review555 of massage for low back pain. We excluded two outdated systematic reviews399, 734 and one systematic review that evaluated case reports of adverse events associated with massage for any condition735.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials Eight unique trials of massage for low back pain were included in the two systematic reviews555, 700, 701 . We did not search for additional trials.
Efficacy of massage versus placebo or sham massage

Conclusions of the two systematic reviews were generally consistent555, 700, 701. Neither systematic review included any trial of massage versus placebo or sham massage. For subacute or chronic low back pain, the Cochrane review700, 701 included one higher-quality trial that found massage moderately superior to sham laser for short- and long-term pain intensity (SMD=-0.80, 95% CI -1.37 to -0.23 and -0.49, 95% CI -1.05 to 0.06, respectively) and substantially superior for short- and long-term functional status (SMD=-1.06, 95% CI -1.65 to 0.47 and SMD=-0.96, 95% CI -1.58 to -0.35, respectively736.
Efficacy of massage versus other interventions

Nearly all trials that compared massage to other interventions only assessed outcomes during or shortly following (within one month) a course of treatment in patients with subacute to chronic low back pain. Interpretation of results from these trials is a challenge because several trials with negative results evaluated superficial massage techniques, brief (10 to 15 minutes) treatment sessions, or few (<5) sessions. For back pain of varying duration, the Cochrane review included three lower-quality trials of massage versus spinal manipulation390, 737, 738. Two of the trials evaluated light or minimal massage techniques390, 737. Two trials found massage inferior to spinal manipulation for immediate (after the first session) relief of pain and improvement in function390, 738. In one of the trials, effects of spinal manipulation and massage were similar by the end of treatment and through three weeks of follow-up738. In the other trial, interpretation of findings is a challenge because of differences in baseline function scores390. A third trial found no differences between spinal manipulation and massage at any time point737. Other interventions have only been compared to massage in one or two trials each. For acute low back pain, one lower-quality trial found no difference between massage and application of a faradic current737. For chronic low back pain, one higher-quality trial found minimal massage inferior to TENS for the proportion of patients with at least 50% reduction in pain during the course of treatment (85% vs. 38%)708. One lower-quality trial found massage moderately superior to relaxation therapy739. Three trials (two higher-quality369, 736) compared massage to other interventions for subacute to chronic low back pain. One lower-quality trial found no differences between minimal massage and TENS or a corset390, 391. One higher-quality trial found no differences between massage and exercise therapy736. Another higher-quality trial found massage moderately superior to acupuncture or self-care education, with beneficial effects persisting through one year of followup369.
EVIDENCE REVIEW
Efficacy of one massage technique versus another

The Cochrane review found no clear difference between results of trials of manual massage and those that used a mechanical device700, 701. One higher-quality study740 found acupuncture massage superior to classical (Swedish) massage for improvements in pain and function. The greatest benefits from massage were observed in trials that used a trained massage therapist with many years of experience or a licensed massage therapist369, 736, 739. No conclusions could be drawn regarding differential effects associated with the number or duration of massage sessions700, 701. Harms One higher-quality trial included in the Cochrane review reported minor adverse events (such as significant pain or discomfort) in 13% of patients who received massage369. No serious adverse events were noted in any of the trials included in the Cochrane review700, 701, though most trials didnt report adverse events at all. One systematic review on safety of massage for any condition included case reports of serious adverse events (one large hematoma with slight anemia and one case of renal embolization) in two patients that received massage for low back pain735. Costs One trial found no significant differences (p=0.15) between HMO-related costs among massage ($139), acupuncture ($252), and a self-care education booklet ($200)369.
Summary of evidence
For acute low back pain, there is insufficient evidence to judge efficacy of massage. One lower-quality trial found no difference between massage and application of a faradic current (level of evidence: poor). For subacute or chronic low back pain, massage was moderately superior to sham laser for short- and long-term pain relief and moderately to substantially superior for functional outcomes in one higher-quality trial (level of evidence: fair). For back pain of varying duration, massage was inferior to spinal manipulation in two of three trials (all lower-quality) for immediate (after the first session) pain relief and improvement in functional status. However, differences were no longer present by the end of treatment sessions in two of three trials, the third trial evaluated groups with significant baseline differences in function scores, and two of the trials evaluated minimal massage interventions (level of evidence: poor). For chronic or subacute low back pain, minimal massage was inferior to TENS in one higherquality trial, but there were no differences between minimal massage and TENS in one lowerquality trial (level of evidence: poor). For chronic or subacute low back pain, one trial found no difference between massage and exercise plus a corset, one trial found massage moderately superior to relaxation therapy, and one trial found massage moderately superior to acupuncture or a self-care education book. Most trials only evaluated short-term outcomes, but one trial found that beneficial effects of
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain massage compared to acupuncture or a self-care education book persisted for one year (level of evidence for each comparison: poor to fair). One higher-quality trial found acupuncture massage superior to classical (Swedish) massage (level of evidence: fair). No serious adverse events were reported in trials of massage for low back pain, though quality of reporting was suboptimal (level of evidence: poor).

The AHCPR guidelines found physical agents and modalities (including massage) of insufficiently proven benefit to justify their cost (strength of evidence: C). The VA/DoD and UK RCGP guidelines reached similar conclusions. The European COST guidelines recommend against massage for acute low back pain and found insufficient evidence to recommend massage for chronic low back pain.
Modified work
Results of search: systematic reviews We excluded two lower-quality systematic reviews on effectiveness of return-to-work interventions for low back pain because neither specifically evaluated benefits or harms associated with modified work645, 741. We excluded another outdated systematic review on modified work for low back pain that only included one randomized trial (discussed below)742. Results of search: trials The systematic reviews all included one lower-quality randomized trial evaluating an occupational intervention (including modified work if necessary) versus a clinical intervention, both interventions, or neither intervention (also reviewed in the section on interdisciplinary interventions)646. We identified one other randomized trial not included in the systematic reviews that evaluated effects of efforts of an intervention to promote utilization of active sick leave, but it did not meet inclusion criteria because it did not evaluate effects of modified work on individual patients743, 744.
Efficacy of modified work versus no modified work

For subacute low back pain, one lower-quality randomized trial that compared an occupational intervention (including modified work), clinical intervention, both interventions, and neither intervention (usual care) found workers randomized to the two arms with the occupational intervention had about half as many lost work days than those randomized to the other two arms (60 and 67 days versus 120 and 131)646. However, it is difficult to assess the effects of modified work from this trial, as the occupational intervention also involved a work site visit and ergonomic adjustments, with modified work (light duties) only prescribed if deemed necessary. The excluded cluster randomized trial randomized municipalities in Norway to a proactive intervention versus a passive or no intervention to increase use of active sick leave743. It found no differences between interventions in median days of sick leave or proportion of patients returning to work before 50 weeks. The proactive intervention only slightly increased use of
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain active sick leave (18% vs. 12%), and the trial was not designed to evaluate effects of modified work on individual patients744. An outdated systematic review on modified work included only one randomized trial (discussed above)742. It also included 12 higher-quality observational studies that were consistent with the conclusion that modified work increases return to work. Only four of the 12 studies specifically evaluated low back pain patients, and only one of the four was prospective. In most studies the modified work intervention was evaluated as part of a more comprehensive occupational intervention. Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
For workers with subacute low back pain, there is insufficient evidence (one lower-quality trial) to evaluate effects of modified work on rates of return to work or other outcomes (level of evidence: poor).

The AHCPR guidelines state that activity recommendations for the employed patient with acute low back symptoms need to consider the patients age and general health, and the physical demands of required job tasks (strength of evidence: D). The VA/DoD guidelines are similar. The European COST guidelines state that temporary modified work (which may include ergonomic workplace adaptations) can be recommended, when needed, in order to facilitate earlier return to work for workers sicklisted due to low back pain (level B).
Spa therapy and balneotherapy

Balneotherapy involves immersion in baths containing thermal mineral waters at temperatures above 20 C. Spa therapy differs from balneotherapy in that it also involves physical therapy interventions (exercise and physical modalities) provided at a spa resort. Results of search: systematic reviews We identified one higher-quality systematic review on efficacy of spa therapy and balneotherapy for low back pain745. Results of search: trials The systematic review included three trials of spa therapy for chronic low back pain746-748 and two trials of balneotherapy for subacute or chronic low back pain696, 749. All trials were conducted in Europe.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Because the systematic review was published after we completed our initial draft of this report, we had already conducted a search for trials. From 88 potentially relevant citations, we identified the same five trials as the systematic review, and rated three higher-quality747-749. The systematic review rated two of the 5 trials higher-quality (at least 3 points on the 5-point Jadad scale)747, 749 and the other received 2 out of 5 points696, 746, 748. This difference did not affect conclusions.
Efficacy of spa therapy or balneotherapy versus no spa therapy or balneotherapy

For chronic low back pain, three trials (two rated higher-quality747, 748 found spa therapy associated with large benefits compared to no spa therapy for pain (differences of 20 to 30 points on a 100 point pain scale) and analgesic intake at the end of a three-week course of treatment, with benefits persisting for up to 9 months746-748. The systematic review calculated a WMD of 26.7 on a 100 point pain scale (95% CI 20.4 to 32.8). In two747, 748 of three746 trials, spa therapy was also superior to no spa therapy for functional status or disability.
Table 54. Trials of spa therapy versus no spa therapy
Author, year Constant, 1995748 (chronic LBP)
Constant, 1998747 (chronic LBP)
n=224 3 months
Guillemin, 1994746 (chronic LBP)
n=104 9 months
Main results Spa therapy vs. no spa therapy (mean improvement from baseline at 6 months) Pain, VAS (0 to 100 scale): -22.4 vs. +1.0, p<0.0001 Overall patient evaluation, (0 to 100 scale): +28.7 vs. +1.6, p<0.0001 RDQ Score (0 to 24): -5.1 vs. -0.9, p<0.0001 Spa therapy vs. no spa therapy (mean improvement from baseline at 3 months) Pain, VAS (0 to 100 scale): -37.6 vs. -14.2, p<0.0001 Overall patient evaluation (0 to 100 scale): +24.8 vs. +3.9, p<0.0001 RDQ Score (0 to 24): -4.0 vs. -1.1, p<0.0001 Spa therapy vs. no spa therapy (mean improvement from baseline at 9 months) Pain, VAS (0 to 100 scale): -34.4 vs. +7.1, p<0.0001 Waddell disability score: +0.09 vs. +0.18, NS
Quality score* 5/9
5/9
4/9
Efficacy of balneotherapy versus other interventions

For subacute to chronic (one to six month duration) low back pain, one higher-quality trial compared balneotherapy plus exercise to exercise alone (Table 55)749. It found no differences in pain scores one month after completion of a three-week course of treatment. For subacute or chronic low back pain, another lower-quality trial found balneotherapy superior to flexion and extension exercises by about 20 points on a 100 point pain scale after four weeks, though there were no differences in pain outcomes after one year696. Daily analgesic use significantly decreased in the balneotherapy group but not in the exercise group. There were no differences between balneotherapy and either underwater traction or underwater massage. Although the systematic review calculated a WMD of 18.8 points (95% CI 10.3 to 27.3) on a 0 to 100 pain scale in favor of balneotherapy based on these two trials, it only pooled early, short-term results.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Data from longer term follow-up (1 month to 1 year) showed smaller effects and no significant differences (Table 55).
Table 55. Trials of balneotherapy versus other interventions
Number of patients Duration of follow-up n=170 1 year Yurtkuran, 1997749 (subacute or chronic LBP) n=50 7 weeks
Author, year Konrad, 1992696 (subacute)
Main results Balneotherapy vs. underwater massage vs. underwater traction vs. exercise (mean improvement from baseline at 1 year) Pain, VAS (0 to 100 scale): -13.9 vs. -10.9 vs. -13.7 vs. -6.6 (NS) Balneotherapy + exercise versus exercise alone (mean improvement from baseline at 1 month) Pain, VAS (0 to 10 scale): -2.95 vs. -1.35 (NS)
Quality score* 4/9
5/9
Summary of evidence
There is no evidence on spa therapy for acute low back pain. For chronic low back pain, spa therapy was moderately to substantially superior to no spa therapy for pain in three trials (two higher-quality, all trials conducted in Europe) up to nine months after a three-week course of treatment, though effects on functional status were mixed (level of evidence: fair). For subacute or chronic low back pain, balneotherapy was no better than underwater massage, underwater traction, or exercise for pain relief after one month in one lower-quality trial (level of evidence: poor). For subacute or chronic low back pain, balneotherapy plus exercise therapy was no better than exercise therapy alone for pain relief after one year in one lower-quality trial, though balneotherapy was moderately superior for short-term pain relief (level of evidence: poor).

The other guidelines do not address spa therapy.
Spinal manipulation
Spinal manipulation refers to manual therapy in which loads are applied to the spine using short or long lever methods. Using these methods, high-velocity thrusts are applied to a spinal joint beyond its restricted or normal range of movement. Spinal mobilization (low-velocity, passive movements within or at the limit of joint range) is often used in conjunction with manipulation.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: systematic reviews We identified 12 systematic reviews of spinal manipulation for low back pain that met inclusion criteria. Six (including a Cochrane review of 39 trials66, 67) were rated higher-quality100, 555, 750-752 and six lower-quality753-758. Four other systematic reviews specifically evaluated harms of spinal manipulation (most including observational studies as well as randomized trials)759-762 and one higher-quality systematic review evaluated whether trials that permitted discretion in manipulation techniques found larger benefits than trials that didnt allow discretion763. We excluded 17 outdated systematic reviews193, 345, 346, 623, 764-776 and three systematic reviews that either evaluated cervical manipulation only or cervical and lumbar manipulation together777-779. Results of search: trials Sixty-nine unique trials on efficacy of spinal manipulation were included in twelve systematic reviews. Nearly all of the trials evaluated patients with non-specific low back pain, mixed populations with and without sciatica, or did not specify presence or absence of sciatica. For example, in the Cochrane review,12 of 39 trials included patients with or without sciatica, but only three reported results specifically in patients with sciatica. The number of manipulation sessions in the trials ranged from 1 to 24. We also identified two large (n=681 and n=1334), recently published trials (the UK BEAM Trial629 and the UCLA Low Back Pain Study780, 781) and one smaller (n=102) trial of spinal manipulation for acute low back pain with sciatica and herniated lumbar disc782 not included in the systematic reviews.
Efficacy of spinal manipulation versus sham, placebo, or therapies judged ineffective

For acute low back pain, the Cochrane review found spinal manipulation slightly to moderately superior to sham manipulation for short-term pain relief in a meta-regression (WMD=-10 mm on a 100 mm VAS, 95% CI=2 to 17)66, 67. However, the only trial that reported pain relief for acute low back pain was a lower-quality trial that included patients with acute or subacute (<3 months duration) sacroiliac pain783 Based on two trials (one higher-quality784, 785), spinal manipulation was moderately more effective than sham manipulation on short-term function (RDQ), but the difference did not reach statistical significance (WMD=-2.8, 95% CI -5.6 to +0.1). Compared to therapies judged to be ineffective or harmful (traction, bed rest, home care, topical gel, no treatment, diathermy, and minimal massage) spinal manipulation was statistically superior for short-term pain relief, but the difference was not clinically significant (WMD=-4 on a 0 100 mm VAS, 95% CI -8 to -1). Compared to ineffective or possibly harmful therapies, difference on the short-term RDQ favored spinal manipulation but did not reach statistical significance (WMD=2.1, 95% CI -4.4 to +0.2). There were no differences between spinal manipulation and either sham manipulation or the therapies judged to be ineffective or harmful in long-term pain or function. Only one lower-quality trial in the Cochrane review evaluated efficacy of spinal manipulation versus sham, placebo, or ineffective therapies specifically in patients with sciatica. It found no significant differences between spinal manipulation and a placebo gel for either acute/subacute
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain or chronic sciatica, though trends favored manipulation356. For acute sciatica with a radiologically confirmed herniated disc, a higher-quality trial not included in the Cochrane review found spinal manipulation substantially superior to sham manipulation for the proportion free of radicular pain after six months (55% vs. 20%, p<0.0001), though there were no significant differences in SF-36 scores (Table 56)782.
Table 56. Trial of spinal manipulation for acute sciatica with prolapsed lumbar disc
Author, year Santilli, 2006782
Main results Manipulation vs. sham manipulation Proportion pain-free (radiating pain) at 180 days: 55% (29/53) vs. 20% (10/49), p<0.0001 Proportion pain-free (local pain) at 180 days: 28% (15/53) vs. 6% (3/49) Use of NSAIDs (days): 1.8 vs. 3.7 days SF-36: No differences Kellner symptom scale: No differences
Quality score 5/9
For chronic low back pain, the Cochrane review found spinal manipulation associated with moderate improvements in short- or long-term pain and short-term function compared to sham manipulation (3 trials) or therapies judged to be ineffective or harmful (5 trials)66, 67 . Against sham manipulation, differences in short- and long-term pain averaged 10 mm (95% CI 3 to 17) and 19 mm (95% CI 3 to 35) on a 100 mm VAS, and differences for short-term function averaged 3.3 points (95% CI 0.6 to 6.0) on the RDQ. Conclusions were insensitive to different cutoffs for classification of studies as higher-quality or to the profession of the manipulator (chiropractor or other). There was insufficient data to judge effects of presence or absence of sciatica on benefits. No trials evaluated efficacy of spinal manipulation under anesthesia66, 67, 786 . A recent technology report funded by the Canadian Coordinating Office for Health Technology Assessment (CCOHTA) reviewed 14 published systematic reviews of spinal manipulation750. It concluded that the Cochrane review66, 67 was the best available summary of clinical effectiveness because it received a high quality score, was published recently, and included the largest number of trials. The CCOHTA report also identified two additional randomized trials and two non-randomized trials that did not change the overall conclusions of the Cochrane review. Four other higher-quality100, 555, 751, 752 and six lower-quality753-758 systematic reviews also found spinal manipulation superior to placebo, sham, or therapies thought to be ineffective. One higher-quality systematic review found that trials that permitted providers to tailor specific spinal manipulation techniques to individual patients did not report better outcomes than trials that did not allow therapeutic discretion763. In fact, spinal manipulation was associated with better short-term outcomes in trials that didnt allow discretion, though long-term outcomes were similar. These conclusions should be interpreted with caution because they involve indirect, cross-trial comparisons.
EVIDENCE REVIEW
Efficacy of spinal manipulation versus usual care or other interventions

For acute low back pain, the Cochrane review found spinal manipulation associated with no clinically or statistically significant advantages over usual general practitioner care or analgesics (3 trials), physical therapy or exercises (5 trials), and back school (2 trials)66, 67. For chronic low back pain, there were no differences between manipulation and general practitioner care or analgesics (6 trials), physiotherapy or exercises (4 trials), and back school (3 trials). For sciatica of varying duration, three trials found no differences between spinal manipulation and other interventions151, 356, 389. Five other higher-quality systematic reviews (including one that focused on patients with sciatica100) also found no clear differences between spinal manipulation and other interventions100, 555, 750-752. The two most comprehensive lower-quality systematic reviews found spinal manipulation superior to some other effective interventions754, 758. However, conclusions regarding superiority of spinal manipulation over other interventions were generally based on sparse data (one to three trials, often lower-quality, and often with small sample sizes) or did not appear to adequately consider inconsistency when results of some trials or outcomes demonstrated no differences. Two large, recently published trials reported results consistent with the Cochrane review. For low back pain of unspecified duration, the higher-quality UCLA Low Back Pain Study found no differences in pain or functional status scores between those randomized to chiropractic versus medical care through 18 months (Table 57), though patients randomized to chiropractic care perceived themselves to be more improved780, 781. The other trial found manipulation slightly superior to usual care for back-specific functional status, pain, and disability in patients with subacute or chronic low back pain, though beneficial effects were diminished after 12 months compared to after 3 months629. There were no significant differences between manipulation and exercise, though trends favored manipulation.
EVIDENCE REVIEW
Table 57. Results of the UK BEAM Trial and the UCLA Low Back Pain Study
Number of patients Duration of follow-up n=681
Author, year Hurwitz, 2002780, 781
Main results Chiropractic care vs. medical care (adjusted betweengroup difference in improvement from baseline) UCLA Low Back Pain 6 months Most severe pain (0 to 10 scale): -0.25 (95% CI -0.96 to Study 0.45) at 6 months, -0.64 (95% CI -1.38 to -0.21) at 18 months Average pain (0 to 10 scale): -0.26 (95% CI -0.81 to 0.29) at 6 months, -0.50 (-1.09 to 0.08) at 18 months RDQ score (0 to 24 scale): -0.37 (95% CI -1.63 to 0.90) at 6 months, -0.69 (-2.02 to 0.65) at 18 months Manipulation + exercise versus manipulation alone UK BEAM Trial, 2004629 n=1334 versus exercise alone (all results are net benefit relative to usual care at 12 months) 12 months RDQ score (0 to 24 scale): 1.30 (95% CI 0.54 to 2.07) vs. 1.01 (95% CI 0.22 to 1.81) vs. 0.39 (95% CI -0.41 to 1.19) Modified Von Korff pain score (0 to 100 scale): 6.71 (95% CI 2.47 to 10.95) vs. 5.87 (95% CI 1.58 to 10.17) vs. 4.90 (95% CI 0.30 to 9.50) Modified Von Korff disability score (0 to 100 scale): 6.71 (95% CI 2.62 to 10.80) vs. 5.65 (95% CI 1.57 to 9.72) vs. 4.56 (95% CI 0.34 to 8.78) * Excludes criteria involving blinding of care providers, for maximum score of 10 ** Excludes criteria involving blinding of patients and care providers, for maximum score of 9
Quality score 7/9**
2/10*
Harms Five systematic reviews consistently found serious adverse events such as worsening lumbar disc herniation or cauda equina syndrome following lumbar spinal manipulation therapy to be very rare555, 759-762. One systematic review found no serious complications reported in over 70 controlled clinical trials760. Including data from observational studies, the estimated risk for serious adverse events was lower than 1 per 1 million patient visits761, 762. Current guidelines recommend against spinal manipulation in patients with severe or progressive neurologic deficits. Costs In the UCLA Low Back Pain Study, costs were higher with chiropractic care relative to medical care ($560 versus $369, p<0.001)787. Because outcomes were very similar for the two interventions, this is essentially a cost-minimization analysis. In the UK BEAM Trial, manipulation was associated with an incremental cost-effectiveness of 4800/QALY (about $9,264/QALY) relative to best care and 2300/QALY ($4,439/QALY) relative to exercise629. Two other trials that compared spinal manipulation to exercise therapy found similar costs and outcomes for the two interventions367, 788, 789. In one of the trials, chiropractic care was more costly then a self-care booklet ($429 versus $153), with only modest differences in patient outcomes367.
EVIDENCE REVIEW
Summary of evidence
For acute low back pain, spinal manipulation was slightly to moderately superior to sham manipulation for pain relief, but results are primarily based on a small, lower-quality trial of patients with acute or subacute sacroiliac symptoms. Spinal manipulation was moderately superior to sham manipulation for functional outcomes in two trials (one higher-quality), but the difference just missed reaching statistical significance. Spinal manipulation was not effective versus sham manipulation for long-term outcomes (level of evidence: fair). For acute low back pain, spinal manipulation was statistically superior to no treatment or therapies thought to be ineffective or harmful for short-term pain relief, but differences were not clinically meaningful. Spinal manipulation was moderately superior for short-term functional status, but the difference just missed reaching statistical significance (level of evidence: good). For acute low back pain, there are no clear differences between spinal manipulation and analgesics/usual care (3 trials), exercise therapy (6 trials), or back school (2 trials) (level of evidence: fair). For chronic low back pain, evidence from eleven trials found spinal manipulation moderately superior to sham, no treatment, or therapies thought to be ineffective or harmful for pain relief and functional status (level of evidence: good). For chronic low back pain, there is no consistent evidence from a number of trials of clinically significant differences between spinal manipulation and other non-invasive interventions thought to be effective (level of evidence: good). For acute sciatica, one higher-quality trial found spinal manipulation substantially superior to sham manipulation for the proportion free of radicular pain after 6 months (level of evidence: fair). For sciatica of mixed duration, outcomes favored spinal manipulation over a placebo gel in one lower-quality trial, but differences were not significant (level of evidence: poor). For sciatica of mixed duration, there were no differences between spinal manipulation and other non-invasive interventions in three trials (level of evidence: fair). In patients without severe or progressive neurologic deficits, serious adverse events such as cauda equina syndrome or worsening lumbar disc herniation following lumbar spinal manipulation are very rare (level of evidence: good).

The AHCPR guidelines found that manipulation can be helpful in patients with acute low back problems without radiculopathy when used within the first month of symptoms (strength of evidence: B). The AHCPR guidelines found insufficient evidence to recommend manipulation in patients with radiculopathy (strength of evidence: C).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The AHCPR guidelines found that a trial of manipulation in patients without radiculopathy with symptoms longer than one month is probably safe, but efficacy unproven (strength of evidence: C). The AHCPR guidelines recommended an appropriate diagnostic assessment to rule out serious neurologic conditions prior to initiating manipulation therapy when progressive or severe neurologic deficits are present (strength of evidence: D). The VA/DoD guidelines for manipulation are essentially identical to the AHCPR guidelines. The UK RCGP guidelines found manipulation superior for short-term improvement in pain and activity levels and higher patient satisfaction compared to comparison treatments in patients with acute and subacute back pain (strength of evidence: **). The UK RCGP guidelines found that the risks of manipulation for low back pain are very low, provided patients are selected and assessed properly and manipulation is performed by a trained therapist or practitioner (strength of evidence: **). The UK RCGP guidelines found no firm evidence regarding what kind of manipulation is most effective, or optimum timing of manipulation (strength of evidence: *). The UK RCGP guidelines recommend against manipulation under general anesthesia (strength of evidence: *). The European COST guidelines recommend considering referral for spinal manipulation patients with acute low back pain who are failing to return to normal activities, and a shortcourse of spinal manipulation/mobilization as a treatment option for chronic low back pain.
EVIDENCE REVIEW
Table 58. Systematic reviews on efficacy of non-pharmacologic therapies for low back pain
Number of trials not included in any other relevant systematic review 11 Overall quality using Oxman scale (1 to 7) 7
Author, year Furlan, 200569, 70
Number of included trials (number rated higher-quality)* 35 (14)
Duration of treatment in included trials 1 to 20 sessions
Interventions evaluated (number of trials) Acupuncture (32), dry needling (3)
Main conclusions Acupuncture vs. no treatment for chronic LBP: SMD=-0.73 (95% CI -1.19 to -0.28) for short-term pain (2 RCTs) and SMD=-0.63 (95% CI -1.08 to -0.19) for short-term function (2 RCTs) Acupuncture vs. sham acupuncture: WMD=-17.79 (95% CI -25.5 to -10.07) for short-term pain (6 RCTs), WMD= -5.74 (95% CI -14.7 to 3.25) for longterm pain (3 RCTs), no difference for function Acupuncture vs. no additional treatment for chronic LBP: SMD=-0.69 (95% CI -0.98 to -0.40) for short-term pain (8 RCTs), SMD=-0.74 (95% CI -1.47 to -0.02) for long-term pain (5 RCTs), SMD=-0.62 (95% CI -0.95 to -0.30) for short-term function (6 RCTs) Acupuncture vs. sham acupuncture: SMD=-0.58 (95% CI -0.36 to -0.80), for short-term pain (4 RCTs), SMD=0.59 (95% CI -1.29 to +0.10) for longterm pain (2 RCTs), no difference for function Back school vs. control: Rate difference for return to work rate ranged from -7% to 29% after 21 to 42 days (4 RCTs); 30% to 37% after 180200 days (3 RCTs); -1% to 42% after 360 days (4 RCTs)
Acupuncture (51 unique trials in three systematic reviews)
Manheimer, 200568
Quantitative
33 (5)
10
1 to 20 sessions
17 to 194 (median=60)
Chinese acupuncture (29), Western acupuncture (4), electroacupuncture (14), acupuncture for antenatal LBP (3)
Back schools (31 unique trials in three systematic reviews) Elders, 2000588 Qualitative and quantitative 6 trials of back schools (quality not assessed) 3 Not reported 51 to 975 (median=194) Not described 3
EVIDENCE REVIEW
Author, year Heymans, 2005

587 586,
Duration of treatment in included trials 1, 4-hour session to 21, 85minutes sessions
Interventions evaluated (number of trials) Swedish or modified Swedish back school (6), Maastricht (2), others (11)
Main conclusions Conflicting evidence from 8 RCTs on effectiveness of back schools for chronic LBP versus wait-list control or placebo for short-, intermediate-, or long-term pain, functional status, and return to work Back school in occupational setting appeared to more effective Back school vs. any control: SMD +0.14 (p=0.026) for pain intensity at <3 months (9 RCTs), SMD=0.44 (p=0.001) for recurring back pain through 6 months (6 RCTs), no significant differences for functional status (7 RCTs) or recurring back pain after 6 months McKenzie therapy versus control (booklet, strength training, spinal mobilization, or massage): WMD= -8.6 (95% CI -13.7 to -3.5) on a 100 point scale for short-term (<3 months ) pain (3 RCTs) and WMD=-5.4 (95% CI -8.4 to -2.4) for short-term disability (5 RCTs); no differences for intermediate-term disability
Maier-Riehle, 589 2001
Quantitative
13 (quality not assessed)
1 to 22 hours (median=5)
29 to 299 (median=76)
Not described
Exercise (seventy-nine unique trials in seven systematic reviews) Clare, 2004616 Quantitative 5 (3) 1 Not reported 25 to 321 All trials evaluated McKenzie method 6
EVIDENCE REVIEW
Author, year Hayden, 2005

615 613,
Type of systematic review Quantitative and qualitative
Duration of treatment in included trials 2 to 150 hours
Interventions evaluated (number of trials) McKenzie (6), extensor (5), flexion (9), isometric (3), aerobics (8), strengthening (16), stretching (12), graded activity (2), other or multiple (17) Outpatient exercise therapy (9), inpatient (3), back school (3), interdisciplinary/ functional restoration (5) Strength/ flexibility (9), multimodal (3), other (4)
Main conclusions Exercise therapy vs. no treatment for acute LBP: WMD=-0.59 (95% CI -12.69 to 11.51) on 100 point scale for short-term pain (3 RCTs), no differences for function Exercise therapy vs. no treatment for chronic LBP: WMD=10.2 (95% CI 1.31 to 19.09) for short-term pain (19 RCTs) and WMD=3.00 (95% CI -0.53 to 6.48) for short-term function (17 RCTs); results similar at longer-term follow-up Exercise vs. usual care: SMD=-0.24 (95% CI -0.36 to -0.11) for number of sick days during first year follow-up (9 RCTs), RR=1.37 (95% CI=1.05 to 1.78) for proportion at work after one year (3 RCTs)
Kool, 2004617
14 (9)
3 weeks to 12 months
80 to 476 (median=166)
Liddle, 2004618
Qualitative
16 (8)
Not reported
28 to 222 (median=99)
Exercise vs. control: 9 of 16 RCTs reported a "positive result" (on any outcome) vs. control (waiting list, advice, or electrotherapy), 7 other RCTs reported "positive result" but no difference compared to control (usually exercise-based); 5 of 7 RCTs reported positive result for backspecific function
EVIDENCE REVIEW
Author, year Machado, 2006

619
Duration of treatment in included trials Not clearly reported
Interventions evaluated (number of trials) All trials evaluated McKenzie method
Main conclusions McKenzie versus passive therapy (educational booklets, bed rest, ice packs, and massage) for acute LBP: WMD=-4.16 (95% CI -7.12 to -1.20) on 100 point scale for pain (4 RCTs) and WMD=-5.22 (95% CI -8.28 to 2.16) for disability at 1 week follow-up; no differences at 4 weeks (4 RCTs) McKenzie versus advice to stay active for acute LBP: WMD=+3.85 (95% CI +0.30 to +7.39) for disability at 12 weeks follow-up (2 RCTs) No differences between McKenzie and other exercise therapy Unable to draw firm conclusions regarding exercise therapy for spondylolysis and spondylolisthesis
McNeely, 2003620
Qualitative (exercise therapy for spondylolysis and spondylolisthesis)
2 (1)
Not reported
44 and 65
Strengthening (1), flexion/ extension (1)
EVIDENCE REVIEW
Number of trials not included in any other relevant systematic review 12 trials not included in systematic reviews of interdisciplinary therapy Overall quality using Oxman scale (1 to 7) 6
Author, year Schonstein, 2003302, 303
Duration of treatment in included trials One session to weekly sessions for 1.5 years
Interventions evaluated (number of trials) Cognitivebehavioral component (10), no cognitivebehavioral component (8)
Main conclusions Physical conditioning vs. usual care for time lost from work: WMD=-45 (95% CI -88 to -3) for number of sick leave days after one year (2 RCTs); OR 0.80 (95% CI 0.58 to 1.09) for proportion off work at 12 months (3 RCTs) Physical conditioning vs. physical conditioning plus psychological treatment: OR=0.93 (95% CI 0.44 to 1.97) for proportion off work at 6 or 12 months (2 RCTs) Strong evidence that intensive (>100 hour) daily interdisciplinary therapy is more effective than usual care or less intensive therapy for function (3 RCTs) Moderate evidence that less intensive (<30 hour) interdisciplinary therapy is no more effective than usual care or non-multidisciplinary therapy (5 RCTs) Moderate evidence that multidisciplinary rehabilitation with a work site visit or more comprehensive occupational health care intervention is more effective than usual care for return to work, sick leave, and subjective disability (2 RCTs)
Functional restoration (18 trials in one systematic review)
Inter-disciplinary therapy (16 unique trials in three systematic reviews) Guzman, 2001643,
644
Quantitative (chronic low back pain)
10 (3)
10
Once weekly to daily sessions
20 to 476 (median=170)
Higher intensity (4), lower intensity (4), other (3)
Karjalainen, 2001299, 300
Qualitative (subacute low back pain)
2 (0)
Not reported
103 and 130
Interdisciplinar y therapy not categorized
EVIDENCE REVIEW
Author, year Tveito, 2004

645
Duration of treatment in included trials Not reported
Interventions evaluated (number of trials) Interdisciplinar y therapy not categorized
Main conclusions Moderate evidence that interdisciplinary therapy has a positive effect on sick leave (4 trials), no evidence of a positive effect on pain (1 trial) Massage superior to sham laser in 1 RCT Relative to other therapies, massage superior to relaxation therapy, acupuncture, and self-care education; massage similar to corset and exercises; light massage inferior to manipulation and TENS Insufficient evidence to assess efficacy of lumbar support versus no treatment (1 RCT); lumbar support superior to other interventions in 1 of 4 RCTs
Massage (8 unique trials in two systematic reviews) Furlan, 2002700, 701 Qualitative 8 (5) Not applicable 5 to 9 sessions 24 to 262 (median=106) Massage with hands (6), massage with mechanical device (2) 6
Lumbar supports (six trials in one systematic review) Jellema, 2001385; Van Tulder, 2000384 Qualitative 6 trials of treatment (2) Not applicable 3 to 8 weeks (median=3. 5 weeks) 19 to 334 (median=190) Lumbar support with rigid stay (2), pneumatic lumbar support (1), other or not specified (3) Neuroreflexoth erapy (3) 7
Neuroreflexotherapy (three trials in one systematic review) Urrutia, 2004582 Qualitative 3 (2) Not applicable 1 to 1.4 treatments 78 to 104 Neuroreflexotherapy substantially superior to sham therapy (2 RCTs) and usual care (1 RCT) 6
EVIDENCE REVIEW
Author, year Hoffman, 2007722
Duration of treatment in included trials Not reported
Interventions evaluated (number of trials) Not described
Main conclusions Any psychological intervention or multidisciplinary intervention vs. wait list controls: SMD=0.50 (95% CI 0.23 to 0.77) for pain intensity (7 RCTs), SMD=0.50 (95% CI 0.00 to 0.83) for health-related quality of life (4 RCTs) Cognitive-behavioral treatment vs. wait list controls: SMD=0.62 (95% CI 0.25 to 0.98) for pain intensity (7 RCTs) Self-regulatory treatment vs. wait list controls: SMD=0.75 (95% CI 0.35 to 1.15) for pain intensity (4 RCTs) Progressive relaxation versus wait list controls: SMD=1.16 (95% CI 0.47 to 1.85) for pain intensity (2 RCTs) Biofeedback versus wait list controls: SMD=0.84 (95% CI 0.32 to 1.35) for pain intensity (3 RCTs) Operant therapy versus wait list controls: SMD=0.29 (95% CI -0.14 to 0.72) for pain intensity (2 RCTs) Cognitive-behavioral therapy: SMD=0.59 (95% CI 0.10 to 1.09) for pain intensity (4 RCTs)
Psychological therapies (35 unique trials in two systematic reviews)
Ostelo, 2005790
Quantitative and qualitative
21 (7)
13
3 to 12 weeks
17 to 161 (median=66)
Cognitive behavioral (14), operant (7), relaxation (11), biofeedback (6)
EVIDENCE REVIEW
Number of trials not included in any other relevant systematic review Not applicable Overall quality using Oxman scale (1 to 7) 7
Author, year Pittler, 2006745
Duration of treatment in included trials 3 to 4 weeks with 3 to 6 sessions weekly
Interventions evaluated (number of trials) Spa therapy (2), balneotherapy (3)
Main conclusions Spa therapy vs. wait list control for chronic LBP: WMD=-26.6 (95% CI -32.8 to -20.4) for pain relief (3 RCTs) Balneotherapy vs. NSAIDs or exercise therapy: WMD=-18.8 (95% CI -27.3 to -10.3) for immediate pain relief (2 RCTs) Spinal manipulation vs. sham for acute LBP: WMD=-10 mm (95% CI -17 to -22) on 100 mm VAS for shortterm pain and WMD=-2.8 (95% CI -5.6 to +0.1) for short-term function (RDQ) Spinal manipulation vs. sham for chronic LBP: WMD=-10 mm (95% CI -17 to -33) on 100 mm VAS for shortterm pain, WMD=-19 mm (95% CI -35 to -3) for long-term pain, and WMD= -3.3 (95% CI -6.0 to -0.6) for shortterm function (RDQ) No differences between spinal manipulation and other therapies judged effective for either acute or chronic LBP Insufficient new evidence to assess efficacy of spinal manipulation (updates previous review by Mohseni774 Bandpei et al )
Spa therapy and balneotherapy (five trials in one systematic review
Spinal manipulation (69 unique trials in twelve systematic reviews) Assendelft, 200366,
67
Quantitative
39 (10)
1 session to 24 sessions over 3 weeks
21 to 741 (median= 103)
Rotational manipulation (6), Maitland method (5), thrust (3), sacroiliac (2), other or not specified (23)
Avery, 2004753
Qualitative
3 (quality not assessed)
Not reported
155 to 323
Chiropractic spinal manipulation (2), osteopathic (1)
EVIDENCE REVIEW
Author, year Bronfort, 2004

754
Sample sizes in included trials 5202 subjects (mean=168)
Interventions evaluated (number of trials) Spinal manipulation (26), mobilization only (5)
Main conclusions Moderate evidence that spinal manipulation is similar to prescriptions non-steroidal anti-inflammatory drugs for chronic low back pain; limited to moderate evidence that spinal manipulation is superior to some other interventions for acute and chronic LBP Spinal manipulation is as effective as other non-invasive treatments Chiropractic spinal manipulation superior to control treatments in 5 of 12 RCTs. Chiropractic manipulation consistently superior to sham manipulation. Beneficial effects usually small or moderate. No clear difference between results for acute vs. chronic low back pain. Spinal manipulation vs. placebo: WMD=7 mm (95% CI 1 to 14) on 100 mm VAS for short-term pain (2 RCTs) Spinal manipulation vs. NSAIDs: WMD=14 mm (95% CI -11 to 40) for short-term pain (2 RCTs) and 6 points (95% CI 1 to 12) on 100 mm scale for disability (2 RCTS) No differences between spinal manipulation and other effective therapies
Brown, 2005750
Qualitative
14 (6) systematic reviews and 2 (2) RCTs 12 (6)
Not reported 4 to 12 sessions
Not reported
Not reported
Ernst, 2003755
Qualitative
12 to 741 (median= 69)
All trials evaluated chiropractic manipulation
Ferreira, 2002752
Quantitative
8 (4)
4 to 12 sessions
19 to 395 (median=63)
Not specified
EVIDENCE REVIEW
Author, year Ferreira, 2003

751
Duration of treatment in included trials 1 to 14 sessions (mean 6.8)
Sample sizes in included trials 3817 subjects (mean=146)
Interventions evaluated (number of trials) High-velocity thrust (11), high-velocity thrust plus other techniques (8), high-velocity thrust plus low-velocity mobilization (7), compared different types of manipulation (1) Distraction manipulation (1) All trials evaluated osteopathic spinal manipulation
Main conclusions High-velocity thrust spinal manipulation vs. sham manipulation or no treatment for LBP <3 months: WMD=18 (95% CI 13 to 24) on a 100 point scale for short-term pain (3 RCTs), WMD=9 (95% CI 1 to 17) on a 100 point scale for short-term disability (3 RCTs) No differences between spinal manipulation and other effective therapies
Gay, 2005756
Qualitative
1 (quality not assessed) 6 (quality not assessed)
Not reported 4 to 11 sessions
30
Insufficient evidence to assess efficacy of distraction manipulation Osteopathic spinal manipulation vs. control treatment: SMD=-0.30 (95% CI -0.47 to -0.13) for pain reduction (8 comparison from 6 RCTs)
Licciardone2005757
Quantitative
30 to 178 (median=93)
EVIDENCE REVIEW
Author, year Woodhead, 758 2005
Interventions evaluated (number of trials) Rotational (8), Maitland (5), sacroiliac (3), other or not specified (46)
Main conclusions Limited evidence that spinal manipulation is more effective than placebo for acute LBP and moderate evidence that spinal manipulation is more effective than placebo for chronic or subacute LBP Moderate evidence that spinal manipulation is more effective than some other interventions for acute LBP and strong evidence that spinal manipulation is more effective than some other interventions for chronic LBP
Superficial heat (9 trials in 1 systematic review) French, 2006398 Quantitative 9 (5) Not applicable Single application to 7 days 36 to 371 (median=90) Superficial heat (9), superficial cold (2) Heat wrap versus oral placebo or nonheated wrap for acute or subacute LBP (4 RCTs): WMD=1.06 (95% CI 0.68 to 1.45 on a 0 to 5 scale) for pain relief up to day 5 (2 RCTs); WMD= -2.10 (95% CI -3.19 to -1.01) for RDQ (2 RCTs) Insufficient evidence to assess efficacy of superficial heat versus superficial cold 7
EVIDENCE REVIEW
Author, year Clarke, 2005676, 677
Duration of treatment in included trials 1 week to 8 weeks
Interventions evaluated (number of trials) Mechanical or manual traction (13), autotraction (6), Tru-Trac (3), underwater (1), other (3)
Main conclusions Strong evidence that continuous traction is not superior to placebo, sham, or no treatment for any outcome at 3 months or 6 weeks in patients with or without sciatica (2 RCTs) Moderate evidence that autotraction is more effective than placebo, sham, or no treatment for pain, global improvement, or work absenteeism in patients with sciatica (2 RCTs); moderate evidence that other forms of traction not more effective than control (8 RCTs) Traction vs. sham traction: 6 RCTs (1 higher-quality) reported negative results (1 RCT inconclusive)
Traction (24 unique trials in three systematic reviews)
Harte, 2003678
Qualitative
13 (1)
1 week to 8 weeks
16 to 334 (median=62)
Mechanical or manual traction (7), autotraction (2), Tru-Trac (2), other (3) TENS given at clinic (1), TENS selfadministered at home (1)
Transcutaneous electrical nerve stimulation (11 trials in six systematic reviews)** Khadilkar, 2005698,
699
Qualitative
2 (1)
Single session and 4 weeks
30 and 145
TENS vs. placebo (2 RCTs, 1 goodquality): TENS not superior to placebo for any outcomes measured (pain, functional status, range of motion, use of medical services) in 1 good-quality RCT In the other RCT, TENS superior for subjective pain intensity for 60 minutes post treatment; no longerterm follow-up
EVIDENCE REVIEW
Author, year Cherkin, 2003555
Number of included trials (number rated higher-quality)* 8 systematic reviews, 9 RCTs (quality not assessed) 12 (4) trials of traction or ultrasound 8 (3) trials of traction, exercise, or spinal manipulation
Duration of treatment in included trials 2 to 12 weeks (RCTs)
Sample sizes in included trials 24 to 262 (RCTs)
Interventions evaluated (number of trials) Acupuncture (20), massage (3), spinal manipulation (26) Traction (10), ultrasound (2) Traction (7), exercise (2), spinal manipulation (2)
Main conclusions Effectiveness of acupuncture unclear. Massage effective for subacute and chronic LBP in 3 RCTs Spinal manipulation equivalent to other commonly used therapies No benefit demonstrated for traction or ultrasound for acute, subacute, or chronic LBP Traction vs. sham, infrared heat, or corset for sciatica: OR=1.2 (95% CI 0.7 to 2.0) for treatment success (4 RCTs) Insufficient evidence to evaluate efficacy of exercise or spinal manipulation for sciatica
Multiple interventions
Philadelphia Panel, 399 2001 Vroomen, 2000100
Qualitative
2 (RCTs of ultrasound) 0
1 to 5 weeks Not reported
16 to 322 (median=60) 44 to 322 (median=77)
Quantitative
*Trials adequately meeting at least half of the quality rating criteria or rated as good or higher-quality if the number of criteria met was not reported **Including trials of TENS included in systematic reviews of acupuncture68, massage701, superficial heat398, and traction676 22 trials of behavioral therapy alone or as part of interdisciplinary rehabilitation CI=confidence interval, ODI=Oswestry Disability Index, LBP=low back pain, OR=odds ratio, RCT=randomized controlled trial, RDQ=Roland-Morris Disability Questionnaire, RR=relative risk, TENS=transcutaneous electrical nerve stimulation, WMD=weighted mean difference
EVIDENCE REVIEW
Table 59. Summary of evidence on non-pharmacologic therapies for acute low back pain
Number of trials (number rated higher-quality by at least one systematic review) 0 4 (3) 1 (0) 0 1 (0) 13 (7) 4 (3)
Intervention Acupressure Acupuncture Back schools Brief educational interventions Dry needling Exercise Functional restoration Hydrotherapy Interdisciplinary rehabilitation Interferential therapy Low-level laser therapy Lumbar supports Massage therapy Neuroreflexotherapy Percutaneous electrical nerve stimulation Psychological therapies
Net benefit* No evidence Unable to estimate Unable to estimate No evidence Unable to estimate Not effective Not effective
Effective vs. placebo, sham, wait list, or no treatment? No evidence Unclear (2 trials) Unclear (1 trial) No evidence No evidence No (9 trials) Yes (3 trials)
Inconsistency? No evidence Some inconsistency Not applicable No evidence Not applicable Some inconsistency Some inconsistency
Directness of evidence? No evidence Direct Direct No evidence Not applicable Direct Direct
Overall quality of evidence No evidence Poor Poor No evidence Poor Good Fair
Comments
Most trials found no effect Most trials found no effect, but studies were heterogeneous
0 0 0 0 1 (0) 1 (0) 0 0
No evidence No evidence No evidence No evidence Unable to estimate Unable to estimate No evidence No evidence
No evidence No evidence No evidence No evidence No evidence No evidence No evidence No evidence
No evidence No evidence No evidence No evidence Not applicable Not applicable No evidence No evidence
No evidence No evidence No evidence No evidence Direct Direct No evidence No evidence
No evidence No evidence No evidence No evidence Poor Poor No evidence No evidence
No evidence
No evidence
No evidence
No evidence
No evidence
EVIDENCE REVIEW
Table 59. Summary of evidence on non-pharmacologic therapies for acute low back pain
Number of trials (number rated higher-quality by at least one systematic review) 1 (0) 0 11 (2) 5 (5) 0
Intervention Shortwave diathermy Spa therapy and balneotherapy Spinal manipulation Superficial heat Traction
Net benefit* Unable to estimate No evidence Small to moderate Moderate No evidence
Effective vs. placebo, sham, wait list, or no treatment? No evidence No evidence Yes (2 trials) Yes (2 trials) No evidence
Inconsistency? Not applicable No evidence No No No evidence
Directness of evidence? Direct No evidence Direct Direct No evidence
Overall quality of evidence Poor No evidence Fair Good No evidence
Comments
Most trials included patients with back pain of varying duration, with or without sciatica
Transcutaneous electrical nerve stimulation (TENS) Ultrasound Yoga
1 (0)
Unable to estimate No evidence No evidence
No evidence
Not applicable
Direct
Poor
0 0
EVIDENCE REVIEW
Table 60. Summary of evidence on non-pharmacologic therapies for chronic or subacute low back pain
Intervention Acupressure
Net benefit* Moderate to substantial
Effective vs. placebo, sham, usual care, or no treatment? No evidence
Important inconsistency? No
Comments Both trials conducted in Taiwan by same set of investigators; physical therapy comparison treatments not standardized Efficacy of acupuncture versus sham acupuncture inconsistent
Acupuncture
24 (8)
Moderate
Yes (12 trials)
Some inconsistency (versus sham acupuncture) Some inconsistency No
Direct
Fair
Back schools
26 (3)
Small
Yes (13 trials)
Direct
Fair
Back schools based on Swedish model appeared most effective Three of four trials were in workers with subacute low back pain
Brief educational interventions Dry needling Exercise Functional restoration Hydrotherapy
4 (3)
Moderate (for return to work) Moderate Moderate Moderate Moderate
Yes (3 trials versus usual care) Yes (1 trial) Yes (24 trials) Yes (7 trials) Unclear (1 trial) Yes (4 trials)
Direct
Good
2 (2) 62 (29) 12 (9) 3 (0)
No No No No
Direct Direct Direct Direct
Fair Good Fair Fair Hydrotherapy similar to landbased exercise therapy in two trials More intense interdisciplinary rehabilitation more effective than less intense interdisciplinary rehabilitation
Interdisciplinary rehabilitation
11 (2)
Moderate
No
Direct
Good
Interferential therapy
3 (1)
Unable to estimate
No evidence
No
Direct
Poor
EVIDENCE REVIEW
Intervention Low-level laser therapy Lumbar supports Massage therapy
Effective vs. placebo, sham, usual care, or no treatment? Unclear (5 trials) Unclear (1 trial) No evidence
Important inconsistency? Some inconsistency Some inconsistency Some inconsistency (versus spinal manipulation) No No
Comments Trials evaluated different types and intensity of laser, with inconsistent findings
2 (1) 4 (3)
Unable to estimate Moderate
Direct Direct
Poor Fair Some trials evaluated minimal or light massage techniques
Neuroreflexotherapy Percutaneous electrical nerve stimulation Psychological therapies
3 (2) 3 (0)
Substantial Unable to estimate
Yes (2 trials) Unclear (2 trials) Yes (11 trials)
Direct Direct
Fair Poor
All trials conducted in Spain by same investigator
35 (11)
Moderate (cognitivebehavioral treatment), substantial (progressive relaxation), unable to estimate (biofeedback), no effect (operant therapy) Not effective Moderate to substantial (for spa therapy), unable to estimate (for balneotherapy) Moderate
Some inconsistency (for EMG biofeedback)
Direct
Good (cognitivebehavioral and operant therapy)) fair (progressive relaxation), poor (biofeedback)
Shortwave diathermy Spa therapy and balneotherapy
1 (0) 5 (3)
No evidence Yes (3 trials of spa therapy)
Not applicable No
Direct Direct
Poor Fair (for spa therapy), poor (for balneotherapy) Good All trials conducted in Europe at spa resorts
Spinal manipulation
29 (15)
Yes (13 trials)
No
Direct
EVIDENCE REVIEW
Number of trials (number rated higher-quality by at least one systematic review) 3 (0) 6 (3) 9 (2)
Intervention Superficial heat Traction Transcutaneous electrical nerve stimulation (TENS) Ultrasound Yoga
Net benefit* Unable to estimate Not effective (for continuous traction) Unable to estimate
Effective vs. placebo, sham, usual care, or no treatment? Unclear (3 trials) No (2 trials) Yes (2 trials)
Important inconsistency? No No Yes (for TENS vs. sham or no treatment) Not applicable No
Directness of evidence? Direct Direct Direct
Overall quality of evidence Poor Fair Poor
Comments Three lower-quality trials
1 (0) 3 (1)
Unable to estimate Moderate (for Viniyoga)
Unclear (1 trial) No evidence
Direct Direct
Poor Fair (for Viniyoga) Insufficient evidence to judge non-Viniyoga techniques
EVIDENCE REVIEW
Table 61. Summary of evidence on non-pharmacologic therapies for radiculopathy or sciatica

Intervention Percutaneous electrical nerve stimulation Spinal manipulation Traction
Effective vs. placebo, sham, or no treatment? Unclear (1 trial)
Important Inconsistency? Not applicable
Comments
3 (0)
Moderate
No evidence
No
Direct
Fair
No clear differences compared to other interventions Other trials of traction included patients with back pain of varying duration
16 (4)
Not effective (continuous or intermittent traction); small to moderate (autotraction) Unclear
No for continuous or intermittent traction (8 trials), yes for autotraction (2 trials) Unclear (1 trial)
Some inconsistency (for autotraction versus continuous or intermittent traction)
Direct
Fair
Ultrasound
1 (0)
Not applicable
Direct
Poor
* Based on evidence showing medication is more effective than placebo, and/or evidence showing medication is at least as effective as other medications or interventions thought to be effective, for one or more of the following outcomes: pain, functional status, or work status. Versus placebo, small benefit defined as 5-10 points on a 100-point Visual Analogue Scale (VAS) for pain (or equivalent), 1-2 points on the Roland Morris Disability Questionnaire (RDQ), 10-20 points on the Oswestry Disability Index (ODI), or a standardized mean difference (SMD) of 0.20.5. Moderate benefit defined as10-20 points on a VAS for pain, 2-5 points on the RDQ, 10-20 points on the ODI, or a SMD of 0.5-0.8. Large benefit defined as >20 points on a 100 point VAS for pain; >5 points on the RDQ, >20 points on the ODI, or a SMD of >0.8. Inconsistency defined as <75% of trials reaching consistent conclusions on efficacy (no effect vs. positive effect considered inconsistent)
EVIDENCE REVIEW
Key Question 5 How effective are decision tools or other methods for predicting which patients are more likely to respond to specific therapies like spinal manipulation or different types of exercise therapy?
Results of search: We identified four systematic reviews (three rated higher-quality791-794) on the reliability and validity of physical exam maneuvers for determining whether manipulative treatments791, 792, 795 or treatments that target the sacroiliac joint794 are indicated. We identified no systematic reviews on effectiveness of decision tools, clinical prediction rules, or other methods for identifying patients more likely to respond to specific therapies. Results of search: trials The systematic reviews included no randomized trials of physical exam maneuvers for identifying manipulable low back pain or sacroiliac joint pain. From 327 potentially relevant citations, we identified one higher-quality randomized trial that prospectively evaluated how well a clinical prediction rule identified patients with back pain (any duration) more likely to respond to spinal manipulation (Table 62)796. The prediction rule was based on a previous study that identified five factors (symptom duration <15 days, Fear Avoidance Beliefs Questionnaire work subscale score <19, lumbar hypomobility, hip internal rotation range of motion >35 degrees, and no symptoms distal to the knee) associated with greater likelihood of success with spinal manipulation797. We also identified one recent, small (n=54) observational study that derived a clinical prediction rule for identifying patients likely to benefit from a stabilization exercise program798 and two recent, higher-quality trials on the effectiveness of using a patient classification system to individualize physical therapy interventions for acute or subacute low back pain799, 800.
Reliability and validity of manual spinal palpatory exam or clinical tests of the sacroiliac joint
Three systematic reviews on the reliability and validity of manual spinal palpatory maneuvers each found suboptimal evidence, poor reproducibility of examination findings, and uncertain validity for identifying manipulable conditions791, 792, 795. One systematic review found poor or inconsistent reliability for most pain provocation and mobility tests for the sacroiliac joint, though two higher-quality studies included in the review found good reliability (kappa=0.61 to 0.80) for the Gaenslen test and thigh thrust793. The same authors found estimates of diagnostic accuracy for pain provocation and mobility tests of the sacroiliac joint inconsistent and difficult to interpret due to poor methodologic quality of the studies, lack of a valid reference standard, and poor test reliability794.
Utility of clinical prediction rules for spinal manipulation

The randomized trial by Childs et al allocated patients (n=131) with a median duration of 27 days of low back pain to spinal manipulation or exercise therapy796. It applied a previously derived clinical prediction rule to all patients and prospectively evaluated whether outcomes from spinal manipulation correlated with classification of patients using the prediction rule. It
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain found treatment effects greatest in the subgroup of patients positive on the rule (met at least 4 of 5 criteria) who received manipulation. Relative to patients who were negative on the rule and received exercise, the odds ratio for a successful outcome (improvement in ODI at least 50%) in this subgroup was 60.8 (95% CI 5.2 to 704.7), compared to 2.4 (95% CI 0.83 to 6.9) for those negative on the rule who received manipulation and 1.0 (CI 0.28 to 3.6) for those positive on the rule who received exercise. Patients positive on the rule who received manipulation had a 92% chance of a successful outcome, with an associated number needed to treat for one successful outcome (relative to treatment with exercise) of 1.9 (95 % CI 1.4 to 3.5). One potential shortcoming of the prediction rule evaluated in this trial is that it may not be readily applied in everyday practice because it requires the clinician to perform and interpret potentially unfamiliar physical exam maneuvers (spinal mobility and hip range of motion tests) and administer a specific, potentially unfamiliar questionnaire (Fear-Avoidance Beliefs Questionnaire). The authors of the trial have developed a pragmatic version of the prediction rule with two factors (duration <16 days and no symptoms extending distal to the knee) that also predicted outcomes with manipulation (positive likelihood ratio=7.2, 95% CI 3.2 to 16.1 in patients meeting both criteria)801. However, this variation of the prediction rule was developed retrospectively and has not yet been prospectively validated.
Table 62. Randomized trial evaluating decision tool for predicting success from spinal manipulation
Number of patients Duration of follow-up n=131 6 months Likelihood of success at 4 weeks, relative to patients negative on rule who received exercise: Positive on rule and received manipulation OR 60.8 (5.2 to 704.7, p=0.002), negative on rule and received manipulation OR 2.4 (0.83 to 6.91), positive on rule and received exercise OR 1.0, 95% CI (0.28 to 3.6) Positive likelihood ratio for positive rule in manipulation group at predicting success at 1 week: 13.2 (3.4 to 52.1)
Author, year Childs, 2004796
Main results Manipulation + exercise vs. exercise alone "Success" at 4 weeks: 44/70 (63%) vs. 22/61 (36%)
Quality score* 7/9
This trial was designed to confirm the predictive ability of a clinical prediction rule in a setting other than the one from which it was originally derived796. One classification scheme categorizes clinical prediction rules validated in this manner as level 2802. This trial does not meet criteria for a level 1 (highest classification) clinical prediction rule, which is defined as one that has been shown to affect clinician behavior and improve outcomes. One method for demonstrating effects of this clinical prediction would be to compare clinical outcomes in patients randomized to receive the prediction rule and spinal manipulation if they met criteria for it, compared to patients who had therapy selected without the aid of the prediction rule.
EVIDENCE REVIEW
Clinical prediction rules for exercise

One study (n=54) prospectively derived a clinical prediction rule for determining which patients with low back pain are more likely to respond to a stabilization exercise program803. It found that presence of three or more of the following factors was associated with a greater likelihood of treatment success (positive likelihood ratio=4.0, 95% CI 1.6 to 10.0, negative likelihood radio=0.52, 95% CI 0.30 to 0.88): positive prone instability test, presence of aberrant movement, average straight leg raise test >91 degrees, or age <40 years. This prediction rule would be classified as level 4 (derived but not validated)802.
Patient classification systems for individualizing physical therapy interventions

One higher-quality trial compared a standardized exercise regimen (low-stress aerobic exercise, general muscle reconditioning, and advice to stay active) with an approach using a classification scheme to match patient signs and symptoms to specific exercises or other treatments (such as manipulation, mobilization, or traction) in workers with back pain for less than three weeks (Table 63)799. It found patients receiving physical therapy according to the classification scheme had greater improvements in ODI scores at 4 weeks (between-group mean difference=10.9, 95% CI 1.9 to 19.9) and at one year (mean difference=9.0, 95% CI 0.30 to 17.7), and were less likely to have continued work restrictions (42% vs. 17%, p=0.017). One difficulty in interpreting results, however, is that the intensity of the standardized exercise regimen was unclear. A second higher-quality trial also used a system to classify patients with acute or subacute low back pain into manipulation, specific exercise, and stabilization exercise subgroups800. Patients randomized to matched treatment (i.e. the treatment matched their classification) experienced slightly greater improvements in ODI scores after 4 weeks and 1 year (6-8 points) than those who received unmatched treatment, and a greater proportion had an improvement in ODI scores at least 20 points or 33% from baseline. The classification system used in this trial has not yet been validated in other settings.
EVIDENCE REVIEW
Table 63. Trial comparing standardized exercise therapy to individualized treatment based on a classification scheme
Author, year Fritz, 2003799
Brennan, 2006800
n=123 1 year
Main results Standard exercise vs. classification-based therapy (mean differences between groups relative to baseline) ODI: 10.9 (95% CI 1.9 to 19.9) at 4 weeks, 9.0 (0.30 to 17.7) at 1 year SF-36 physical component summary: 5.6 (0.6 to 10.7) at 4 weeks, 3.6 (-2.1 to 9.3) at 1 year SF-36 mental component summary: 5.7 (1.8 to 9.5) at 4 weeks, 3.6 (-1.4 to 8.7) at 1 year Continued work restrictions after four weeks: 42% (15/36) vs. 17% (7/41) "Matched" vs. "unmatched" therapy ODI, change from baseline: 29.9 vs. 23.3 at 4 weeks (p=0.03), 27.9 vs. 19.6 at 1 year (p=0.006) Proportion with improvement in ODI >20 points or at least 33%: 78% vs. 60% at 4 weeks (p=0.039)
Quality score* 7/9
5/9
Harms No trial reported adverse events. Costs The trial that compared standardized exercise therapy to classification-based treatment found higher total median costs with the former ($1,004 versus $774), though the difference was not significant (p=0.13)799.
Summary of evidence
Five systematic reviews found spinal palpatory tests for manipulable low back pain and clinical tests of the sacroiliac joint have poor or inconsistent reproducibility and uncertain validity (level of evidence: fair). For back pain of any duration, a decision tool accurately identified patients who experienced benefit from spinal manipulation. However, the tool has only been validated in one study, evidence of beneficial effects on clinical outcomes from applying the decision tool is not yet available, and the tool may not be practical for use in many primary care setting. A more pragmatic version has not yet been prospectively validated (level of evidence: fair). A decision tool for identifying patients likely to benefit from stabilization exercise has not yet been validated (level of evidence: poor). For acute low back pain, one recent, higher-quality trial found a standardized exercise regimen inferior to physical therapy tailored according to patient symptoms and physical exam findings. However, the intensity of the standardized exercise regimen in this trial was unclear (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For acute or subacute low back pain, one recent, higher-quality trial found patients randomized to a physical therapy intervention that matched their symptoms and physical exam findings had slightly superior outcomes compared to those who received an unmatched physical therapy intervention. The classification system has not yet been validated in other populations and settings (level of evidence: fair).

The UK RCGP guidelines found no firm evidence that it is possible to select which patients will respond to manipulation (strength of evidence: **). The European COST guidelines found insufficient evidence to recommend the use of spinal palpatory and range of motion tests to identify patients with manipulable lesions.
Key Question 6 How effective is referral from primary care providers to back specialty providers for improving patient outcomes? What are the outcomes for patients who are managed by different types of care providers or by multidisciplinary or interdisciplinary clinics?
Results of search: systematic reviews We found no systematic review on effects of referral by a primary care provider (defined here as a family practitioner, general internist, or general practitioner) to a non-surgical back specialist (defined here as a neurologist, rheumatologist, physiatrist, occupational medicine physician, neurologist, or pain physician) on patient outcomes. The efficacy of interdisciplinary rehabilitation, behavioral therapies, acupuncture, and spinal manipulation is reviewed in Key Question 4, and the efficacy of surgical and non-surgical invasive interventions in Key Questions 8 and 9. In general, trials focused on the intervention rather than the provider managing care, and did not specify whether patients were referred by a primary care provider, managed without a referral, or co-managed by multiple providers. Results of search: trials From 525 potentially relevant citations, we found no trial on effects of referral from primary care providers to back specialty providers on patient outcomes. One recent large, higher-quality trial (the UCLA Low Back Pain Study) evaluated chiropractic versus medical care for patients with low back pain of unspecified duration (Table 64)780, 781. We also identified one well-designed, prospective cohort study on outcomes of acute low back pain episodes in patients managed by different provider types19.not in the qual table
Efficacy of referral to back specialty providers on patient outcomes from low back pain
The UCLA Low Back Pain Study found no significant differences in pain or disability through 18 months in patients (n=339) randomized to chiropractic care versus medical care without physical therapy, with specific chiropractic interventions chosen at the discretion of the assigned providers780, 781. Adding physical therapist care (including of one or more of the following: heat or cold, ultrasound, electrical muscle stimulation, soft tissue and joint mobilization, traction,
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain and/or supervised exercises) to medical care was associated with statistically significant but small benefits in pain scores (<1 point on a 10 point scale) and the RDQ (1.7 to 2.1 points) through 18 months. However, the addition of physical therapy care was associated with substantially increased costs (average $760 vs. $369 per patient)787.
Table 64. Results of UCLA Low Back Pain Study
Number of patients Duration of follow-up n=681 for all four arms 6 months
Author, year Hurwitz, 2002780, 781
Main results Chiropractic care vs. medical care (adjusted between group difference in improvement from baseline) Most severe pain (0 to 10 scale): -0.25 (95% CI -0.96 to 0.45) at 6 months, -0.64 (95% CI -1.38 to -0.21) at 18 months Average pain (0 to 10 scale): -0.26 (95% CI -0.81 to 0.29) at 6 months, -0.50 (-1.09 to 0.08) at 18 months RDQ score (0 to 24 scale): -0.37 (95% CI -1.63 to 0.90) at 6 months, -0.69 (-2.02 to 0.65) at 18 months Medical care + physical therapist care vs. medical care alone Most severe pain: -0.61 (95% CI -1.31 to 0.10) at 6 months, -0.95 (95% CI -1.69 to -0.21) at 18 months Average pain: -0.63 (95% CI -1.19 to -0.08) at 6 months -0.76 (-1.35 to -0.17) at 18 months RDQ score: -1.78 (95% CI -3.05 to -0.51) at 6 months, -2.11 (95% CI -3.46 to -0.77) at 18 months Chiropractic care + physical modalities vs. chiropractic care Most severe pain: -0.15 (95% CI -0.85 to 0.55) at 6 months, +0.25 (-0.49 to 0.98) at 18 months Average pain: -0.26 (95% CI -0.81 to 0.29) at 6 months, +0.12 (-0.46 to 0.71) at 18 months RDQ score: +0.12 (95% CI -1.15 to +1.38) at 6 months, -0.01 (95% CI -1.35 to +1.32) at 18 months
Quality score* 7/9
A well-designed prospective observational study from North Carolina found little difference in time to functional recovery, return to work, and complete recovery in patients with acute back pain managed by primary care providers, chiropractors, or orthopedic surgeons19. Despite similar baseline pain and back-related disability, orthopedists were more likely to order CT or MRI of the spine compared to primary care providers (17% vs. 6-11%). Chiropractors saw patients an average of 9 to 13 visits for the acute back episode, compared to around 2 visits for primary care providers and orthopedists. Satisfaction with care was greater with chiropractors than with the other providers. The mean cost per episode was higher for orthopedic or chiropractic care ($611 to $783: 1993 dollars) than with primary care providers ($435 to $508). A survey of physicians from the early 1990s found that given the same clinical situations, use of diagnostic tests varied considerably among eight medical specialties (family practice, internal medicine, osteopathic general practice, physical medicine, rheumatology, neurology, orthopedic surgery, neurosurgery)15. Neurosurgeons and neurologists were more likely to order imaging studies, physiatrists and neurologists more likely to order electromyograms, and rheumatologists more likely to order laboratory tests.
EVIDENCE REVIEW
Summary of evidence
There is no direct evidence on effects of referral from primary care to back specialty providers on patient outcomes, though evidence on effects of interventions offered by specialty providers is reviewed elsewhere. For low back pain of unspecified duration, one recent large, higher-quality trial found medical care and chiropractic care associated with similar patient outcomes. Observational data also suggests no significant differences for back pain episodes managed by different provider types, though patterns of care varied (level of evidence: fair).

All guidelines recommend consideration of referral to a back specialist if low back pain is not improving despite non-invasive, usual interventions (strength of evidence: not assessed). For active duty personnel who have not improved after 4 to 6 months, the VA/DoD guidelines specifically recommend consideration of referral to the Medical Evaluation Board for possible reclassification or discharge from service (strength of evidence: not assessed).
Key Question 7 What is the diagnostic accuracy and what are the potential harms associated with invasive tests for identifying patients who may benefit from invasive procedures? How effective is prior use of these tests for selecting patients for invasive procedures in improving outcomes? Provocative discography
Provocative discography involves the injection of radiographic contrast material into the nucleus of an intervertebral disc, which may elicit pain. It is most commonly performed in patients with persistent, chronic low back pain in order to help identify those who may benefit from invasive procedures intended to treat discogenic back pain. The usefulness of provocative discography in patients with low back pain remains controversial804. Much of the debate centers on whether provocative discography is accurate for identifying painful lumbar discs, the uncertain natural history of discogram-positive low back pain (in one retrospective study, 68% of un-operated patients improved805), and whether use of provocative discography improves patient outcomes or leads to unnecessary and potentially harmful interventions. Many studies show good correlation between results of provocative discography and abnormalities on CT or MRI imaging806, 807. However, because the presence of radiographic degeneration or other abnormalities is not necessarily associated with patient symptoms, imaging is considered an inadequate reference standard for assessing diagnostic accuracy. Nonetheless, no other reliable reference standard for discogenic low back pain is available. We focused our review on several specific types of studies of provocative discography. First, we identified studies on rates of positive discography responses in populations of persons without serious back pain. Studies that addressed this type of questionDo test results in patients with the target disorder differ from those in normal people?have been categorized as the lowest level (Phase I) on a hierarchy of diagnostic research808. Because Phase II (Are
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain patients with certain test results more likely to have the target disorder than patients with other test results?) and Phase III (Does the test result distinguish patients with and without the target disorder among patients in whom it is clinically reasonable to suspect that the disease is present?) studies cannot be reliably interpreted in the absence of an appropriate reference standard, we did not review the literature comparing provocative discography to CT or MRI imaging results. However, we searched for studies that evaluated accuracy of provocative discography based on alternative reference standards. We also included studies that evaluated whether use of provocative discography to select patients for procedures intended to treat presumed discogenic back pain improves clinical outcomes compared to not using provocative discography. Such evidence addresses the highest level (Phase IV) question in the hierarchy of diagnostic research808Do patients who undergo this diagnostic test fare better in their ultimate health outcomes than similar patients who are not tested? Results of search: systematic reviews We identified two lower-quality systematic reviews on lumbar discography for low back pain806, 809 . We also included a lower-quality systematic review on risk of discitis following discography810. We excluded an earlier version807 of one of the reviews809 Results of search: primary studies The systematic reviews included a total of six higher-quality studies published since 1990 (when the Walsh criteria811 were first introduced) on rates of positive pain responses to provocative discography in patients without significant chronic back pain811-817. Four other studies included in the systematic reviews evaluated factors associated with a higher likelihood for positive pain responses in patients with chronic low back pain818-821 and one study included in the systematic reviews compared surgical outcomes in patients selected for fusion by results of discography versus those in whom fusion was performed without prior discography822. From 323 potentially relevant citations, we identified one additional study not included in the systematic reviews that evaluated positive responses to provocative discography after incorporation of low-pressure criteria in patients without significant chronic back pain816, one study that used a novel reference standard to estimate diagnostic accuracy of discography823, and one study that evaluated outcomes of surgery in patients selected for fusion based on response to temporary external transpedicular fixation with or without positive responses to provocative discography in adjacent discs824. We excluded two studies from the 1960s that reported high rates of positive provocative discography because they used outdated techniques825, 826. Quality ratings of provocative discography studies are shown in Appendix 8.
Rates of positive pain responses to provocative discography in patients without significant back pain
We included 7 studies that evaluated positive pain responses to provocative discography in patients without significant low back pain (Table 65). A study published by Walsh and colleagues in 1990 found that in ten asymptomatic, healthy young men undergoing provocative discography, none met criteria for a positive test811. By contrast, 6 of 7 (86%) of patients with low back pain for more than 6 months had a positive test. A positive test by the Walsh criteria
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain was defined as an abnormal disc in conjunction with pain rated as more severe than moderate plus pain-related behavior (at least two of the following: guarding/bracing/withdrawal, rubbing, grimacing, sighing, or verbalizing). Carragee and colleagues subsequently conducted a series of studies that evaluated the rate of positive pain responses to provocative discography (as defined using the Walsh criteria) in patients without serious back pain, including asymptomatic persons (Table 65). They found that patients with somatization or abnormal psychometric testing had high rates of positive responses (70% to 83%), as did those who were disabled (86% or 5/6) or had an active workers compensation or personal injury claim (89% or 8/9)813, 814. Patients with pain outside the back also frequently had positive results (50% or 4/8 following iliac crest harvest and 40% or 4/10 in those with neck pain following cervical surgery)814, 815. In patients with previous discectomy, positive pain responses were seen in 40% (8/20) of those with good surgical results813. More recently, investigators proposed adding pressure threshold criteria to the requirements for a positive response, to reduce potential false-positive findings827. With this adaptation, pain that is only provoked with high injection pressures (which can occur in normal discs) is not considered a positive response. In one study, 0% (0/16) of asymptomatic volunteers had a positive response when incorporating pressure criteria, compared to a 35% (100/282) rate of positive discograms in patients with chronic low back pain817. However, asymptomatic subjects in this study mainly consisted of physicians, which could limit generalizability of results828. In a re-analysis of data reported in earlier studies, Carragee and colleagues also reported no positive pain responses (0/10) in asymptomatic, low-risk patients without low back pain after incorporating pressure threshold criteria816. However, 36% (5/14) of patients without back pain but with either chronic pain or somatization, 25% (5/20) of pain-free patients following disc surgery, and 28% (7/25) of patients with mild low back pain would still be classified as having positive tests after incorporation of pressure threshold criteria.
EVIDENCE REVIEW
Table 65. Rates of positive pain responses to provocative discography in persons without serious back pain
Definition of positive pain response Walsh criteria, with added criteria of 'low pressure' response defined as pain provoked with static pressure of less than 22 psi Quality score* 8/9
Author, year Carragee, 2006816
Derby, 2005817
Carragee, 812 2002
Negative discogram=no pain described as 'familiar', no pain 6/10 at pressures 50 psi above opening pressure and 3.5 ml total injected volume Walsh criteria
Rates of positive pain responses A: No LBP, but with chronic pain or somatization: 36% (5/14); 30% (3/10) in patients with chronic pain and 50% (2/4) in patients with somatization B: No LBP, history of prior successful lumbar discectomy (n=20): 25% (5/20) C: Mild persistent low back pain but not seeking or receiving treatment for it (also s/p cervical surgery): 28% (7/25); 23% (3/13) in patients with no chronic pain and 33% (4/12) in patients with chronic pain D: No LBP, no chronic pain: 0% (0/10) A: Asymptomatic volunteers: 0% (0/16) B: Chronic low back pain with unremitting pain despite conservative treatment: 35% (100/282) of discs positive A: Patients with mild persistent low back pain but not seeking or receiving treatment for it and s/p cervical spine surgery: 36% (9/25); 23% (3/13) in patients with good cervical surgery outcomes and 50% (6/12) in patients with worst cervical surgery outcomes B: Patients undergoing discography for consideration of surgery: 73% (38/52) In group A, 5/5 (100%) of patients with daily opioid had positive discogram vs. 3/17 (18%) without opioids A: No low back pain 2 to 10 years following successful lumbar disc surgery, no depression: 40% (8/20) B: Chronic persistent or recurrent low back and leg problems 14 months to 6 years following posterior discectomy: 63% (17/27); 43% (3/7) in patients with normal psychometric scores and 70% (14/20) in those with abnormal scores A: No low back pain, status post cervical discectomy and/or fusion 2 to 4 years previously with good surgical outcomes: 10% (1/10) B: No low back pain, status post cervical discectomy and/or fusion 2 to 4 years previously with poor surgical outcomes: 40% (4/10) C: No low back pain, somatization disorder and chronic pain present: 83% (5/6) Disabled: 86% (5/6) Active workers compensation or personal injury claim: 89% (8/9) A: No low back pain, status post iliac bone graft harvesting for reasons unrelated to lumbar spine: 50% (4/8) A: Low back pain >6 months: 86% (6/7) B: No low back pain: 0% (0/10)
7/9
9/9
Carragee, 813 2000
Walsh criteria
9/9
Carragee, 814 2000
Walsh criteria
9/9
Carragee, 1999815 Walsh, 1990811
Walsh criteria Walsh criteria
8/9 8/9
*See Methods for quality criteria used to evaluate studies assessing positive rates from provocative discography
EVIDENCE REVIEW
Factors associated with higher rates of positive discography in patients with chronic low back pain
Two higher-quality studies of patients with chronic low back pain reported higher rates of positive pain responses to provocative discography in patients with abnormal psychometric testing818 or abnormal pain drawings (Table 66)821. One other study found no clear association between presence or absence of somatization disorder and positive pain responses to provocative discography, but subjects appeared more highly selected, as they had already undergone negative testing for facet joint mediated pain as well as an epidural steroid injection820. A lower-quality study reported positive pain responses in 38% (51/136) of unoperated discs in patients with chronic low back pain following lumbar surgery, though the rate was higher in previously operated discs (72% or 73/102)819.
Table 66. Trials evaluating predictors of positive pain responses to provocative discography in patients with chronic back pain
Definition of positive pain response NASS criteria Quality score* 5/9
Author, year Manchikanti, 2001820
Heggeness, 1997819 Block, 1996818
Reproduction of patient's typical pain pattern Similar or exact pain reproduction
Rates of positive pain responses A: Low back pain, negative testing for facet joint mediated pain and epidural steroids, with somatization disorder: 48% (12/25) B: Low back pain, negative testing for facet joint mediated pain and epidural steroids, without somatization disorder: 56% (14/25) A: Postoperative disks: 72% (73/102) B: Unoperated disks: 38% (51/136) A: Low back pain, with at least 1 nondisrupted disc: 47% (34/72) Discordant pain response associated with higher scores on hysteria and hypochondriasis subscales of MMPI A: Low back pain with abnormal pain drawing: 50% (18/36) B: Low back pain with normal pain drawing: 12% (13/105)
2/9 7/9
Ohnmeiss, 1995821
Similar or exact pain reproduction
5/9
*See Methods for quality criteria used to evaluate studies assessing positive rates from provocative discography
Estimating accuracy of provocative discography

One recent, higher-quality prospective cohort study by Carragee and colleagues (2006) attempted to estimate the positive predictive value of provocative discography by comparing the rate of successful surgical outcomes in patients with presumed discogenic pain by provocative discography relative to the rate of successful surgical outcomes in patients with single-level, unstable spondylolisthesis (a condition for which surgery is widely considered appropriate) (Table 67)823. Patients in the provocative discography group (n=32) were selected if they met low-pressure criteria for a positive response at a single level, failed conservative therapy, had negative facet joint and sacroiliac joint blocks, and had no other spinal or pelvic pathology or comorbidities associated with poorer surgical outcomes. Patients in the spondylolisthesis group (n=34) also had no comorbidities and had single-level Grade I or II isthmic spondylolisthesis of either L5-S1 or L4-L5 with radiologic segmental instability. Patients appeared well-matched on baseline demographics, pain scores, functional status, and other important covariates.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The rate of highly successful outcomes two years following spinal fusion was 72% (23/32) in the spondylolisthesis group compared to 27% (8/30) in the positive discography group (p=0.0004). The proportion of patients who met criteria for minimal acceptable outcomes as assessed by blinded and independent observers was 91% (29/32) in the spondylolisthesis group compared to 43% (13/30) in the positive discography group. The positive predictive value (rate of success in the positive discography group relative to rate of success in the spondylolisthesis group) was 42% to 43% for both outcomes. Using the most favorable assumptions about dropouts (2 dropouts in discogenic pain group considered successes and 2 dropouts in spondylolisthesis group considered failures), the positive predictive value of discography would be 55% to 57%.
Table 67. Study evaluating rates of successful surgical outcomes in highly selected patients with positive discography relative to patients with isthmic spondylolisthesis
Author, year Carragee, 2006823
Main results Surgery for presumed discogenic pain (positive discography) vs. unstable single-level isthmic spondylolisthesis "Success" (pain VAS 2/10, ODI 15, no opioid or daily analgesic use, return to full employment): (27% (8/30) vs. 72% (23/32) Minimal acceptable outcome (pain VAS <4/10, ODI <30, no opioid use, return to at least partial employment): 43% (13/30) vs. 91% (29/32 ) Pain VAS <2 (0 to 10 scale): 30% (9/30) vs. 84% (27/32) ODI score <15: 33% (10/30) vs. 72% (23/32) No opioid or daily analgesic: 30% (9/30) vs. 88% (28/32) Working in usual occupation: 30% (9/30) vs. 81% (26/32) Positive predictive value (positive outcome in discography group relative to spondylolisthesis group: 42% for success, 43% for minimal acceptable outcome)
Quality score* 6/9
Although this study met criteria for a higher-quality prospective cohort study, the reference standard is quite atypical because it compares outcomes following the same surgical procedure in patients with two different underlying conditions (rather than comparing results to a reference test in the same set of patients). Interpretation of positive predictive value estimates from this study depends on the key assumptions that surgical morbidity should be similar in both groups and that surgery for true discogenic pain should achieve similar outcomes as surgery for unstable spondylolisthesis in a matched group of patients without risk factors for poor surgical outcomes. A potential alternative interpretation of study results is that even though surgery for discogenic pain identified by provocative discography is associated with a lower rate of success compared to surgery for unstable spondylolisthesis in highly selected patients without comorbidities, this observation could reflect an imperfect treatment rather than an incorrect diagnosis. However, the authors of the study argue that surgical removal of the disc and annulus (the presumed pain generators) should be the definitive treatment if the disc is the true source of pain823.
EVIDENCE REVIEW
Effects of provocative discography for selecting patients for spinal fusion on clinical outcomes
One lower-quality observational study compared outcomes in patients selected for spinal fusion based on positive discography to those who underwent surgery without prior discography (Table 68)822. It was rated lower-quality because it used a historical control group, did not describe independent or blinded assessment of outcomes, and did not adjust for baseline differences or confounders. It found that after 2.4 to 2.8 years of follow-up, there were no significant differences in rates of satisfactory outcomes (defined as score of <40 on ODI), pain, or psychologic testing. Another lower-quality observational study found that in patients who underwent fusion based on a positive temporary external transpedicular fixation trial, the likelihood of a successful outcome was not associated with presence of or absence of a positive response to provocative discography in adjacent disc segments824.
Table 68. Study of outcomes in patients selected for spinal fusion with or without provocative discography
Number of patients Duration of follow-up n=82 Mean 80 months Madan, 2002822 n=73 2.4 to 2.8 years
Author, year Willems, 2007824
Main results Positive provocative discography in adjacent disc vs. no positive provocative discography in adjacent disc Fusion successful (>30% improvement in pain score): 45% vs. 45%, p=0.58 Discography vs. no discography "Excellent" or "good" ODI outcome: 81% vs. 76% "Excellent" ODI outcome: 62% vs. 58% ODI (mean scores): 34 vs. 34 Psychologic (mean scores): 22 vs. 15 Pain (VAS, 0-10): 4.2 vs. 4.4 Core set of surgical outcomes (range 10 to 50): 24 vs. 25
Quality score* 4/9
4/9
* Excludes criteria involving blinding of patients and care providers, for maximum score of 9
Two other retrospective cohort studies were excluded because they didnt compare outcomes in patients who did and did not undergo discography prior to surgery. One found successful surgery more likely in patients with positive discography and an abnormal MRI compared to positive discography and a normal MRI (75% vs. 50%)829. The other found success rates higher with abnormal discs and positive pain provocation compared to patients with abnormal discs and no pain provocation (88% vs. 52%)830. Both studies failed to report independent or blinded assessment of outcomes and did not adjust for baseline differences or potential confounders. Harms The most common serious complication following discography is discitis. In one systematic review of observational studies, 12 cases of discitis occurred in 5,091 patients (13,205 disc injections) who underwent discography without prophylactic antibiotics (mean 0.24% using the number of patients as the denominator and 0.09% using the number of disc injections as the denominator)810. In the single study of patients who received prophylactic antibiotics (200
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain patients, 435 discs), no cases were reported831. Other rare complications that have been reported after discography include disc herniation after injection, retroperitoneal hemorrhage, and dural penetration810. Increased pain following the procedure is frequent but usually transient. However, one small study found that 20% to 67% of patients without back pain but with somatization or chronic pain at other sites reported persistent back pain one year after provocative discography814. Long-term effects of discography have not been well-studied, though one small study (n=36) found no increase in degenerative disc changes 10 to 20 years after discography832. Costs We found no studies evaluating costs.
Summary of evidence
In healthy, asymptomatic volunteers, positive responses to provocative discography were uncommon in several series of patients (level of evidence: fair). In patients without significant back pain, provocative discography was frequently and consistently associated with high rates of positive pain responses in patients with chronic pain at other sites, those with somatization, those with previous disc surgery, and those disabled or seeking monetary compensation (level of evidence: fair). In higher-risk subgroups of patients without significant low back pain (see above bullet), incorporation of pressure criteria into the definition for a positive response to provocative discography did not eliminate positive results in one small study (level of evidence: fair). In patients with chronic low back pain, previous back surgery, chronic pain, and abnormal psychometric testing were associated with increased rates of positive discography in several series of patients (level of evidence: fair). In patients without risk factors for poor surgical outcomes, one higher-quality cohort study found that relative to the rate of successful surgery for single-level isthmic spondylolisthesis, the rate of successful surgery for presumed discogenic back pain (based on provocative discography) was 43-44%. Interpretation of this finding as a positive predictive value depends on the critical assumptions that surgical morbidity and rates of successful surgery for presumed discogenic back pain should be equivalent to rates of successful surgery for isthmic spondylolisthesis if the disc is the true source of symptoms (level of evidence: fair). In patients who underwent spinal fusion, one lower-quality observational study found surgery outcomes similar with or without the use of provocative discography to select patients for surgery (level of evidence: poor). In patients who underwent spinal fusion based on results of a temporary external transpedicular fixation trial, one lower-quality observational study found presence of or absence of a positive response to provocative discography in adjacent discs did not predict clinical outcomes (level of evidence: poor).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Discitis following provocative discography appears rare with or without antibiotics. Other serious adverse events also appear rare. In one study, persistent pain was reported in patients with somatization or chronic pain outside the back, but no back pain at the time of provocative discography, but who had somatization or chronic pain at other sites (level of evidence: fair).

The AHCPR guidelines recommend against discography in patients with acute low back pain because it is invasive and interpretation is equivocal (strength of evidence: C). The AHCPR guidelines recommend against CT-discography over MRI or CT for assessing patients with suspected nerve root compression due to lumbar disc hernia (strength of evidence: C). The European COST guidelines recommend against discography for diagnosis of discogenic pain in patients with chronic low back pain.
Diagnostic selective nerve root block

Diagnostic selective nerve root blocks involve the injection of local anesthetic around spinal nerves under fluoroscopy. A positive response is defined as relief of usual radicular symptoms and is thought to indicate that the target nerve is the source of those symptoms. Results of selective nerve root blocks correlate well with radiologic or surgical evidence of nerve compression833. However, because nerve root compression can usually be identified by noninvasive imaging, the main roles of diagnostic nerve root blocks are to evaluate the appropriate target level for interventions when multiple nerve roots are involved or to confirm radiculopathy when imaging is equivocal or when there is discordance between clinical findings and imaging. No reliable reference standard (such as electrophysiologic testing329) is available for estimating diagnostic accuracy of selective nerve root blocks for identifying true nerve root pain in these situations. We therefore focused our review on evidence on whether use of selective nerve root blocks to select patients for procedures intended to relieve nerve root compression improves clinical outcomes compared to not using selective nerve root blocks to select patients (Phase 4 evidence on the diagnostic research hierarchy808). Results of search: systematic reviews We identified one lower-quality systematic review on diagnostic accuracy of selective nerve root blocks834. However, it included no studies that evaluated whether use of diagnostic selective nerve root blocks to identify patients for procedures intended to relieve nerve root compression improves clinical outcomes compared to relying only on imaging or other non-invasive diagnostic methods to select patients. We excluded an earlier version of the systematic review833. Results of search: other studies From 381 potentially relevant citations, we identified no relevant studies.
EVIDENCE REVIEW
Effects of selective nerve root block on clinical outcomes

We could not assess effects on clinical outcomes of using selective nerve root blocks to select or guide procedures for relieving nerve root compression compared to using non-invasive diagnostic methods alone (no evidence). Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
There are no studies on how use of diagnostic selective nerve root blocks to evaluate patients for suspected nerve root compression affects choice of therapy and clinical outcomes compared to use of non-invasive methods alone.

The other guidelines do not address diagnostic selective nerve root blocks.
Diagnostic intra-articular facet joint block and medial branch block

Diagnostic intra-articular facet joint blocks involve the injection of local anesthetic under fluoroscopic guidance into the facet (zygapophysial) joints. Medial branch blocks involve injection of local anesthetic around the medial branches of the dorsal rami, which innervate the facet joints. Both procedures are performed in order to evaluate whether the facet joint is the source of low back pain. A positive response to either diagnostic procedure is defined as the relief of usual back pain. In a number of studies, positive intra-articular facet joint blocks and medial branch blocks have been reported in 15% to 45% of patients with chronic low back pain835. Use of control blocks can reduce the rate of positive responses by up to 50% compared to relying on a single block. However, as in other invasive diagnostic procedures for low back pain, no reliable reference standard for facet joint pain is available to estimate the diagnostic accuracy of intra-articular facet joint blocks and medial branch blocks. It is therefore unknown whether the decreased rate of positive responses is due to fewer false positives, fewer true positives, or some combination. Furthermore, results of intra-articular facet joint blocks and medial branch blocks do not correlate well with findings on imaging studies. We therefore focused our review on evidence that evaluated whether use of intra-articular facet joint blocks or medial branch blocks to select patients for procedures intended to treat presumed facet joint pain improves clinical outcomes compared to not using facet joint blocks to select patients (Phase 4 evidence on the diagnostic research hierarchy808). Results of search: We identified two systematic reviews on diagnostic utility or accuracy of intra-articular facet joint blocks and medial branch blocks836, 837. Neither included any study that evaluated whether use of intra-articular facet joint blocks or medial branch blocks to select patients for procedures
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain intended to treat presumed facet joint pain improves clinical outcomes compared to relying on other methods to select patients such procedures. We excluded an earlier version835 of one of the systematic reviews837. Results of search: primary studies From 46 potentially relevant citations, we identified one lower-quality trial not included in the systematic reviews that evaluated clinical outcomes in patients selected for percutaneous facet joint cryodenervation based on a positive uncontrolled medical branch block versus those selected based on a positive uncontrolled pericapsular block838.
Effects of facet joint block on clinical outcomes

One lower-quality trial found no clear differences in pain relief between patients selected for percutaneous facet joint cryodenervation based on a positive uncontrolled medial branch block, versus those selected based on a positive uncontrolled pericapsular block (Table 69)838. However, results are difficult to interpret because efficacy of facet joint cryodenervation has not been evaluated in randomized trials. We found no trials that assessed whether use of intraarticular facet joint blocks or medial branch blocks to select patients for facet joint interventions evaluated in randomized trials (e.g., radiofrequency denervation or facet joint steroid injection) improves clinical outcomes compared to selection of patients based on non-invasive methods alone.
Table 69. Trial of outcomes in patients selected for percutaneous facet joint cryodenervation with uncontrolled medial branch versus uncontrolled pericapsular block
Author, year Birkenmaier, 838 2007
Main results Facet joint cryodenervation based on positive uncontrolled medial branch block versus cryodenervation based on positive pericapsular block Mean pain score (0 to 10 scale): 4.2 vs. 3.2 at 2 weeks (p=0.33), 4.2 vs. 2.3 at 3 months (p=0.049), 4.0 vs. 2.7 at 6 months (p=0.148) Percent improvement in pain score: 51% vs. 44% at 2 weeks (p=0.61), 40% vs. 33% at 6 months (p=0.52) McNab score (0 to 3): 1.3 vs. 1.7 at 2 weeks, 1.5 vs. 2.0 at 6 months
Quality score* 4/10
* Excludes criteria involving blinding of care providers, for maximum score of 10
Summary of evidence
For presumed facet joint pain, one lower-quality trial found no clear differences in pain relief between patients selected for percutaneous facet joint cryodenervation based on a positive
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain uncontrolled medial branch block, versus those selected based on a positive uncontrolled pericapsular block (level of evidence: poor) There are no other studies on how use of facet joint injections or medial branch blocks to evaluate patients for facet joint pain affects choice of therapy and clinical outcomes compared to use of non-invasive methods alone. Evidence on interventions for treating presumed facet joint pain is outlined in Key Question 7. In all trials of facet joint interventions, patients were enrolled based on positive (primarily uncontrolled) diagnostic intra-articular facet joint blocks or medial branch blocks.

The European COST guidelines recommend against facet joint blocks for the diagnosis of facet joint pain.
Diagnostic sacroiliac joint block

Diagnostic sacroiliac joint injection involves the injection of local anesthetic into or around the sacroiliac joint in order to evaluate whether the sacroiliac joint is the source of low back pain. A positive response is defined as the relief of usual back pain. Rates of positive intra-articular sacroiliac joint blocks range from 2% to 27%, depending in part on the population evaluated and the method of block used (e.g. controlled or uncontrolled, fluoroscopic guidance or no fluoroscopic guidance)83. However, as in other invasive diagnostic procedures for low back pain, no reliable reference standard for sacroiliac pain is available for estimating the diagnostic accuracy of sacroiliac joint blocks. Furthermore, results of sacroiliac joint blocks may not correlate well with findings on imaging studies. We focused our review on studies that evaluated whether use of sacroiliac joint blocks to select patients for procedures intended to treat presumed sacroiliac joint pain improves clinical outcomes compared to other methods to select patients (Phase 4 evidence on the diagnostic research hierarchy808). Results of search: systematic reviews We identified one systematic review on diagnostic utility or accuracy of sacroiliac joint blocks83. It did not include any study that evaluated whether use of a diagnostic sacroiliac joint block to select patients for procedures intended to treat presumed sacroiliac joint pain improves clinical outcomes compared to relying on other methods to select patients for such procedures. We excluded an earlier version186 of this systematic review. Results of search: primary studies From 46 potentially relevant citations, we identified no relevant trials.
Effects of diagnostic sacroiliac joint block on selection of therapies and clinical outcomes
We found no studies that assessed whether use of diagnostic sacroiliac joint blocks to select patients for interventions targeting the sacroiliac joint improves clinical outcomes compared to selection based on non-invasive methods alone.
EVIDENCE REVIEW
Summary of evidence
There are no studies on how use of sacroiliac joint blocks to evaluate patients for sacroiliac joint pain affects choice of therapy and clinical outcomes compared to use of non-invasive methods alone. Evidence on interventions for presumed sacroiliac joint pain is reviewed in Key Question 7.

The other guidelines do not address diagnostic sacroiliac joint blocks.
Key Question 8 How effective are injections (and different injection interventions) and other interventional therapies for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? Injections outside the spine Local injections
Local injections involve the placement of a local anesthetic (with or without corticosteroid) into the muscles or soft tissues of the back via a catheter. We defined trigger point injections, a type of local injection, as an injection performed at a tender area with a palpable nodule or band. Results of search: systematic reviews We identified one higher-quality Cochrane review94 and one lower-quality systematic review92 on local or trigger point injections. We excluded two earlier versions of the Cochrane review86, 187 . Results of search: trials Four randomized trials131, 140, 142, 165 of local injections were included in the two systematic reviews.92. We rated two trials higher-quality131, 140. We identified no other trials of local injections for low back pain that met inclusion criteria. We excluded one non-randomized trial of local injections200.
Efficacy of local injections versus placebo injection

For non-specific subacute or chronic low back pain, three small (n=15 to 41), placebo-controlled trials (one rated higher-quality131) each found local anesthetic injections superior to placebo for short-term (1 to 2 weeks post-injection) pain relief for subacute or chronic back pain (Table 70)131, 142, 165. However, the trials evaluated heterogeneous injection methods and patient populations. One trial evaluated a local anesthetic injection over the iliac crest for iliac crest
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain pain131, one evaluated local anesthetic plus corticosteroid injections over the iliolumbar ligament for non-specified low back pain165, and one evaluated local anesthetic trigger point injections for lumbar or cervical (2 of 15 patients) myofascial pain syndrome142. None evaluated longer-term outcomes. A higher-quality systematic review also found no strong evidence to support local injections.94.
Table 70. Randomized, placebo-controlled trials of local injections
Sample size Type of LBP Duration of symptoms n=41 Iliac crest syndrome Duration not reported Garvey, 1989140 n=63 Non-specific low back pain At least subacute Hameroff, 1981
142
Author, year Collee, 1991131
Duration of follow-up 2 weeks
2 weeks
n=15 Myofascial low back pain Duration unclear
1 week
Sonne, 1985165
n=30 Non-specific low back pain At least subacute
2 weeks
Main results Iliac crest local anesthetic vs. saline injection Pain score: 30.5 vs. 43.8 at 2 weeks, p<0.05 Improved or much improved (patient rated): 52% vs. 30%, NS Morning stiffness and medication use: No differences (data not reported) Trigger point injection with lidocaine vs. trigger point injection with lidocaine plus corticosteroid vs. dry needle-stick vs. topical ethyl chloride plus acupressure Proportion improved: 31% (4/13) vs. 36% (5/14) vs. 55% (11/20) 50% (8/16) (p=0.09 for trigger point groups vs. other groups) Trigger point injection with bupivacaine vs. etidocaine vs. saline (mean percent improvement from baseline at 7 days, p values vs. saline) Average pain (0 to 100 VAS): -7% (p=0.005) vs. -12% (p=0.001) vs. +13% % time pain felt (0 to 100 VAS): -3% (NS) vs. -5% (NS) vs. +7% Effect of pain on activity (0 to 100 VAS): -3% (NS) vs. -11% (NS) vs. +5% Effect of pain on sleep (0 to 100 VAS): -1% (NS) vs. -10% (NS) vs. +2% Effect of pain on mood (0 to 100 VAS): +2% (NS) vs. -11% (p=0.026) vs. +9% Iliolumbar ligament steroid/local anesthetic vs. saline injection Good or excellent improvement (patient rated): 64% (9/14) vs. 20% (3/15), p<0.05
Quality score* 7/11
8/11
2/11
4/11
Efficacy of local injection versus epidural steroid injections

Efficacy of trigger point injections versus epidural steroid injections is reviewed in the section on epidural steroids.
Efficacy of local injection versus dry acupuncture needlestick or topical ethyl chloride plus acupressure
For subacute or chronic back pain without sciatica, one higher-quality trial found no significant differences between trigger point injections (with steroid, lidocaine, or both) and a single dry
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain acupuncture needlestick (RR=1.47, 95% CI 0.74 to 2.92) or topical ethyl chloride plus 20 seconds of acupressure (RR=1.71, 95% CI 0.71 to 4.14) (Table 71)140. Interpretation of these results is a challenge because the comparator interventions could be considered active treatments.
Efficacy of trigger point injection with a local anesthetic versus a steroid

For acute or subacute sciatica, one higher-quality trial found that the addition of a corticosteroid to a local anesthetic trigger point injection was no better than a local anesthetic alone (proportion improved 45% with corticosteroid versus 40% without corticosteroid)140. Harms One placebo-controlled trial reported no adverse events142 and another didnt report adverse events165. In two other trials, adverse events following local injections included pain at the injection site, temporary paresthesia, and nausea (in both active and control injection groups)131, 140 . Costs We found no studies evaluating costs.
Summary of evidence
For acute or subacute sciatica, adding a steroid to a local anesthetic for trigger point injections did not result in superior outcomes compared to a local anesthetic alone in one lower-quality trial (level of evidence: poor). For subacute or chronic low back pain, three small, lower-quality trials found local or trigger point injections with a local anesthetic superior to saline injection for short-term pain relief. The trials evaluated heterogeneous injection methods, and none evaluated longer-term outcomes (level of evidence: poor). There is no evidence on efficacy of local injections for long-term pain relief. For subacute or chronic low back pain without sciatica, one lower-quality trial found no significant differences between a trigger point injection with local anesthetic (with or without a steroid) and either a single dry needlestick or topical ethyl chloride plus acupressure (level of evidence: poor). One lower-quality trial found no difference between trigger point injection with a local anesthetic plus corticosteroid versus a local anesthetic alone (level of evidence: poor). See section on epidural steroids for summary of evidence on local or trigger point injections versus epidural steroids.

The AHCPR guidelines recommend against invasive trigger point injections in the treatment of patients with acute low back problems (strength of evidence: C). The VA/DoD guideline recommendations are similar.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The UK RCGP guidelines found that studies of trigger point injections included patients with chronic low back pain, and findings were equivocal, with little evidence specifically in acute low back pain patients (strength of evidence: *). The European COST guidelines found insufficient evidence to recommend trigger point injections for chronic low back pain.
Botulinum toxin
Botulinum toxin is a product of the bacterium Clostridum botulinum that has anti-spasmodic activity. Injections of botulinum toxin have been shown to reduce pain associated with movement disorders and certain painful conditions. Two antigenically distinct serotypes of botulinum toxin (A and B) are available for use in clinical practice. Results of search: systematic reviews We excluded one review on botulinum toxin for low back pain because it did not report systematic methods175. We identified no other systematic reviews of botulinum toxin for low back pain. Results of search: trials We identified one small (n=31), higher-quality randomized trial105 of botulinum toxin injection for chronic low back pain met inclusion criteria. We excluded one non-randomized trial207 and two trials202, 204 of botulinum toxin for neck pain.
Efficacy of botulinum toxin versus saline injection or no injection

For chronic low back pain failing to respond to standard treatments, a small (n=31), higherquality RCT found botulinum toxin A superior to saline injection for short-term pain relief (proportion of patients with >50% pain relief 73% vs. 25% at 3 weeks, p=0.012, and 60% vs. 12.5% at 8 weeks, p=0.09) and for rates of improvement in ODI scores (67% vs. 19%, p=0.011) (Table 83)105. Effects were no longer present in most (60%) of responders after three to four months.
Table 71. Randomized, placebo-controlled trial of botulinum toxin injection
Sample size Type of LBP Duration of symptoms n=31 Non-specific low back pain Chronic
Author, year Foster, 2001105
Main results Botulinum toxin A vs. saline injection Degree of pain relief >50%: 73% (11/15) vs. 25% (4/16) at 3 weeks (p=0.012), 60% (9/15) vs. 12.5% (2/16) at 8 weeks (p=0.009) ODI, proportion with improvement at 8 weeks: 67% (10/15) vs. 19% (3/14) (p=0.011) 6/10 responders in botulinum toxin A group reported cessation of analgesic effect after 3 to 4 months
Quality score* 9/11
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Harms No side effects were reported in one randomized trial105. A case report exists of fatal anaphylaxis following injection of botulinum toxin A for chronic neck and back pain839. Costs We found no studies evaluating costs.
Summary of evidence
There is no evidence on effectiveness of botulinum toxin injection for acute low back pain. For chronic low back pain, a single, small, higher-quality trial found botulinum toxin injection moderately superior to saline injection for short-term pain relief and functional status in patients who failed to respond to standard treatments, though effects were no longer present in most responders after 3 to 4 months (level of evidence: poor). There is no evidence comparing botulinum toxin injection to other interventions. There is insufficient evidence to evaluate harms of botulinum toxin in patients with low back pain, though one case of fatal anaphylaxis has been reported.

The other guidelines do not address botulinum toxin.
Prolotherapy
Prolotherapy (also referred to as sclerotherapy) is a technique that involves the repeated injection of irritants into ligaments and tendinous attachments in order to trigger an inflammatory response. This theoretically leads to subsequent strengthening of ligaments and improvement in pain and disability. Prolotherapy injections are often supplemented by co-interventions such as trigger point injections, manipulation, and exercises that are thought to enhance the effectiveness of treatment or address underlying dysfunction contributing to the back pain. Because of the irritant nature of prolotherapy injections, patients are expected to experience transient pain at the injection site after receiving treatment. Results of search: systematic reviews We identified one recent, higher-quality Cochrane review on efficacy of prolotherapy for chronic low back pain76. We excluded an earlier version of the Cochrane review197 and a review that did not clearly use systematic methods182. Results of search: trials We identified five randomized, placebo-controlled trials of prolotherapy134, 146, 151, 156, 168. All were included in the Cochrane review76. We rated four of the five trials higher-quality134, 146, 156, 168.
For chronic non-specific low back pain, three trials (two higher-quality134, 168) found no difference between prolotherapy and either saline or local anesthetic control injections for short- or longterm (up to 24 months) pain or disability
Efficacy of prolotherapy versus control injection
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain (Table 72)151. One higher-quality trial found prolotherapy associated with increased likelihood of short-term improvement in pain or disability versus control injection (RR=1.47, 95% CI 1.04 to 2.06), but both treatment groups received a number of co-interventions including spinal manipulation, local injections, exercises, and walking146. In the fifth trial, effects of prolotherapy could not be determined because the prolotherapy group received strong manipulation and the control injection group only light manipulation156. A higher-quality Cochrane review rated all five placebo-controlled trials higher-quality76. It also found prolotherapy to be ineffective when used alone for chronic low back pain.
Table 72. Randomized, placebo-controlled trials of prolotherapy
Sample size Duration of symptoms n=74 Chronic Non-specific low back pain Klein, 1993146 Non-specific low back pain Mathews, 151 1987 Non-specific low back pain Ongley, 1987156 Non-specific low back pain n=82 Chronic 6 months n=80 Chronic 6 months Duration of follow-up 6 months
Author, year Type of LBP Dechow, 1999134
n=22 Chronic
Up to 1 year
Yelland, 168 2004 Non-specific low back pain
n=110 Chronic
2 years
Main results Prolotherapy vs. saline/lignocaine injection, mean scores at 6 months, estimated from graphs) Pain (0 to 10): 5.2 vs. 4.4, NS ODI (0 to 100): 36 vs. 35, NS Zung Depression Score: 34.2 vs. 37.0, NS Present Pain Score: 1.9 vs. 1.9, NS Prolotherapy vs. saline/lidocaine injection >50% improvement in pain score: 77% (30/39) vs. 52% 21/40), p=0.04 RDQ (mean score at 6 months): 4.04 vs. 4.38, p=0.07 Pain score (0 to 8 VAS): 2.29 vs. 2.85, p=0.06 Prolotherapy vs. tender point local anesthetic injection Proportion recovered (pain score 5 or 6 on a 1 to 6 scale): 63% (10/16) vs. 33% (2/6) at 3 months (NS), no significant differences at 6 or 12 months. No further pain: 12% (2/16) vs. 17% (1/6) up to 1 year Prolotherapy plus forceful manipulation vs. saline injection plus non-forceful manipulation >50% improvement in disability score: 88% (35/40) vs. 39% (16/41) at 6 months, p<0.003 Disability score (mean, 0 to 33 scale): 8.37 vs. 4.00 at 1 month (p<0.001) < 8.29 vs. 3.43 at 6 months (p<0.001) Pain score (mean, 0 to 7.5 scale): 3.06 vs. 2.13 at 1 month (p<0.01), 3.08 vs. 1.50 at 6 months (p<0.001) Prolotherapy vs. saline injection (mean change from baseline, positive values indicate improvement) Pain intensity (0 to 100 VAS): 18.6 vs. 18.4 at 1 year (p=0.96), 18.4 vs. 16.4 at 2 years (p=0.93) Modified RDQ (0 to 23): 5.5 vs, 4.5 at 1 year (p=0.85), 4.9 vs. 4.2 at 2 years (p=0.60) Analgesic use (0 to 4): -0.1 vs. -0.1 at 1 year (p=0.60) Days of reduced activity in last 28 days: 3.2 vs. 2.4 at 1 year (p=0.66), 2.5 vs. 1.8 at 2 years (p=0.75) SF-12 Physical Component Summary score: 5.5 vs. 6.0 at 1 year (p=0.76), 1.4 vs. 3.3 at 2 years (p=0.30) SF-12 Mental Component Summary score: 0.6 vs. -0.2 at 1 year (p=0.75), -0.8 vs. 1.1 at 2 years (p=0.48)
Quality score* 9/11
9/11
4/11
6/11
11/11
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Harms In three146, 156, 168 of the four trials, nearly all participants experienced the expected temporary increases in back pain and stiffness following prolotherapy injections. In two other trials, either no adverse events151 or only a few cases of post-injection pain134 were reported. Post-injection headaches suggestive of lumbar puncture occurred in two to four percent of patients in two trials146, 168. Other adverse events included postmenopausal spotting, leg pain (attributable to herniated disc), diarrhea, and other, generally transient symptoms, but there were no significant differences in any adverse event between treatment and control groups. Costs We found no studies evaluating costs.
Summary of evidence
There is no evidence on efficacy of prolotherapy for acute low back pain. For chronic low back pain, three of four higher-quality trials found no differences between prolotherapy versus control injections. The fourth trial found prolotherapy more effective than control injections, but is difficult to interpret because both groups received a number of cointerventions (level of evidence: good). Serious adverse events have not been reported following prolotherapy treatments, though nearly all patients in most trials report increased back pain due to the irritant nature of the injections (level of evidence: fair).

The AHCPR guidelines recommend against ligamentous and sclerosant injections for patient with acute low back problems (strength of evidence: C). The VA/DoD and UK RCGP guideline recommendations are similar. The European COST guidelines recommend against injections of sclerosants (prolotherapy) for nonspecific chronic low back pain.
Intraspinal steroid injections and chemonucleolysis Epidural steroid injection

Epidural steroid injections involve administration of corticosteroids via a catheter into the space between the dura and the spine. The anti-inflammatory effects of the steroid are thought to reduce swelling and pain. Epidural steroid injections have been used for sciatica, spinal stenosis, and non-specific low back pain. Epidural injections can be performed by the translaminar approach (via the interlaminar space in the spine), the transforaminal approach (through the neuroforamen ventral to the nerve root), or the caudal approach (through the sacral hiatus at the sacral canal). Results of search: systematic reviews We identified four higher-quality84, 86, 94, 100 and five lower-quality systematic reviews 71, 74, 77, 92, 97 of epidural steroids for low back pain. We excluded eleven outdated or already updated
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain systematic reviews170, 172, 178, 183, 184, 187, 189, 190, 193, 194, 386 and three reviews that were not clearly systematic173, 181, 188. Results of search: trials We identified forty randomized trials (reported in 39 articles) of epidural steroid injections for low back pain101, 102, 106, 107, 111, 115, 123, 125, 126, 128, 130, 132, 135, 143, 145, 147, 148, 151, 152, 158, 159, 161, 164, 167, 169, 840-853. 33 trials were included in at least one of the nine systematic reviews71, 74, 77, 84, 86, 92, 94, 97, 100 and we identified seven additional trials101, 102, 106, 107, 111, 115, 842. 21 trials (with two trials reported in one article148) were placebo-controlled106, 123, 125, 126, 128, 130, 132, 135, 143, 145, 147, 148, 151, 152, 158, 159, 161, 164, 167, 169 . We rated eleven placebo-controlled trials higher-quality106, 115, 123, 128, 130, 135, 145, 152, 159, 167, 843 .
Efficacy of translaminar or caudal epidural steroid injection versus epidural placebo injection (saline or local anesthetic) for low back pain with sciatica or radiculopathy
For low back pain with radiculopathy, we found inconsistent results for short-term (up to one month following injection) benefits from 20 placebo-controlled trials of epidural steroid injection Table 73)123, 125, 126, 128, 130, 132, 135, 143, 145, 147, 148, 151, 152, 158, 159, 161, 164, 167, 169. Ten of 17 trials (including three of seven higher-quality trials) showed no differences in pain or function between epidural steroid and placebo injection123, 125, 126, 128, 130, 132, 135, 143, 145, 147, 151, 152, 158, 161, 164, 167, 169. Results were more consistent after trials were stratified according to whether the control intervention was an epidural or non-epidural (soft tissue) injection. Five123, 135, 143, 158, 167 of six151 trials found an epidural steroid injection associated with short-term benefits compared to a nonepidural (primarily interspinous ligament) placebo injection, including all three higher-quality trials123, 135, 167. Only two126, 128 of eleven125, 130, 132, 145, 147, 152, 161, 164, 169 trials found an epidural steroid injection associated with short-term benefits compared to epidural placebo (saline or local anesthetic). One of the positive trials was rated higher quality128; both were small (n=23 and 35) trials of caudal epidural injections. Three other trials reported mixed or unclear results145, 147, 161. Stratification of trials according to duration of symptoms, use of imaging to confirm presence of prolapsed disc, or study quality did not appear to reduce inconsistency in short-term findings.
EVIDENCE REVIEW
Table 73. Randomized, placebo-controlled trials of epidural steroid injection for radiculopathy
Sample size Type of LBP Duration of symptoms n=228 Radiculopathy Subacute or chronic
Author, year Arden, 2005123
Duration of follow-up 1 year
Beliveau, 125 1971
n=48 Radiculopathy
1 to 3 months
Main results Interlaminar epidural steroid vs. interspinous ligament saline injection (all results at 52 weeks unless otherwise noted) ODI, proportion with >75% improvement: 12% (15/120) vs. 4% (4/108) at 3 weeks (p=0.016), 32% (39/120) vs. 30% (32/108) at 52 weeks (p=0.64) SF-36: No significant differences Leg pain, >50% improvement: 48% (58/120) vs. 44% (48/108) (NS) Back pain (VAS 0 to 100), mean improvement from baseline: -8 vs. -9 (NS) Required surgery: 13% vs. 13% Off work due to sciatica: 24% (29/120) vs. 22% (24/108) Epidural steroid vs. epidural local anesthetic Proportion improved or completely relieved (clinician rated): 75% (18/24) vs. 67% (16/24)
Quality score* 9/11
1/11
Breivik, 1976126
Duration not reported n=35 Radiculopathy Chronic (mean duration not reported) n=23 Radiculopathy
Unclear
Caudal epidural steroid/local anesthetic vs. epidural local anesthetic plus large volume (100 cc) saline injection Considerable pain relief: 56% (9/16) vs. 26% (5/19), duration of follow-up unclear, p<0.05
5/11
Bush, 1991 128
1 year
Caudal epidural steroid vs. epidural saline injection (mean scores) Pain (0 to 100): 16 vs. 45 at 4 weeks (p not reported), 14 vs. 30 at 1 year (NS) Function/lifestyle (6 to 18): 16 vs. 14 at 4 weeks (p not reported), 17 vs. 16 at 1 year (NS) Deterioration of symptoms: 8% (1/12) vs. 36% (4/11) (NS) (duration unclear) Translaminar epidural steroid vs. epidural saline injection (mean treatment effect, negative values favor epidural steroid)) ODI: -2.5 (CI -7.1 to 2.2) at 3 weeks, -1.9 (CI, -9.3 to 5.4) at 3 months Pain score (0 to 100 VAS): -8.6 (CI -17.5 to 0.3) at 3 weeks, -4.0 (CI, -15.2 to 7.2) at 3 months McGill Present Pain Intensity (0 to 5): 0.0 (CI -0.4 to 0.4) at 3 weeks, 0.2 (-CI -0.3 to 0.7) at 3 months Sickness Impact Profile (0 to 100): -2.5 (CI -5.1 to 0.1) at 3 weeks, -1.2 (CI -5.2 to 2.8) at 3 months ODI <20: 3.2% (CI -8.6% to 15.0%) at 3 weeks, -4.1% (CI -19.6 to 11.3%) at 3 months Marked or very marked improvement: 3.4% (CI -11.4% to 18.2%) at 3 weeks, -0.4% (CI -16.5% to 15.7%) at 3 months Subsequent surgery: 26% vs. 25% (p=0.90) at 12 months
6/11
Carette, 1997130
Subacute or chronic n=158 Radiculopathy with imagingconfirmed disk prolapse Subacute or chronic
3 months
10/11
EVIDENCE REVIEW
Sample size Type of LBP Duration of symptoms n=73 Radiculopathy or neurogenic claudication Primarily chronic (mean duration 14 to 17 months) n=100 Radiculopathy Primarily subacute and chronic (10% acute) Helliwell, 143 1985 n=39 Radiculopathy Primarily chronic (mean duration 8 to 13 months) n=160 Radiculopathy Subacute or chronic 3 months
Author, year Cuckler, 1985132
Duration of follow-up 13 to 30 months
Dilke, 1973135
3 months
Karppinen, 145 2001
1 year
Klenerman, 1984147
n=73 Radiculopathy Duration not reported (inclusion criteria <6 months)
2 months
Main results Interlaminar epidural steroid vs. epidural saline/local anesthetic injection Average improvement: 42% vs. 44% (NS) Treatment success (>75% improvement): 32% (7/22) vs. 36% (5/14) in herniated disc group at 24 hours, 26% (6/23) vs. 15% (2/13) in herniated disc group at long-term (13 to 30 month) follow-up (NS for all comparisons) Surgery: 43% (10/23) vs. 23% (3/13) in herniated disk group at long-term (13 to 30 month) followup Interlaminar epidural steroid vs. interspinous ligament saline injection Analgesic consumption (pain clearly relieved, based on decrease on average daily dosing by 2 or more): 46% (16/35) vs. 11% (4/36) (p<0.01) during admission Pain "none" (patient assessment): 39% (16/41) vs. 24% (8/34) at 3 months Pain "none" or "not severe" (patient assessment): 98% (40/41) vs. 82% (28/34) at 3 months Analgesic consumption "none": 50% (19/38) vs. 38% (11/29) at 3 months Not returned to work: 8% (3/36) vs. 40% (14/35) at 3 months Epidural steroid injection vs. interspinous ligament saline injection Pain (mean change, 0 to 10 VAS): -2.4 vs. -0.6 at 1 month (p<0.01) and -2.5 vs. -0.3 at 3 months (p<0.01) (estimated from figure) Analgesic consumption decreased by 50% or more: 64% (7/11) vs. 40% (4/10) at 3 months, NS Overall outcome "definite improvement" (patient rated): 70% (14/20) vs. 26% (5/19) at 3 months, p<0.001 Transforaminal epidural steroid vs. epidural saline injection (mean difference in change from baseline, positive values favor epidural steroid except for sick leave) Leg pain (0 to 100): 12.5 (CI, 1.6 to 23.4, p=0.02) at 2 weeks, 2.3 (CI, -8.7 to 13.4, NS) at 4 weeks, NS or favors saline injection at 3, 6 and 12 months Back pain: 5.1 (CI, -0.3 to 10.4, NS) at 2 weeks, -12.2 (CI, -23.5 to -1.0, p=0.03) at 3 months, -8.4 (CI, -18.9 to 2.1, NS) at 12 months ODI: 5.1 (-0.3 to 10.4, NS) at 2 weeks, NS through 12 months Sick leave (days): -0.5 (CI, -4.9 to 3.9, NS) at 2 weeks, NS through 12 months Nottingham Health Profile: No difference at any follow-up time on sleep, pain, mobility, energy, and emotional reaction dimensions Epidural steroid vs. epidural local anesthetic vs. epidural saline vs. dry needle stick Proportion failed (clinician assessment): 21% (4/19) vs. 31% (5/16) vs. 31% (5/16) vs. 17% (2/12) at 2 months, p=0.74 Pain (0 to 100 VAS, mean score): 25 vs. 16 vs. 19 vs. 29 at 2 months (p not reported)
Quality score* 5/11
7/11
2/11
10/11
2/11
EVIDENCE REVIEW
Sample size Type of LBP Duration of symptoms n=133 Radiculopathy with imagingconfirmed disk prolapse Kraemer, 1997b148 Duration not reported n=49 Radiculopathy with imagingconfirmed disk prolapse Mathews, 151 1987 Duration not reported n=57 Radiculopathy Acute and subacute (mean duration 4 weeks) n=88 Radiculopathy Chronic Ridley, 1988158 n=39 Radiculopathy Riew, 2000159,
160
Author, year Kraemer, 1997a148
Duration of follow-up 3 months
Main results Targeted epidural steroid injection via an oblique interlaminar approach vs. standard interlaminar epidural steroid injection vs. epidural saline plus intramuscular steroid injection Good results (based on modified MacNab criteria): 68% vs. 53% vs. 26%, p not reported
Quality score* 2/11
3 months
Targeted epidural steroid injection via an oblique interlaminar approach vs. epidural saline plus intramuscular steroid injection Good results (based on modified MacNab criteria): 54% vs. 40% (estimated from graph), p not reported
5/11
Up to 1 year
Caudal epidural steroid vs. sacral hiatus or tender point local anesthetic injection Proportion recovered (pain score 5 or 6 on a 1 to 6 scale): 67% (14/21) vs. 56% (18/32) at 1 month (NS) Proportion with no further pain: 39% (9/23) vs. 41% (14/34) at up to 1 year
4/11
Ng, 2005152
12 weeks
2 weeks
Transforaminal epidural steroid injection vs. epidural local anesthetic injection ODI improved >10 points: 55% (24/43) vs. 35% (15/43) at 12 weeks Leg pain improved >20 points: 48% (20/43) vs. 42% (18/43) Leg pain, mean improvement (100 mm VAS): 22 vs. 21 at 6 weeks, 23 vs. 22 at 12 weeks (NS) Back pain, mean improvement: 9.9 vs. 6.3 at 6 weeks, 4.8 vs. 8.0 at 12 weeks (NS) ODI, mean improvement: 7.8 vs. 12.9 at 6 weeks, 10.8 vs. 12.3 at 12 weeks (NS) Walking distance, mean improvement: No significant differences Interlaminar epidural steroid vs. interspinous ligament saline injection Rest pain, median improvement: 46% vs. 0% at 2 weeks Walking pain, median improvement: 69% vs. 0% at 2 weeks Some improvement observed: 89% (17/19) vs. 19% (3/16), p<0.0005 Transforaminal epidural steroid vs. epidural local anesthetic injection Failure of injection (proportion of patients who underwent surgery): 8/28 (29%) vs. 18/27 (67%), p<0.004 at 1 year; among non-surgery patients after 1 year, 3/12 vs. 1/9 (p=0.42) underwent surgery at minimum five years follow-up (8 patients lost to follow-up)
11/11
5/11
Mixed (51% >6 months) n=55 Radiculopathy with imagingconfirmed disk prolapse or spinal stenosis Subacute or chronic
5 years
9/11
EVIDENCE REVIEW
Sample size Type of LBP Duration of symptoms n=30 Radiculopathy Snoek, 1977164 Chronic n=51 Radiculopathy with imagingconfirmed disk prolapse Mixed (mean duration 11 to 12 weeks) WilsonMacDonald, 2005167 n=92 Radiculopathy with imagingconfirmed disk prolapse Subacute or chronic n=63 Radiculopathy or neurogenic claudication with imagingconfirmed disk prolapsed or spinal stenosis Primarily chronic (mean 14 to 17 months) 20 to 21 months 2 years or longer
Author, year Rogers, 1992161
Duration of follow-up 1 month (primary outcomes) 2 days and 8 to 20 months
Main results Interlaminar epidural steroid vs. epidural saline injection Pain "none" or "mild": 47% (7/15) vs. 20% (3/15) at 1 month Work status "full": 53% (8/15) vs. 33% (5/15) at 1 month Analgesic intake reduced: 53% (8/15) vs. 40% (6/15) at 1 month Underwent surgery: 27% (4/15) vs. 27% (4/15) at mean 20 to 21 months Epidural steroid vs. epidural saline injection (mean improvement from baseline) Relief of low back pain: 33% vs. 25% at 2 days, p=0.88 Relief of radiating pain : 25.9% vs. 12.5% at 2 days, p=0.37 Relief of pain interfering with sleep: 53.8% vs. 23.1% at 2 days, p=0.24 Discontinuation of analgesic medications: 40.0% vs. 15.8% at 2 days, p=0.19 Improvement (physiotherapist rated): 70.0% vs. 42.8% at 2 days, p=0.22 Improvement (patient rated): 66.7% vs. 41.7% at 2 days, p=0.13 Underwent surgery: 51.9% (14/27) vs. 58.3% (14/24) at 8 to 20 months follow-up Interlaminar epidural vs. intramuscular/interspinous ligament steroid injection Proportion of patients undergoing surgery: 41% (18/44)vs. 31% (15/48), p=0.45 Pain: favors epidural group at 35 days (p<0.004) but raw data not reported
Quality score* 5/11
4/11
9/11
Zahaar, 1991169
Caudal epidural steroid vs. epidural saline injection Success (>75% improvement) at 24 hours: 65% (24/37) vs. 61% (16/26) overall; 74% (14/19) vs. 71% (10/14) in herniated disc group Success at mean 20 to 21 months: 49% (18/37) vs. 50% (13/26) overall; 58% (11/19) vs. 64% (9/14) in herniated disc group Underwent surgery at mean 20 to 21 months: 35% (13/37) vs. 38% (10/26) overall
3/11
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Four126, 143, 148, 159 of 18 trials reported long-term (greater than three months) benefits following epidural steroid injection, but three of these126, 143, 148 were rated lower-quality and did not report statistical significance of results. Two159, 854 of seven123, 130, 161, 164, 167 trials found epidural steroid injection associated with lower rates of subsequent surgery compared to placebo injection. Among four higher-quality trials (total n=533)123, 130, 159, 167, only one small (n=55) trial159 reported this effect. Three higher-quality systematic reviews reached discordant conclusions regarding short-term benefits following epidural steroid injection for sciatica or radiculopathy84, 86, 100. For non-acute (>4 weeks) sciatica, a Cochrane review found no difference between epidural steroid versus placebo injection for short-term (<6 weeks) pain relief, but only pooled data from four trials (RR=0.93, 95% CI 0.79 to 1.09)86. The highest quality and largest (n=158) trial reported results very similar to the pooled estimates (RR=0.94, 95% CI 0.76 to 1.15 and RR=1.00, 95% CI 0.71 to 1.41)129. A second, qualitative systematic review found no differences between epidural steroid and placebo injections in 5 of 7 trials, including 3 of 4 higher-quality trials84. The third higher-quality systematic review found epidural steroid superior to placebo injection for improvement in symptoms (OR=2.2, 95% CI 1.0 to 4.7) for acute or chronic sciatica100. Its conclusions may be sensitive to inclusion of a trial that reported an unusually high odds ratio for short-term reduction in inpatient analgesic consumption immediately following the injection (OR=6.8, compared to 1.1 to 2.8 in the other trials)135. Although this trial was excluded from the Cochrane review because it allowed enrollment of patients with acute symptoms, only 10% had symptoms less than four weeks. Three lower-quality systematic reviews each found some evidence for short-term pain relief following epidural steroid injections for radiculopathy, but also at least some inconsistency between trials71, 74, 97. None of the systematic reviews evaluated results stratified according to use of epidural or soft tissue placebo injection. One higher-quality trial found that if a first epidural injection was not effective, additional injections within the first six weeks were no more effective123.
Efficacy of translaminar or caudal epidural steroid injection versus epidural placebo injection (saline or local anesthetic) for spinal stenosis
For spinal stenosis, one higher-quality trial found an epidural steroid plus local anesthetic injection or an epidural local anesthetic injection alone superior to an epidural saline injection on improved walking distance at one week, though no differences were observed after 3 months (Table 74)106. Two lower-quality trials that enrolled mixed populations of patients did not find beneficial effects of epidural steroids in small subgroups of patients with spinal stenosis132, 169.
EVIDENCE REVIEW
Table 74. Randomized, placebo-controlled trials of epidural steroid injection for spinal stenosis
Sample size Type of LBP Duration of symptoms n=73 Radiculopathy or neurogenic claudication Primarily chronic (mean duration 14 to 17 months) Fukusaki, 1998106 n=53 Spinal stenosis (diagnosed by orthopedist) Subacute or chronic Zahaar, 169 1991 n=63 Radiculopathy or neurogenic claudication with imaging-confirmed disk prolapsed or spinal stenosis Primarily chronic (mean 14 to 17 months) 20 to 21 months 3 months
Author, year Cuckler, 1985132
Main results Interlaminar epidural steroid vs. epidural saline/local anesthetic injection Average improvement: 42% vs. 44% (NS) Treatment success (>75% improvement): 25% (5/20) vs. 18% (3/17) in spinal stenosis group at 24 hours, 22% (5/23) vs. 14% (2/14) in spinal stenosis group at long-term (13 to 30 month) follow-up (NS for all comparisons) Surgery: 26% (6/23) vs. 29% (4/14) in spinal stenosis group at long-term (13 to 30 month) follow-up Interlaminar epidural steroid + local anesthetic vs. epidural local anesthetic alone vs. epidural saline Walking distance: 87 vs. 92 vs. 23 at 1 week, 10 vs. 13 vs. 11 at 3 months (p<0.05 for A and B vs. C at week 1 only) Good results (walk >100 m): 63% vs. 56% vs. 12% at 1 week, 5% vs. 6% vs. 6% at 3 months (no difference between A and B) Caudal epidural steroid vs. epidural saline injection Success (>75% improvement) at 24 hours: 56% (10/18) vs. 50% (6/12) in spinal stenosis group Success at mean 20 to 21 months: 49% (18/37) vs. 50% (13/26) overall; 58% (11/19) vs. 64% (9/14) in herniated disc group Underwent surgery at mean 20 to 21 months: 35% (13/37) vs. 38% (10/26) overall
Quality score* 5/11
6/11
3/11
Efficacy of translaminar or caudal epidural steroid injection versus placebo epidural injection (saline or local anesthetic) for low back pain without sciatica or radiculopathy
For low back pain without sciatica or radiculopathy (non-specific low back pain), one small, lower-quality trial found no differences in pain or functional outcomes between epidural steroid injection versus intrathecal midazolam852.
Efficacy of transforaminal epidural steroid injection versus placebo epidural injection for low back pain with sciatica or radiculopathy
Most placebo-controlled trials evaluated the interlaminar or caudal approach. Three higherquality, placebo-controlled trials evaluating the transforaminal approach reported mixed results (Table 75)145, 152, 159. One small (n=55), higher-quality trial found that a lower proportion of patients with radicular pain randomized to transforaminal epidural steroid injection proceeded to surgery compared to those randomized to an epidural local anesthetic injection, after 1 year (29% vs. 67%)159. Five-year rates of surgery have also been reported from this trial, but results are difficult to interpret because 40% of the patients in the epidural steroid arm who had not
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain undergone surgery by 23 months were lost to follow-up160. Two other higher-quality trials found transforaminal epidural steroid injection to be no better on most outcomes compared to epidural saline injection through 12 months145, or compared to epidural local anesthetic injection after 6 to 12 weeks152 (Table 75).
Efficacy of epidural steroid injection versus local injections or dry needling of the interspinous ligament
For sciatica or radiculopathy, one higher-quality trial found no statistically significant differences between epidural steroid and trigger point injection in the proportion of patients who recovered after one month (67% vs. 56%), though the epidural steroid was superior at three months151. In a lower-quality trial, transforaminal and interlaminar epidural steroid injections were associated with a greater likelihood for a good overall response at 3 months (68% and 53%, respectively) compared to a paravertebral local anesthetic injections (26%)148. However, it appeared that the paravertebral injection was meant to serve as a placebo control in this trial, and it was not clear if the local anesthetic was administered at tender points. One trial did not meet inclusion criteria because it was not randomized208. It found transforaminal epidural steroid injection superior to trigger point injection for the proportion of patients with a successful outcome at 12 months (84% vs. 48%). For sciatica, one small (n=74), lower-quality trial found no differences between epidural steroid injection and dry needling of the interspinous ligament in the proportion of patients improved or cured according to clinician assessment (79% or 15/19 versus 83% or 10/12)147.
Efficacy of epidural steroid injection versus other interventions

One small (n=60), lower-quality trial found no differences between transforaminal epidural injection with steroid versus transforaminal epidural injection with hyaluronidase in patients with failed back surgery syndrome844. Five other recent (published since 2004) trials also compared epidural steroid injection to other interventions (Table 76)102, 107, 115, 167, 842. One higher-quality trial of patients with sciatica for at least six weeks found no significant difference between epidural steroid injection compared to intramuscular steroid (methylprednisolone 80 mg) plus local anesthetic injection in rates of subsequent surgery (41% vs. 31%) after two years or more167. Pain relief favored the epidural group at 35 days (p<0.004, other data not provided) but differences were no longer significant at subsequent follow-up. For large herniated disc with no improvement after at least six weeks, another higher-quality trial found epidural steroid injection moderately to substantially inferior to discectomy for most short-term (1-3 months) outcomes842. Differences were no longer observed for most outcomes by 2-3 years, but results are difficult to interpret because about one-third of the patients assigned to epidural steroids crossed over to surgery, and intention-to-treat results were not reported. Among patients randomized to epidural steroids that did not cross over to surgery, 42% to 56% considered their treatment a success, compared to 92% to 98% in patients allocated surgery and 82% to 93% in patients who crossed over to surgery. For chronic (more than two years duration) back pain with no response to a previous epidural steroid injection, a
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain third trial found epidural steroid alone substantially inferior to adhesiolysis either with or without hypertonic saline for pain relief, functional status, opioid intake, and psychiatric outcome measures115. Even though this trial was rated higher-quality, its results needs to be confirmed because of an unusually low response rate (defined as >50% pain relief at 12 months) in the epidural steroid group (0%) and very high response rates in the adhesiolysis groups (72% and 60%). Two lower-quality trials not included in the systematic reviews evaluated efficacy of oxygenozone (O2-O3) injections. For acute or chronic low back pain with sciatica, one trial found a transforaminal epidural steroid injection inferior to transforaminal O2-O3 injection on rates of resolution of pain and return to normal activities in the subgroup of patients with a herniated or bulging disc at 6 months (58% vs. 74%), but not at earlier follow-up102. There were no differences between epidural steroid injection and oxygen-ozone injection in the subgroup of patients without a herniated or bulging disc. In patients with radicular pain and lumbar disc prolapse, the second trial found intradiscal plus intraforaminal epidural steroid injection inferior to intradiscal plus intraforaminal epidural steroid plus oxygen-ozone injection for achieving an ODI <20 at 6 months (47% vs. 74%, p<0.01), though differences were not significant at earlier follow-up107.
EVIDENCE REVIEW
Table 75. Recent trials of epidural steroid injection versus other interventions
Author, year Bonetti, 2005102
Buttermann, 2004842
n=100 3 years
Gallucci, 2007107
n=159 6 months
Manchikanti, 2004115
n=75 12 months
Wilson-MacDonald, 2005167
n=93 2 years or longer
Main results Transforaminal oxygen-ozone injection vs. transforaminal corticosteroid injection Herniated or bulging disc group: Excellent result (resolution of pain and return to baseline activity): 85% vs. 80% (NS) at 1 week, 78% vs. 68% (p=0.13) at 3 months, and 74% vs. 58% (0.002) at 6 months Non-disc disease group: Excellent result: 80% vs. 78% (NS) at 1 week, 78% vs. 70% (p=0.25) at 3 months, and 76% vs. 63% (p=0.099)at 6 months Epidural steroid versus discectomy Motor deficit: 72% vs. 38% at 1-3 months (p<0.05), 9% vs. 4% at 2-3 years (NS) Back pain, mean score (0 to 10 VAS): 3 vs. 2 at 1-3 months (p<0.05), 1.8 vs. 2.4 at 2-3 years (NS) Leg pain, mean score (0 to 10 VAS): 4.1 vs. 1.4 at 1-3 months (p<0.05), 0.8 vs. 1.5 at 2- 3 years (NS) ODI, mean score: 34 vs. 22 at 1-3 months (p<0.05), 8 vs. 16 at 2-3 years (NS) Much less use of medication: 16% vs. 24% at 1-3 months, 57% vs. 32% at 2-3 years Transforaminal and intradiscal corticosteroid + oxygen-ozone versus corticosteroid alone Treatment success (<20 on ODI): 88% vs. 90% (p=0.72) at 2 weeks, 78% vs. 67% (p=0.14) at 3 months, 74% vs. 47% (p<0.001) at 6 months Epidural steroid vs. adhesiolysis with hypertonic saline vs. adhesiolysis with isotonic saline Proportion with >50% pain relief at 12 months: 0% vs. 72% vs. 60% (p<0.001) ODI score at 12 months: 32 vs. 23 vs. 24 (p<0.001) VAS pain score (0 to 10) at 12 months: 7.7 vs. 4.6 vs. 5.2 Taking opioids: 52% vs. 16% vs. 16% (p<0.001) Epidural steroid versus intramuscular steroid plus local anesthetic Proportion of patients undergoing surgery after at least 2 years: 41% vs. 31%, p=0.45
Quality score 4/11
5/9*
5/11
8/11
9/11
*Excludes criteria involving blinding patients and care providers, for maximum score of 9
Efficacy of different approaches for administering epidural steroids

In six trials (two rated higher-quality101, 843) that directly compared different methods for administering epidural steroids, no approach was clearly superior148, 846, 847, 853. One higherquality trial (n=90) found the transforaminal approach superior to both the interlaminar and caudal approaches101, but two lower-quality trials reported inconsistent results for the transforaminal versus interlaminar approaches846, 853. One lower-quality trial found an oblique interlaminar approach modestly superior to a standard translaminar approach148, and another lower-quality trial found no differences between the caudal and translaminar approaches847. Radiologic confirmation of epidural placement with the translaminar approach was either not reported or not performed in any of these trials, and no trial compared outcomes of epidural
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain steroid injection with fluoroscopic guidance versus epidural steroid injection without fluoroscopic guidance. One recent higher-quality trial compared epidural steroid via the caudal approach versus targeted steroid placement during spinal endoscopy in patients with radicular back pain for at least six months, with needle placement confirmed by fluoroscopy for both methods (Table 76)843. It found no difference in any outcome between the two approaches.
Table 76. Trial of epidural steroid via caudal approach versus targeted placement during spinal endoscopy
Author, year Dashfield, 2005843
Main results Epidural steroid via caudal approach versus targeted placement during spinal endoscopy Pain (VAS) mean improvement: -1.4 vs. -1.22 (NS) Present pain intensity, mean improvement: -0.8 vs. -1.0 (NS) Short-form McGill Pain Questionnaire (sensory), mean improvement: -2.3 vs. +0.5 (NS) Short-form McGill Pain Questionnaire (affective), mean improvement: 0 vs. 0 (NS) HAD-anxiety and -depression scales: no significant differences between groups
Quality score 7/11
Harms Although there are case reports of serious adverse events, including paralysis and infection, following epidural steroid injection855-857, serious adverse events were rarely reported in randomized trials. However, reporting of harms was suboptimal. Ten placebo-controlled trials didnt report harms at all.106, 126, 132, 145, 151, 158, 159, 161, 167, 169 When reported, adverse events were typically transient and minor and included headache, nausea, irregular periods, pruritus, and increased sciatic pain. A recent, large (n=228), high-quality trial reported post-injection headache in 3.3% (4/120) receiving epidural steroid, postdural puncture headache in 0.8% (1/120), nausea in 1.7% (2/120), and other adverse events in 4.2% (5/120)123. Serious adverse events were also uncommon in trials that evaluated the transforaminal approach145, 148, 152, 846. One trial reported a 1.9% incidence of headache148, one trial reported one episode of acute hypertension846, and another reported one retroperitoneal bleed in a patient on anticoagulation145. One trial found that all patients who underwent targeted placement of steroids during spinal endoscopy reported increased back pain, though no post-spinal headache, dural tap, or infection was observed843. Costs One trial found no significant differences between transforaminal steroid and saline injections for cost per one response ($3,740 versus $3,629)145. However, a subgroup analysis suggested transforaminal steroid injection was more cost-effective for contained herniations ($4,432 versus $17,098 per responder, p=0.0073) than for extrusions ($7,165 versus $2,484, p=0.0058). Another trial estimated 44,701/QALY (about $86,273 U.S./QALY) for up to three translaminar
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain epidural steroid injections and 25,746 (about $49,689 U.S./QALY) for one injection from the provider perspective123, 858. From the purchaser perspective, incremental cost-effectiveness for one injection was 167,145 (over $300,000 U.S./QALY).
Summary of evidence
For low back pain with sciatica, evidence of beneficial effects following epidural steroid injections by translaminar or caudal approaches is mixed. Although some higher-quality trials report short-term benefits versus placebo injection, results are inconsistent. Most trials found no longer-term benefits following epidural steroid injection, and one higher-quality trial found no additional benefits from repeated injections. Most evidence on epidural steroid injections is in patients with low back symptoms of at least one months duration (level of evidence: fair). For low back pain with sciatica, evidence on efficacy of epidural steroid injection by the transforaminal approach is mixed, with two of three higher-quality trials showing no benefit compared to control injections (level of evidence: fair). For low back pain with sciatica, one higher-quality randomized trial found epidural steroid injection no better than trigger point injections at one month for overall outcomes, though modestly superior at three months. Other trials that compared epidural steroids and local injections were either not randomized or did not clearly inject tender points (level of evidence: fair). There is insufficient evidence (one lower-quality trial) to accurately judge relative efficacy of epidural steroids compared to dry-needling of the interspinous ligament (level of evidence: poor). For low back pain with sciatica, epidural steroid injections were not clearly superior to intramuscular steroids for long-term outcomes in one higher-quality trial (level of evidence: fair). For low back pain without radiculopathy, there is insufficient evidence (one lower-quality trial showing no benefit) to accurately judge efficacy of epidural steroids (level of evidence: poor). For spinal stenosis, one small, higher-quality trial found epidural steroids have no sustained effects on walking distance compared to a placebo injection and two small subgroup analyses found no clear benefits associated with epidural steroid injection (level of evidence: poor). In patients with chronic low back pain who failed a previous epidural steroid injection, one small, higher-quality trial found epidural steroid injection alone inferior to epidural adhesiolysis, but reported high rates of response in the adhesiolysis group (60% to 72%) and unusually low rates in the epidural arm (0%) (level of evidence: poor). For lumbar disc prolapse, one trial found epidural steroids superior to discectomy for shortterm but not longer-term outcomes, but results are difficult to interpret because crossover rates were high and intention-to-treat results not reported (level of evidence: poor). For low back pain with sciatica, two lower-quality trials found transforaminal epidural steroid injections inferior to tranforaminal oxygen-ozone injections in patients with bulging or herniated disc for resolution of pain and improvement in function at 6 months, but not at earlier
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain assessments. One of the trials also evaluated intradiscal injections of steroids with and without oxygen-ozone (level of evidence: poor) Six trials (two higher-quality) that directly compared different approaches for administration of epidural steroids found inconsistent results, or no clear differences (level of evidence: poor). One higher-quality trial found no differences between caudal epidural steroid and targeted steroid placement during spinal endoscopy, with needle placement for both methods confirmed by fluoroscopy (level of evidence: fair). Serious adverse events were rare in trials of epidural steroid injections, but adverse events were generally not well reported (level of evidence: poor).

The AHCPR guidelines found no evidence to support the use of invasive epidural injections of steroids, local anesthetics, and/or opioids as a treatment for acute low back pain without radiculopathy (strength of evidence: D). The AHCPR guidelines recommend epidural steroids as an option for short-term relief of radicular pain after failure of conservative treatment and as a means of avoiding surgery (strength of evidence: C). The VA/DoD guidelines found limited evidence to support the use of epidural steroid injections for acute low back pain with nerve root pain and radicular neurologic deficit (strength of evidence: C). The UK RCGP guidelines found that epidural steroids with or without local anesthetic appear to produce better short-term relief of acute low back pain with sciatica than comparison treatments (strength of evidence: **). The UK RCGP guidelines found limited evidence that epidural injections are not beneficial for acute low back pain without radiculopathy (strength of evidence: *). The UK RCGP guidelines found that because of its invasive nature, epidural injections pose rare but serious potential risks (strength of evidence: **).
The European COST guidelines recommend against epidural steroid injections for acute
nonspecific low back pain and found insufficient evidence to recommend epidural injections for chronic, nonspecific low back pain.
Facet joint injection and medial branch block

Facet joint injections involve administration of corticosteroids into the facet joints in order to reduce pain and inflammation. The epidural space is not entered. Nerve blocks of the medial branch of the posterior primary ramus (medial branch blocks) are primarily used as a diagnostic procedure to identify patients with facet joint pain. However, some trials have also evaluated clinical or therapeutic effects of medial branch blocks.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: systematic reviews We identified one higher-quality Cochrane review on efficacy of facet joint injections for chronic low back pain94. We also identified three lower-quality systematic reviews75, 92, 93. We excluded three systematic reviews that have already been updated86, 171, 187. Results of search: trials Eight randomized trials evaluated facet joint injection or medial branch block.116, 129, 150, 859-864 Seven were included in at least one of four systematic reviews75, 92-94 and we identified one additional trial.116 Two trials (both evaluating facet joint injection) were placebo-controlled.129, 150 We excluded one trial that focused on the utility of bone scintigraphy for guiding facet joint injections 206.
Efficacy of facet joint steroid injection versus control (saline) facet joint injection
A higher-quality trial (n=101) by Carette et al enrolled patients with chronic low back pain who responded to a single local anesthetic injection into the facet joint with immediate pain relief (Table 77)129. It found no difference in the likelihood of pain relief in patients randomized to steroid or saline either one month or three months after the injection (RR=0.89, 95% CI 0.65 to 1.21 and RR=0.90, 95% CI 0.69 to 1.17, respectively). Although a higher proportion of patients in the corticosteroid injection group experienced marked or very marked improvement after six months (46% vs. 15%, p=0.002), the biologic rationale for such a delayed (after three months) benefit from steroids is unclear. In addition, differences at six months were attenuated after controlling for the increased use of co-interventions in the steroid group. The difference in the proportion of patients that experienced sustained improvement (improvement at one, three, and six months) was not statistically significant (22% vs. 10%, p=0.19); half of the 22 patients with improvement at 6 months did not show benefits at earlier time periods. A second, lower-quality trial found no difference in mean pain scores between facet joint intracapsular or pericapsular steroid and bupivacaine injection compared to saline injection150, 860. In this trial, patients were enrolled based on clinical criteria, and did not require a positive response to diagnostic facet joint blocks.
EVIDENCE REVIEW
Table 77. Randomized, placebo-controlled trials of facet joint steroid injection

Sample size Type of LBP Duration of symptoms n=101 Presumed facet joint pain with positive response to uncontrolled facet joint block Chronic
Author, year Carette, 1991129
860
Lilius, 1989150,
n=109 Presumed facet joint pain without diagnostic facet joint block Chronic
3 months
Main results Facet joint intra-articular steroid vs. saline injection Very marked or marked improvement: 42% vs. 33% at 1 month (p=0.53), 36% vs. 28% at 3 months (p=0.51), 46% vs. 15% at 6 months (p=0.002) Sustained improvement through 6 months: 22% vs. 10%, p=0.19 Mean VAS pain score (0 to 10): 4.5 vs. 4.7 at 1 month (NS), 4.0 vs. 5.0 at 6 months (p<0.05) McGill Pain Questionnaire (0 to 5): 2.3 vs. 2.6 at 1 month (NS), 2.1 vs. 2.9 at 6 months (p<0.05) Sickness Impact Profile Overall score (0 to 100): 9.3 vs. 9.8 at 1 month (NS), 7.8 vs. 10.8 at 6 months (NS) Mean days with complete restriction in main activity in last 2 weeks: 3.2 vs. 2.2 at 1 month (p=0.22), 1.3 vs. 2.9 at 6 months (p=0.07) Facet joint intra-articular steroid vs. pericapsular steroid vs. intra-articular saline Return to work: No differences (data not reported) Pain score: No differences (data not reported) Pain improvement (categorical): No differences (data not reported)
Quality score* 7/11
4/11
A higher-quality Cochrane review94 and two92, 93 and two of three lower-quality systematic reviews also found no clear benefits associated with facet joint steroid versus placebo injection. A third lower-quality systematic review found moderate evidence that facet joint injections are associated with short-term improvement75. Reasons for the discrepancy in this reviews conclusions include its exclusion of the trial by Lilius et al because it did not use diagnostic facet joint blocks to select patients150, 860, its classification of the trial by Carette et al as showing benefits of facet joint injection129, its classification of an active-controlled trial as demonstrating efficacy of facet joint injection because both intervention groups improved compared to baseline859, and its inclusion of evidence from several small (N<100), non-randomized studies.
Efficacy of medial branch block versus placebo

We found no trials comparing therapeutic medial branch block versus placebo.
Efficacy of facet joint injection versus medial branch block

One higher-quality trial found no difference in pain relief one to three months after a facet joint injection with a steroid and local anesthetic compared to medial branch block of the posterior primary ramus with a steroid and local anesthetic863. A second, lower-quality trial not included in any previously published systematic review reported no benefit with either facet joint injection
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain with local anesthetic plus steroid or medial branch block with local anesthetic only, but outcomes were reported using unconventional and difficult to interpret methods (paired sequential analysis) (Table 78)116.
Table 78. Additional trial not included in previously published systematic reviews of facet joint vs. medial branch nerve block
Number of patients Duration of follow-up n=67 1 month
Author, year Nash, 1989116
Main results Facet joint injection vs. medial branch nerve block Overall outcome: No difference between groups based on paired sequential analysis (12 pairs nerve blockade more beneficial, 11 pairs intra-articular injection more beneficial, no results for 8 pairs)
Quality score 2/11
Efficacy of facet joint injection plus home stretching versus home stretching alone
For patients with presumed lumbar segmental rigidity, one lower-quality trial found bilateral lumbar facet joint corticosteroid injection plus a home stretching program to be no more effective than stretching alone for pain or function864.
Efficacy of different types of facet joint injection

For non-radicular back pain and at least moderate facet joint osteoarthritis on imaging (facet joint blocks not performed), one higher-quality trial859 included in one systematic review75 found no clear differences in pain, back-specific functional status, or other outcomes between facet joint injection with a steroid versus facet joint injection with hyaluronic acid (Table 79). All patients received facet joint injections bilaterally at the L3-L4, L4-L5, and L5-S1 levels.
Table 79. Facet joint injection with steroid vs. facet joint injection with hyaluronic acid
Author, year Fuchs, 2005859
Main results Facet joint injection with steroid vs. facet joint injection with hyaluronic acid (mean improvement from baseline) Pain (0 to 100 scale): -35.3 vs. -31.2 (hyaluronic acid non-inferior) RDQ (0 to 24 scale): -4.2 vs. -5.4 (p not reported) ODI (0 to 50 scale): -5.4 vs. -8.1 (p not reported) Low Back Outcome Score (0 to 75 scale): +11.4 vs. +14.1 (p not reported) SF-36: similar in both groups
Quality score 6/11
Efficacy of different types of medial branch blocks

One small (n=73), higher-quality trial found no differences in outcomes become a medial branch block with a local anesthetic, Sarapin (a substance derived from the pitcher plant), and methylprednisolone versus the same intervention without the methylprednisolone862. A small (n=60) trial compared nerve block injection with bupivicaine alone, bupivicaine plus steroid,
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain bupivicaine plus sarapin, or bupivicaine plus sarapin and steroid. Response rates ranged from 73% to 93% at 3 to 12 months861. Harms No adverse events other than transient local pain at the injection sites were reported in the lone higher quality trial129. Costs We found no studies evaluating costs.
Summary of evidence
There is no evidence on efficacy of facet joint injections or medial branch blocks for acute low back pain. For presumed chronic facet joint pain, two randomized trials found facet joint steroid injection no more beneficial than facet joint control injections for short-term pain relief or sustained pain relief (level of evidence: fair). There is no evidence on efficacy of medial branch block versus placebo injection for chronic low back pain. For presumed chronic facet joint pain, two trials (one higher-quality) found no difference between facet joint steroid injection and medial branch block with or without steroid (level of evidence: fair). For patients with presumed lumbar segmental rigidity, one lower-quality trial found no differences between bilateral facet joint corticosteroid injections plus stretching versus stretching alone (level of evidence: poor). For chronic non-radicular back pain with radiographic findings of at least moderate facet joint osteoarthritis, facet joint steroid injection and facet joint hyaluronic acid injection were associated with similar outcomes in one higher-quality trial (level of evidence: fair). For presumed chronic facet joint pain, there is insufficient published evidence (one small, higher-quality trial and one unpublished trial) to evaluate efficacy of medial branch blocks with local anesthetic plus Sarapin, with or without corticosteroid (level of evidence: poor).

The AHCPR guidelines recommend against facet joint injections in patients with acute low back problems (strength of evidence: C). The VA/DoD guideline recommendation is similar. The UK RCGP guidelines found that facet joint injection do not produce pain relief or global improvement, with neither the type of agent injected nor the site of injection making a significant difference to outcomes (strength of evidence: *). The UK RCGP guidelines also found no evidence on the efficacy of facet injections in acute low back problems (strength of evidence: *).
EVIDENCE REVIEW
Sacroiliac joint steroid injection

Sacroiliac joint injections with corticosteroids are intended to decrease pain and inflammation in the sacroiliac joint when it is the presumed source of low back pain. However, diagnosis and treatment of sacroiliac joint pain in patients without spondyloarthropathy remains controversial (see Key Question 6). Results of search: systematic reviews We identified one higher-quality systematic review on sacroiliac joint injections83. We excluded an earlier version of this review186. Results of search: trials We identified one higher-quality randomized trial of patients with presumed sacroiliac pain not due to spondyloarthropathy114. It evaluated a periarticular sacroiliac steroid injection and was excluded from the previously published systematic review. We excluded one small (n=10) randomized trial of sacroiliac joint injection that was included in the previously published systematic review because it enrolled patients with spondyloarthropathy205.
Efficacy of sacroiliac joint injection versus control injection

For chronic pain in the sacroiliac joint area and at least one physical exam finding for sacroiliac joint pain, one small (n=24), higher-quality trial found a periarticular sacroiliac steroid injection substantially superior to control (local anesthetic) injection for improvement in one-month pain scores (Table 80)114.
Table 80. Randomized, placebo-controlled trial of sacroiliac joint injection for suspected sacroiliac joint pain
Sample size Type of LBP Duration of symptoms n=24 Sacroiliac joint pain without spondyloarthropathy Chronic
Author, year Luukainen, 2002114
Duration of follow-up 1 month
Main results Periarticular sacroiliac joint steroid injection vs. control injection VAS (0 to 100), improvement in median scores: -40 vs. -13, p=0.046 Pain index (0 to 12), improvement in median scores: -3 vs. 0, p=0.017
Quality score* 6/11
Harms No adverse events were reported in the trial. Costs We found no studies evaluating costs.
EVIDENCE REVIEW
Summary of evidence
In patients thought to have sacroiliac pain not related to spondyloarthropathy, one higherquality but very small (n=24) trial found sacroiliac joint steroid injection substantially superior to local anesthetic injection for short-term pain relief (level of evidence: poor).

The European COST guidelines found insufficient evidence to recommend the use of corticosteroid injections for nonspecific chronic low back pain.
Intradiscal steroid injections

Intradiscal steroid injections differ from epidural steroid injections because they involve direct injection of steroids into the intervertebral disc, rather than into the epidural space. They have been performed for presumed chronic discogenic low back pain and lumbar disc prolapse with sciatica or radiculopathy. Results of search: systematic reviews We identified a higher-quality Cochrane review on intradiscal steroid injections for low back pain81, 82. We excluded two outdated versions of this review176, 177. Results of search: trials Six randomized trials evaluated intradiscal steroid injection104, 109, 112, 119, 865, 866. Two trials were included in the Cochrane review81, 82 and we identified four additional trials104, 109, 112, 119, 867. Three trials were placebo-controlled104, 112, 119. All three placebo-controlled trials evaluated intradiscal steroid injection for degenerative disc disease.
Efficacy of intradiscal steroid versus control or no injection for presumed discogenic low back pain
For chronic low back pain with MRI evidence of degenerative disc disease and positive results on provocative discography, two trials (one higher-quality119) found no significant differences between intradiscal steroid and control injections (saline or local anesthetic) for either short- or long-term pain relief or improvement in functional status (Table 81)112, 119. In the trial that reported longer-term outcomes, the median pain score was unchanged in both groups at one year112. A third, lower-quality trial found that in patients with degenerative disc disease who failed an epidural steroid injection, intradiscal steroid injection was superior to discography only in the subgroup of patients with inflammatory endplate changes on MRI104. However, changes in outcome scores and levels of statistical significance were poorly reported in this study. At 1 to 2 years, rates of success (not clearly defined) in the subgroup with inflammatory endplate changes were 25% in patients randomized to discography plus intradiscal steroid, and 0% in the group randomized to discography alone. The proportion of patients who subsequently underwent fusion in this subgroup was 50% among those randomized to intradiscal steroid compared to 76% among those randomized to discography alone. Intradiscal steroid injection was also superior for functional status (ODI), though not for pain scores.
EVIDENCE REVIEW
Table 81. Randomized, placebo-controlled trials of intradiscal steroids

Sample size Type of LBP Duration of symptoms n=171 Presumed discogenic back pain Chronic
Author, year Buttermann, 2004104
Duration of follow-up 1-2 years
Main results Discography + intradiscal steroid vs. discography alone (estimated from graphs) Inflammatory end-plate changes present: Pain, mean improvement in VAS (0 to 10): -0.3 vs. +0.6 ODI (0 to 100), mean improvement: -18 vs. +9 "Success" (not defined): 10/40 (25%) vs. 0/38 (0%) Underwent fusion: 50% vs. 76% No inflammatory end-plate changes present: Pain, mean improvement in VAS: -1.2 vs. +0.6 ODI (0 to 100), mean improvement: -1 vs. -1 "Success" (not defined): 5/46 (11%) vs. 1/47 (2%) Underwent fusion: 78% vs. 89% Much less use of medication: 16% vs. 24% at 1-3 months, 57% vs. 32% at 2-3 years Intradiscal methylprednisolone vs. intradiscal saline ODI, mean improvement (percent): 2.28 vs. 3.42 (p=0.71) VAS pain score (0 to 10), median change: 0 vs. 0 (p=0.72) Intradiscal methylprednisolone vs. intradiscal bupivicaine Proportion improved overall: 3/14 (21%) vs. 1/11 (9%) (NS) Proportion improved on VAS pain scale: 43% vs. 36% (NS) Proportion improved on ODI: 36% vs. 27% (NS)
Quality score* 5/11
Khot, 2004112
n=120 Presumed discogenic back pain
1 year
4/11
Simmons, 1992119
At least subacute n=25 Presumed discogenic back pain At least subacute
10-14 days
6/11
Efficacy of intradiscal steroid versus chemonucleolysis for low back pain with sciatica or radiculopathy
In patients with sciatica, two French-language trials865, 866 (one higher-quality865) included in the Cochrane review found no differences between intradiscal steroid injection and chemonucleolysis for risk of failure or no improvement (OR=1.20, 95% CI 0.61 to 2.38). For chronic back pain and sciatica unresponsive to non-invasive therapy, a lower-quality trial also reported similar rates of success (defined as the proportion virtually pain-free) with intradiscal steroids and chemonucleolysis (Table 82)109.
EVIDENCE REVIEW
Table 82. Trial of intradiscal steroid versus chemonucleolysis not included in previously published systematic review
Number of patients Duration of follow-up n=40 Duration of followup unclear
Author, year Graham, 1975109,

867
Main results Intradiscal steroids vs. chemonucleolysis "Success" (proportion virtually pain-free): 45% vs. 45%
Quality score 4/11
Summary of evidence
For presumed chronic discogenic low back pain (positive results on provocative discography), there is consistent evidence from three trials (one higher-quality) that intradiscal steroids are not associated with improved outcomes compared to control injections (level of evidence: good). For presumed chronic discogenic low back pain, a subgroup analysis from one lower-quality trial found intradiscal steroids superior to discography alone in a selected subgroup of patients that failed epidural steroid injections and had inflammatory endplate changes on MRI (level of evidence: poor). For prolapsed lumbar disc or sciatica, three trials (one higher-quality) consistently found no differences between intradiscal steroid injection and chemonucleolysis (level of evidence: good). None of the trials reported safety outcomes.

The European COST guidelines recommend against intradiscal steroids for chronic low back pain.
Chemonucleolysis
Chemonucleolysis involves the injection of a proteolytic enzyme into an intervertebral disc in order to break down the gelatinous nucleus. The goal of chemonucleolysis is to reduce disc size and relieve pressure on compressed nerve roots. Chemonucleolysis has most frequently been studied using chymopapain (derived from papaya) injections, though collagenase (which may be less likely to induce an allergic reaction) has also been used. Chemonucleolysis is practiced infrequently in the U.S.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: systematic reviews We identified one higher-quality Cochrane review of surgery for lumbar disc prolapse that included trials of chemonucleolysis81, 82. We excluded two outdated Cochrane reviews176, 177 and two other outdated systematic reviews190, 191. Results of search: trials 22 randomized trials evaluated chemonucleolysis110, 122, 127, 133, 136, 144, 162, 366, 865, 866, 868-878. Nineteen trials were included in the Cochrane review81, 82 and we identified three additional trials110, 122, 366. One compared chemonucleolysis to spinal manipulation366, one compared longterm outcomes of chemonucleolysis with chymopapain versus chemonucleolysis with collagenase122, and one compared transforaminal posterolateral endoscopic discectomy plus low-dose chymopapain versus transforaminal posterolateral endoscopic discectomy alone110. Six of the 22 trials were placebo-controlled.127, 133, 136, 144, 162, 871 We rated the four of the five English-language placebo-controlled trials higher-quality127, 133, 144, 162. A sixth, small (n=39) French-language placebo-controlled trial was included in the Cochrane review871.
Efficacy of chemonucleolysis versus placebo for lumbar disc prolapse with radiculopathy
For lumbar disc prolapse, three133, 136, 144 of four162 English-language trials found chymopapain chemonucleolysis superior to placebo for achieving treatment success (variably defined) (Table 83). A fifth trial found collagenase chemonucleolysis superior to placebo, but 40% of patients in this trial were no longer blinded after 8 weeks127. Based on pooled results, a higher-quality Cochrane review that also included a French-language trial871 found chymopapain chemonucleolysis associated with a lower-likelihood for a poor patient-reported overall outcome (no success) compared to placebo after one year (OR=0.24, 95% CI 0.12 to 0.49, 2 trials), and lower likelihood of open discectomy within 6 to 24 months (OR=0.41, 95% CI 0.25 to 0.68, 5 trials)81, 82. A total of 446 patients enrolled in 5 trials were included in the pooled results of subsequent surgery rates133, 136, 144, 162, 871.
EVIDENCE REVIEW
Table 83. Randomized, placebo-controlled trials of chemonucleolysis

Sample size Type of LBP Duration of symptoms n=30 Radiculopathy Duration not reported Dabezies, 1988133 n=173 Radiculopathy with imageconfirmed disk prolapse Duration not reported 6 months
Author, year Bromley, 1984127
Fraser, 1982136, 137
n=60 Radiculopathy Primarily nonchronic
6 months through 10 years
Main results Chemonucleolysis with collagenase vs. intradiscal saline Pain improvement "good" or "fair" (patient rated): 80% (12/15) vs. 33% (5/15) at 8 weeks, p<0.005 Pain improvement "good" or "fair" (clinician rated): 80% (12/15) vs. 33% (5/15) at 8 weeks, p<0.005; 80% (12/15) vs. 20% (3/15) at mean 16.8 months (p<<0.005) Chemonucleolysis with chymopapain vs. intradiscal placebo (cysteine-edetateiothalamate) Overall outcome moderately improved or pain-free (investigator-rated, lost to follow-up excluded): 73% (56/74) vs. 52% (42/81) at 6 weeks (p=0.01), 72% (46/64) vs. 49% (37/76) at 3 months, p=0.01, 71% (44/62) vs. 45% (33/74) at 6+ months (p=0.01) Treatment success (lost to follow-up considered failure): 72% (56/78) vs. 52% (42/81) at 6 weeks, 59% (46/78 vs. 46% (37/81) at 3 months, 56% (44/78) vs. 41% (33/81) at 6 months Subsequent surgery: 4% (7/78) vs. 25% (20/81) Chemonucleolysis with chymopapain vs. intradiscal saline Treatment success (patient-rated): 73% vs. 37% at 6 weeks (p=0.004), 80% vs. 57% at 6 months (p=0.047), 73% vs. 47% at 2 years (p<0.05), 80% (24/30) vs. 34% (9/26) at 10 years (p=0.0006) Sciatica moderately improved or pain-free (patient-rated): 83% vs. 50% at 6 weeks (NS), 83% vs. 60% at 6 months (p=0.038), 77% vs. 47% at 2 years (p<0.05), 77% (23/30) vs. 38% (10/26) at 10 years (p=0.0004) Back pain moderate improved or pain-free (patient-rated): 70% vs. 53% at 6 weeks (NS), 77% vs. 50% at 6 months (p=0.23), 73% vs. 43% at 2 years (p<0.05), 77% (23/30) vs. 38% (10/26) at 10 years (p=0.004) Sciatica moderately improved or pain-free (investigator-assessed): 77% vs. 53% at 6 months (p=0.052), 77% vs. 47% at 2 years (p<0.05), 77% (23/30) vs. 38% (10/26) at 10 years Subsequent surgery: 17% (5/30) vs. 37% (11/30) at 6 months, 20% (6/30) vs. 40% (12/30) at 2 years, 20% (6/30) vs. 47% (14/30) at 10 years
Quality score* 8/11
7/11
5/11
EVIDENCE REVIEW
Table 83. Randomized, placebo-controlled trials of chemonucleolysis
Sample size Type of LBP Duration of symptoms n=108 Radiculopathy Duration not reported Schwestschenau, 1976162 n=68 Radiculopathy Mixed duration Mean 20 to 25 weeks
Author, year Javid, 1983144
Main results Chemonucleolysis with chymopapain vs. intradiscal saline Success (composite outcome): 75% (41/55) vs. 45% (24/53) at 6 weeks, p=0.003 Overall response at least "fair" (patient rated): 85% (47/55) vs. 55% (29/53) at 6 weeks, p<0.001 Overall response at least "fair" (physician rated: 80% (44/55) vs. 47% (25/53) at 6 weeks, p<0.001 Chemonucleolysis with chymopapain vs. intradiscal saline Overall outcome good or excellent: 29% (9/31) vs. 31% (11/35) at mean 20 to 25 weeks follow-up (p=0.21), 29% (9/31) vs. 37% (13/35) at 1 year Returned to full activity within 3 months: 29% (9/31) vs. 26% (9/35) Surgery rate: 32% (10/31) vs. 46% (16/35) at 1 year
Quality score* 9/11
6/11
Efficacy of chemonucleolysis versus standard discectomy for lumbar disc prolapse with radiculopathy
The Cochrane review81, 82 included five lower-quality trials (total number of subjects 680) of chymopapain chemonucleolysis versus standard surgical discectomy869, 870, 874, 875, 878. It found consistent trends towards poorer results with chemonucleolysis, though most differences did not reach statistical significance. In addition, some between-study heterogeneity was present, and outcomes were inconsistently reported. At one year, patient randomized to chemonucleolysis were more likely to report overall outcomes as unchanged or worse compared to those randomized to placebo (2 trials, OR=1.64, 95% CI 0.81 to 3.33), and surgeons were also more likely to rate outcomes as poor (3 trials, OR=2.70, 95% CI 0.95 to 7.69). Chemonucleolysis was associated with a much higher rate of subsequent surgery compared to the rate of repeat surgery in patients who underwent initial discectomy (4 trials, OR=14.29, 95% CI 5.56 to 50). About 30% of patients who received chemonucleolysis subsequently underwent lumbar disc surgery within two years.
Efficacy of chemonucleolysis versus other interventions for lumbar disc prolapse with radiculopathy
One lower-quality trial not included in the Cochrane review found no significant differences after one year between patients randomized to chymopapain chemonucleolysis or spinal manipulation, though short-term outcomes (through six weeks) favored manipulation (Table 84)103.
EVIDENCE REVIEW
Table 84. Trial of chemonucleolysis versus spinal manipulation

Author, year Burton, 2000103
Main results Chemonucleolysis vs. spinal manipulation (mean improvement from baseline at 12 months) Leg pain (0 to 10): -1.38 vs. -1.87 (NS) Back pain (0 to 10): -1.18 vs. -1.52 (NS) RDQ score: -4.68 vs. -6.03 (NS)
Quality score* 3/9
* Excludes criteria involving blinding of patients and care providers, for maximum score of 9
Three lower-quality trials that compared chemonucleolysis and intradiscal steroid injections are reviewed in the section on intradiscal steroids109, 865, 866. None reported any differences between interventions.
Efficacy of chemonucleolysis versus automated percutaneous discectomy, endoscopic discectomy, or microdiscectomy for lumbar disc prolapse with radiculopathy
One lower-quality trial included in the Cochrane review found chemonucleolysis associated with a greater likelihood of success at one year compared to automated percutaneous discectomy (OR=2.26, 95% CI 1.17 to 4.37)876. Another small (n=22), lower-quality trial included in the Cochrane review found no clear differences between chymopapain chemonucleolysis and automated percutaneous discectomy on ODI scores and neurologic symptoms, though outcomes were poorly reported873.
Efficacy of low-dose chymopapain chemonucleolysis plus endoscopic discectomy versus endoscopic discectomy alone for lumbar disc prolapse with radiculopathy
A lower-quality trial not included in the Cochrane review found no clear differences between low-dose chymopapain chemonucleolysis plus endoscopic discectomy versus endoscopic discectomy alone through two years of follow-up, except for a slightly lower rate of recurrent herniation with combination therapy (6.9% vs. 1.6%, p=0.045)110. Pain and McNab scores were similar in the two groups.
EVIDENCE REVIEW
Table 85. Trial of chemonucleolysis plus endoscopic discectomy versus endoscopic discectomy alone
Author, year Hoogland, 2006110
Main results Low-dose chymopapain chemonucleolysis plus transforaminal posterolateral endoscopic discectomy vs. endoscopic discectomy alone (mean improvement from baseline) Leg pain (0 to 10 scale): 6.4 vs. 6.3 at 1 year, 6.37 vs. 6.03 at 2 years Back pain: (0 to 10 scale): 5.7 vs. 5.7 at 1 year, 5.35 vs. 5.6 after 2 years McNab result 'excellent': 62.7% vs. 50.8% McNab 'excellent' or 'good': 89.9% vs. 84.6% Recurrent herniation: 1.6% vs. 6.9% (p=0.045) in first year after surgery
Quality score 3/11
Efficacy of different chemonucleolysis methods

One trial included in the Cochrane review found no differences between low- and standard dose chymopapain chemonucleolysis868. Another trial found no differences between chemonucleolysis with chymopapain versus chemonucleolysis with collagenase872. One lowerquality trial not included in the Cochrane review evaluated long-term (five year) outcomes following chemonucleolysis with chymopapain or collagenase122. It found a greater proportion of good or excellent results in the chymopapain group (72% vs. 52%) using the McNab criteria, with much of this difference due to the proportion of patients proceeding to surgery (18% vs. 28%), who were considered failures (Table 86). However, improvements in pain scores were similar in the two groups (-7.8 vs. -7.7 on a 10 point scale).
Table 86. Additional trial not included in Cochrane review on efficacy of chemonucleolysis with chymopapain versus collagenase
Author, year Wittenberg, 2001122
Main results Chemonucleolysis with chymopapain vs. collagenase "Good" or "excellent" result at 5 years (with patients requiring surgery considered poor results): 72% vs. 52% Leg pain score, mean improvement (0 to 10 scale): -7.6 vs. -7.7 Required surgery: 18% vs. 28%
Quality score 4/11
Harms Earlier trials reported allergic reactions (including anaphylaxis) in 1.5% to 2% of patients who received chymopapain chemonucleolysis127, 879, 880. Estimates of allergic reactions may vary depending on how allergic reactions are assessed and defined and may be decreased by use of a lower test dose first. A more recent trial reported 12% of patients in the chymopapain arm experienced allergic reactions (flushing and itching), including one case of slight anaphylaxis122.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Rare serious complications that have been reported following chemonucleolysis include lumbar subarachnoid hemorrhage and paraplegia881, 882. Costs We identified two studies of costs associated with chemonucleolysis but excluded them because they used unreliable cost and outcomes data from a single observational study883, 884.
Summary of evidence
For prolapsed lumbar disc with radiculopathy, chemonucleolysis with chymopapain was moderately superior to placebo in five trials (four higher-quality) (level of evidence: good). There is insufficient evidence to accurately judge efficacy of chemonucleolysis with collagenase compared to placebo (one lower-quality trial) (level of evidence: poor). For prolapsed lumbar disc with radiculopathy, chemonucleolysis was consistently associated with trends towards worse outcomes compared to standard discectomy in five lower-quality trials, with about 30% of patients who underwent chemonucleolysis going on to discectomy (level of evidence: fair). For prolapsed lumbar disc with radiculopathy, chemonucleolysis with chymopapain and intradiscal steroid injections were consistently associated with similar outcomes in three lowerquality trials (level of evidence: fair). One lower-quality trial found no differences between chemonucleolysis with chymopapain and spinal manipulation after one year, though manipulation was superior at short-term (through 6 weeks) follow-up (level of evidence: poor). For prolapsed lumbar disc with radiculopathy, two lower-quality trials found inconsistent evidence on efficacy of chymopapin chemonucleolysis versus automated percutaneous discectomy, with one trial finding chemonucleolysis superior and the other finding no differences in functional status scores or rates of neurologic deficits (level of evidence: poor). One lower-quality trial found low-dose chymopapain chemonucleolysis plus transforaminal posterolateral endoscopic discectomy associated with a slightly lower rate of recurrent herniation compared to endoscopic discectomy alone, but there were no differences on other outcomes (level of evidence: poor). Chymopapain and collagenase chemonucleolysis were associated with similar pain outcomes in two lower-quality trials (one with five year follow-up), but chymopapain was associated with a trend towards reduced rate of subsequent surgery in one of the trials (level of evidence: fair). Chemonucleolysis with chymopapain is associated with allergic reactions in up to 12% of patients, though reporting of allergic reactions in trials was suboptimal. Serious complications (including anaphylaxis) with chymopapain appear uncommon and may be reduced by using lower or test doses or using collagenase (level of evidence: poor).
EVIDENCE REVIEW

The AHCPR guidelines recommend chymopapain as an acceptable treatment for patients with a herniated disc, severe, disabling sciatica, evidence of nerve root compromise, and persistence after at least one month of therapy, though it is less efficacious than standard or microdiscectomy. Testing patients for chymopapain allergic sensitivity could reduce the incidence of anaphylaxis (strength of evidence: C).
Radiofrequency denervation, intradiscal electrothermal therapy (IDET), and related procedures Radiofrequency denervation
Radiofrequency denervation is the destruction of nerves using heat generated by a radiofrequency current. It involves the placement of a catheter or electrode near or in the target nerve. Once the position of the catheter is confirmed by fluoroscopy, a radiofrequency current is applied in order to heat and coagulate adjacent tissues, including the target nerve. Radiofrequency denervation has been evaluated for treatment of presumed facet joint pain (target nerve medial branch of the primary dorsal ramus), presumed discogenic back pain (ramus communicans), and radicular back pain (dorsal root ganglia). Results of search: systematic reviews We identified a higher-quality Cochrane review (seven trials, six rated higher-quality) on efficacy of radiofrequency denervation for chronic low back pain90, 91. We also identified four other systematic reviews75, 78, 92, 93. One was rated higher-quality78. We excluded two earlier versions171, 185 of one of the systematic reviews75 and one other systematic review because it focused on technical aspects and did not evaluate efficacy180. Results of search: trials Nine randomized trials evaluated radiofrequency denervation108, 117, 118, 120, 121, 139, 149, 166, 885. Four trials were included in at least one of five systematic reviews75, 78, 90-93 and we identified five additional trials108, 117, 118, 120, 121. Eight of nine trials were placebo-controlled108, 117, 118, 120, 121, 139, 149, 166 .
Efficacy of radiofrequency denervation of the medial branch of the primary dorsal ramus versus sham or placebo for facet joint pain
For presumed facet joint pain, six placebo-controlled trials of radiofrequency denervation are difficult to interpret (Table 87)117, 120, 121, 139, 149, 166. The only trial (n=40) to use controlled facet joint blocks to select patients and an ablation technique believed to be optimal180 found radiofrequency denervation superior to sham treatment by -1.4 to -1.6 points (0 to 10 VAS scale) for improvement in generalized, back, and leg pain after 6 months, but the difference was not statistically significant for back pain (the main symptom thought to be associated with facet pain)117. In addition, baseline pain scores in the radiofrequency denervation group averaged 1.6 points higher (p<0.05 for differences) than in the sham group, which suggests unsuccessful randomization and could be associated with regression to the mean or differential potential for improvement. Furthermore, final pain scores in both groups were identical. Three other trials
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain met criteria to be classified as higher-quality but used uncontrolled diagnostic facet joint blocks to select patients, may have used suboptimal techniques180, 886, 887, and reported conflicting results121, 149, 166. One trial (n=30) found radiofrequency denervation associated with moderately greater improvement in mean VAS pain (-2.4 vs. -0.4 on a 0 to 10 scale, p<0.05) and ODI scores (-11.1 vs. +1.7, p<0.05) versus sham through 2 months166. Radiofrequency denervation was also associated with greater likelihood of experiencing at least a 2 point reduction in VAS pain score and greater than 50 percent improvement in global effect at 8 weeks (67% vs. 37.5%, p=0.003) and 12 months (46.7% vs. 12.5%, p=0.02). The second trial (n=70) found radiofrequency denervation superior to sham treatment for mean improvement in RDQ scores at four weeks (-8.4 vs. -2.2, p=0.05), but there were no statistically significant differences in ODI or VAS pain scores149. At twelve weeks, the difference in RDQ scores was no longer present. The third trial (n=82) found no differences between radiofrequency and sham intervention on any outcome121.
EVIDENCE REVIEW
Table 87. Randomized, sham-controlled trials of radiofrequency denervation for presumed facet joint pain
Sample size Type of LBP Duration of symptoms n=41 Presumed facet joint pain with positive uncontrolled facet joint block Leclaire, 2001149 Chronic n=70 Presumed facet joint pain with positive uncontrolled facet joint block Nath, 2008117 Chronic n=40 Presumed facet joint pain with positive controlled facet joint blocks Chronic 6 months 3 months
Author, year Gallagher, 1994139
Main results Radiofrequency denervation vs. sham denervation Unable to interpret changes in VAS pain scores and McGill Pain Questionnaire, no intention-totreat analysis and baseline differences in pain scores
Quality score* 3/11
Radiofrequency denervation vs. sham denervation (mean difference in change from baseline, positive values favor radiofrequency denervation) RDQ (transformed to 0 to 100 scale): 6.2 (CI, -1.3 to 13.8, p=0.05) at 4 weeks, 2.6 (CI, -6.2 to 11.4) at 12 weeks ODI (0 to 100): 0.6 (CI, -4.5 to 5.7, NS) at 4 weeks, 1.9 (CI, -3.2 to 7.0) at 12 weeks Pain (0 to 100): 4.2 (CI, -6.9 to 15.4) at 4 weeks, -7.6 (CI, -20.3 to 5.1) at 12 weeks Radiofrequency denervation vs. sham denervation, changes from baseline Generalized pain (0 to 10 VAS): -1.9 vs. -0.4, p=0.02 Back pain (0 to 10 VAS ): -2.1 vs. -0.7, p=0.08 Leg pain (0 to 10 VAS): -1.6 vs. -0.1, p=0.046 Analgesic consumption (6 point scale): -1.40 vs. -0.60, p=0.04 Walking (6 point scale): -0.40 vs. -0.40, p=1.0 Sitting (6 point scale): -0.75 vs. -0.15, p=0.04 Sleep (6 point scale): -0.65 vs. -0.35, p=0.20 Standing (6 point scale): -1.00 vs. -0.25, p=0.04 Work (6 point scale): -1.60 vs. -0.15, p=0.004 Subjective global assessment (6 point scale): -1.1 vs. -0.30, p=0.004 Pulsed radiofrequency denervation vs. conventional radiofrequency denervation vs. sham denervation Pain, mean VAS score (0 to 10): 2.9 vs. 2.3 vs. 3.1 at 6 months (p<0.05 for sham versus pulsed or conventional denervation); 3.5 vs. 2.4 vs. 3.9 at 1 year (p<0.05 for conventional vs. pulsed or sham) ODI, mean score (0 to 100): 25 vs. 25 vs. 29 at 6 months (p<0.05 for conventional vs. sham) and 28 vs. 28 vs. 34 at 1 year (p<0.05 for pulsed or radiofrequency denervation vs. sham) Patient satisfaction good or excellent: 85% vs. 95% vs. 70% (p=0.03 for sham vs. denervation groups) Analgesic use: 75% vs. 40% vs. 95% (p not reported)
9/11
8/11
Tekin, 2007120
n=60 Presumed facet joint pain (clinical criteria only) Chronic
1 year
5/11
EVIDENCE REVIEW
Table 87. Randomized, sham-controlled trials of radiofrequency denervation for presumed facet joint pain
Sample size Type of LBP Duration of symptoms n=30 Presumed facet joint pain with positive uncontrolled facet joint block Chronic
Author, year van Kleef, 1999166
van Wijk, 2005121
n=81 Presumed facet joint pain with positive uncontrolled facet joint block Chronic
3 months
Main results Radiofrequency denervation vs. sham denervation (difference in change from baseline at 8 weeks) VAS-mean (0 to 10 scale): unadjusted 1.94 (CI, 0.24 to 3.64, p<0.05); adjusted 2.46 (CI, 0.72 to 4.20, p<0.05) Global perceived effect (-3 to +3 scale): unadjusted -0.96 (CI, -1.70 to -0.22, p<0.05); adjusted -1.10 (CI, -1.89 to -0.30, p<0.05) Physical impairment (Waddell, 0 to 7 scale): unadjusted 0.27 (CI, -0.69 to 1.22, NS); adjusted 0.31 (CI, -0.74 to 1.35, NS) Analgesic tablets per 4 days: unadjusted 3.88 (CI, 1.19 to 6.57, p<0.05); adjusted 3.24 (CI, -0.13 to 6.60, NS) ODI (0 to 100): unadjusted 15.75 (CI, 4.16 to 21.35, p<0.01); adjusted 10.90 (CI, 1.76 to 20.0, p<0.05) Quality of life (COOP/WONCA, 0 to 35): unadjusted 1.51 (CI, -1.85 to 4.97, NS); adjusted 2.27 (CI, -1.77 to 6.30, NS) Treatment success (2 point reduction in VAS-mean or VAS-high and >50% global perceived effect): 67% vs. 38% at 8 weeks (OR unadjusted 3.33, CI 0.97 to 11.5; OR adjusted 9.53, CI 1.50 to 60.5); 47% vs. 12% at 12 months Radiofrequency denervation vs. sham injection Clinical success (defined as at least 50% improvement in VAS-leg score, without drop in daily activities score or rise in analgesics rating scale, or improvement of at least 2% in VAS-leg score, daily activities score, and analgesic use score) at 3 months: 28% vs. 29% (p=0.86) Leg pain, change in VAS (0-10) score: -1.1 vs. -0.7 (NS) Back pain, change in VAS (0-10) score: -2.1 vs. -1.6 (NS) Change in daily activities: 1.5 vs. 0.9 (NS) Change in analgesics use: -0.1 vs. -0.2 (NS)
Quality score* 7/11
11/11
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain A lower-quality trial (n=60) found conventional but not pulsed radiofrequency denervation superior to sham denervation for pain, the ODI, and analgesic use through 1 year120. Effects on pain were small to moderate (0.8 to 1.5 points on a 0 to 10 scale) and on the ODI were small (4 to 6 points). Another sham-controlled trial had serious methodological shortcomings, including lack of intention-to-treat analysis139. Two higher-quality78, 90, 91 and two lower-quality92, 93 systematic reviews also found uncertain or inconsistent benefits associated with radiofrequency denervation for presumed facet joint pain, though none included the three117, 120, 121 most recently published sham-controlled trials. A fifth systematic review concluded there is moderate evidence supporting benefits from radiofrequency denervation75. It excluded a higher-quality trial149 with more neutral findings because it used a single block to identify facet joint pain, leaving only a single, small (n=31) higher-quality randomized trialwhich also did not appear to use controlled blocks to select patientsdemonstrating benefits166. Although it included observational studies, criteria for differentiating positive from negative trials were poorly defined (results were considered positive if the treatment was effective by defined criteria [e.g., 50% pain relief] for the designated period of time). Three of the ten observational studies included in this review found that fewer than 50% of patients experienced pain relief.
Efficacy of radiofrequency denervation of the ramus communicans nerve versus sham or placebo for presumed discogenic back pain
One small (n=49), lower-quality trial of patients with presumed discogenic back pain (nonradicular back pain with positive discography) who had failed IDET found radiofrequency denervation of the ramus communicans nerves associated with substantially better mean VAS pain scores (3.8 vs. 6.3 on a 0 to 10 scale, p<0.05), and moderately better SF-36 bodily pain (43.7 vs. 32.4, p<0.05) and physical function scores (58.9 vs. 46.5, p<0.05) compared to lidocaine injection after 4 months (Table 88)118.
Table 88. Randomized, sham-controlled trials of radiofrequency denervation for presumed discogenic back pain
Sample size Type of LBP Duration of symptoms n=49 Presumed discogenic back pain Chronic
Author, year Oh, 2004118
Main results Radiofrequency denervation vs. lidocaine injection Pain, mean VAS (0-10) score at 4 months: 3.8 vs. 6.3 (p<0.05) SF-36 bodily pain subscale: 43.7 vs. 32.4 (p<0.05) SF-36 physical function subscale: 58.9 vs. 46.5 (p<0.05) 77% of patients in radiofrequency denervation group decreased analgesics by at least 50%
Quality score* 5/11
EVIDENCE REVIEW
Efficacy of radiofrequency denervation versus sham or placebo for radicular low back pain
One higher-quality trial of patients with chronic (>6 months) radicular pain and a positive selective nerve root block found no difference between radiofrequency denervation of the dorsal root ganglia and sham treatment for the proportion with clinical success (16% vs. 25%, p=0.43), SF-36 scores, or use of analgesics (Table 89)108. There was a trend towards a higher proportion of patients in the sham intervention group reporting >50% reduction in VAS-pain scores for the leg (21% vs. 42%, p=0.051). Out of 1001 patients originally evaluated for potential inclusion, only 83 were enrolled.
Table 89. Randomized, sham-controlled trial of radiofrequency denervation for radiculopathy
Sample size Type of LBP Duration of symptoms n=83 Radiculopathy Chronic
Author, year Geurts, 2003108
Main results Radiofrequency denervation vs. sham injection Clinical success (see definition in van Wijk above): 16% vs. 25% (p=0.43) Leg pain, change in VAS (0-10) score: -0.7 vs. -2.0 (p=0.02) Back pain, change in VAS (0-10) score: -0.6 vs. -1.1 (p=0.32) Change in daily activities: -0.5 vs. -0.4 (p=0.85) Change in analgesics use: 0.1 vs. -0.2 (p=0.23)
Quality score* 11/11
Efficacy of intra-articular versus extraarticular radiofrequency denervation for presumed facet joint pain
One small (n=34), lower-quality trial885 included in one systematic review78 found extra-articular radiofrequency denervation substantially inferior to intra-articular radiofrequency denervation on mean pain scores and the ODI. However, baseline differences in ODI scores appeared to be present. No other RCT has evaluated intra-articular radiofrequency denervation. Harms One trial reported a case of subjective mild lower limb weakness following radiofrequency denervation for presumed discogenic back pain that resolved within two weeks118. In two trials of patients with presumed facet joint pain, adverse events did not differ between treatment and sham radiofrequency denervation, though there was a trend towards a higher rate of increased pain following true radiofrequency denervation108, 121. Costs We found no studies evaluating costs.
Summary of evidence
For presumed facet joint pain, evidence on efficacy of radiofrequency denervation of the medial branch of the primary dorsal ramus is difficult to interpret. The only trial (n=60) to use
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain controlled facet joint blocks to select patients and an ablation technique believed to be optimal found radiofrequency denervation to be moderately superior to sham denervation, but baseline differences between groups could invalidate results. Two of three other small (n=30 to 81), higher-quality trials showed no benefits of radiofrequency denervation compared to sham denervation. Interpretation of these results is controversial because these trials used uncontrolled facet joint blocks to select patients and the radiofrequency denervation technique may have been suboptimal in some of the trials (level of evidence: poor). For presumed facet joint pain, intra-articular radiofrequency denervation was superior to extraarticular radiofrequency denervation in one small trial. No other trial evaluated efficacy of intra-articular radiofrequency denervation (level of evidence: poor). For chronic radicular pain and a positive selective nerve root block, radiofrequency denervation of the dorsal root ganglion was not effective compared to sham in one small, higher-quality trial (level of evidence: poor). For presumed discogenic low back pain with positive discography, radiofrequency denervation of the ramus communicans nerve was moderately to substantially superior to sham denervation in one small, lower-quality trial (level of evidence: poor). Adverse events were poorly reported, but serious adverse events were not described in the trials following radiofrequency denervation.

The European COST guidelines found insufficient evidence to recommend radiofrequency denervation of dorsal root ganglion for chronic low back pain.
Intradiscal electrothermal therapy (IDET)

Intradiscal electrothermal therapy (IDET) involves the placement of an electrode or catheter into the intervertebral disc annulus or nucleus. The catheter is then slowly heated and kept at a predetermined temperature for a predetermined time. This is thought to treat pain by altering the biomechanics of the disc (possibly by shrinking collagen fibers) and/or destroying adjacent nociceptive pain receptors. It is used in patients with presumed discogenic back pain. We considered IDET and a similar procedure that uses radiofrequency energy rather than thermal energy, percutaneous intradiscal radiofrequency thermocoagulation (PIRFT), separately (see below). Results of search: systematic reviews We identified one higher-quality Cochrane review of IDET for chronic low back pain79, 80. We also identified one other higher-quality systematic review99 and three lower-quality systematic reviews72, 73, 87. We excluded two older versions of the Cochrane review176, 255 and three review articles that were not systematic174, 179, 196. Results of search: trials Two higher-quality, sham-controlled, randomized trials evaluated IDET138, 157. The five previously published systematic reviews each included one73, 87 or both72, 79, 80, 99 of the trials.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain One lower-quality, non-randomized prospective cohort study203 comparing IDET to PIRFT was also included in three systematic reviews72, 73, 99. We excluded two non-randomized trials198, 203.
Efficacy of intradiscal electrothermal therapy versus sham

For chronic low back pain with positive response to provocative lumbar discography, two small (n=57 and n=64), higher-quality trials of IDET versus sham IDET reported conflicting results (Table 90)157, 888. In one trial, IDET was associated with moderately greater improvements in mean VAS pain scores (0-10 scale, mean change 2.4 vs. 1.1, p=0.0045) and slightly greater improvements in mean ODI scores (0 to 100 scale, mean change 11 vs. 4, p=0.050) compared to sham IDET, but was no better on the SF-36 bodily pain or physical functioning subscales157. The proportion of patients with at least a two-point improvement in VAS pain scores also favored IDET (56% or 18/32 compared to 38% or 9/24). From a potential cohort of 4253 who were assessed for trial eligibility, 64 patients were enrolled. The other higher-quality trial (n=64) found no differences between IDET and sham IDET on the Low Back Pain Outcome Score, ODI, SF-36, or Zung Depression Index888.
Table 90. Randomized, sham-controlled trials of intradiscal electrothermal therapy
Sample size Type of LBP Duration of symptoms n=57 Presumed discogenic back pain with positive provocative discography Chronic
Author, year Freeman, 2005138
Pauza, 2004157
n=64 Presumed discogenic back pain with positive provocative discography Chronic
6 months
Main results IDET vs. sham IDET, difference in mean improvement from baseline through 6 months Low Back Outcome Score: -1.708, p=0.111 ODI: -2.156, p=0.489 Zung: -0.873 , p=0.693 MSPQ: -0.873, p=0.945 SF-36, physical functioning: 1.044, p=0.819 SF-36, bodily pain index: -1.997, p=0.659 Low back pain outcome score improved >7 points: 0% vs. 0% SF-36 Physical Functioning and Bodily Pain Index improved >1 standard deviation: 3/36 (8.3%) vs. 3/19 (15.8%) IDET vs. sham IDET VAS for pain (0-10), mean change: 2.4 vs. 1.1, p=0.0045 SF-36, bodily pain (0-100), mean change: 17 vs. 9, p=0.086 SF-36, physical functioning (0-100), mean change: 15 vs. 11, p=0.548 ODI (0-100), mean change: 11 vs. 4, p=0.050 Pain improved by >2.0 on VAS: 18/32 (56%) vs. 9/24 (38%) Pain improved by >75%: 7/32 (22%) vs. 1/24 (4.2%)
Quality score* 8/11
8/11
Two higher quality79, 80, 99 and one lower-quality87 systematic review also found inconsistent data on efficacy of IDET. Two other lower-quality systematic reviews concluded that IDET is effective, largely based on pooled rates of response to IDET from mostly observational
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain studies72, 73. In the only controlled observational study included in these reviews, IDET was associated with substantially better VAS pain scores at 3 months (3.5 vs. 8.0 on a 0 to 10 scale, p<0.0005) and 24 months (3.0 vs. 7.5, p=0.028), as well as a higher proportion pain-free at 24 months (20% or 7/35 vs. 0% or 0/17)198. The other observational studies included in these reviews were uncontrolled.
Efficacy of intradiscal electrothermal therapy versus percutaneous intradiscal radiofrequency ablation

One lower-quality, small (n=42), prospective cohort study found IDET substantially superior to PIRFT for improvement in pain (mean difference -21.8 on a 0 to 100 scale) at 1 year203. This study did not meet inclusion criteria because it was not a randomized trial. In addition, differences did not become statistically significant until 3 months after the procedure, and some baseline differences were present. Harms Most studies of IDET reported transient and mild complications ranging in incidence from 0% (0/58) to 15% (5/33)89. These included increased radicular pain (5/33), paresthesias and numbness (2/79), and foot drop (1/79). In one study, one patient developed a CSF leak889. There have also been case reports of cauda equina syndrome and vertebral osteonecrosis89. In one systematic review73, five of seventeen studies (including observational data) did not mention complications at all, and in another, eleven of fourteen observational studies reported no periprocedural complications72. Rates of complications in the other three studies ranged from 9% to 16% but were primarily minor. Costs We found no studies evaluating costs.
Summary of evidence
For chronic low back pain with positive provocative discography, there is conflicting evidence from two higher-quality trials on efficacy of IDET relative to sham IDET. In the one trial finding benefits from IDET, effects were moderate for pain relief and small for functional status in a highly selected population (level of evidence: poor). For chronic low back pain with positive provocative discography, no trial of IDET versus PIRFT met inclusion criteria. One small, non-randomized study found IDET superior to PIRFT for improvement in pain, but differences were not statistically significant until 3 months after the procedure, and some differences in baseline pain scores were present (level of evidence: poor). Periprocedural complications associated with IDET were poorly reported but generally appeared mild or transient, though there are case reports of cauda equina syndrome and vertebral osteonecrosis after IDET (level of evidence: poor).
EVIDENCE REVIEW

The European COST guidelines found insufficient evidence to recommend IDET for nonspecific or discogenic chronic low back pain.
Percutaneous intradiscal radiofrequency thermocoagulation (PIRFT) and Coblation nucleoplasty

Like IDET, percutaneous intradiscal radiofrequency thermocoagulation (PIRFT) involves the insertion of an electrode or catheter into an intervertebral disc presumed to be the source of low back pain. Unlike IDET, the electrode or catheter itself does not become hot. Instead, heat is generated in surrounding tissues by an alternating radiofrequency current. Also, the catheter is placed into the center of the disc rather than around the annulus. Like IDET, PIRFT is not intended to shrink or coagulate tissue, but is thought to work by altering the biomechanics of the disc or nociceptive nerve fibers. Coblation nucleoplasty uses a bipolar radiofrequency current to create a series of channels in the target disc. Unlike IDET and PIRFT, the goal of Coblation nucelopasty is tissue reduction. Coblation nucleoplasty has been used both for treatment of contained lumbar disc prolapse as well as presumed discogenic low back pain. Results of search: systematic reviews Two different higher-quality Cochrane reviews evaluated efficacy of PIRFT for chronic low back pain79, 80, 90, 91. We identified one other higher-quality systematic review99 and one lower-quality systematic review of PIRFT87, and one lower-quality systematic review of Coblation nucleoplasty88. Results of search: trials Two randomized trials124, 890 evaluated PIRFT. Both evaluated PIRFT for chronic low back pain and were included in previously published systematic reviews. Only one was a sham-controlled trial124. One excluded non-randomized study that compared IDET and PIRFT is discussed in the section on IDET203. We identified no trials of Coblation nucleoplasty.
Efficacy of PIRFT versus sham therapy for presumed discogenic low back pain
For chronic, presumed discogenic low back pain (based on a positive response to analgesic discography), one small (n=28), higher-quality randomized trial124 found no significant differences between PIRFT and sham PIRFT for improvement in VAS pain scores, global effect, ODI, or proportion of treatment success, defined as the number of patients with a 2-point reduction on a 10 point VAS pain scale and >50% pain reduction on global perceived effect (1/13 in active treatment group and 2/15 in sham group) (Table 91)124. A second trial compared two different durations of radiofrequency thermocoagulation890. It found no differences and minimal improvement with either intensity of PIRFT.
EVIDENCE REVIEW
Table 91. Randomized, sham-controlled trials of percutaneous intradiscal radiofrequency thermocoagulation for presumed discogenic back pain
Sample size Type of LBP Duration of symptoms n=28 Presumed discogenic back pain with positive provocative discography Chronic
Author, year Barendse, 2001124
Duration of follow-up 1 year
Main results PIRFT vs. sham Proportion classified as 'success' at 8 weeks: 1/14 vs. 2/14 (AOR 1.1, 0.04 to 33.3) Proportion classified as 'success' at 1 year: 1/14 vs. 0/14 Change in VAS: -0.61 vs. -1.14 (NS) Change in global perceived effect: 0.09 vs. 0.21 (NS) Change in Waddell impairment: 0.00 vs. 0.29 (NS) Change in number of analgesic tablets per 4 days: -1.38 vs. 0.43 (NS) Change in ODI: -2.62 vs. -4.93 (NS) Change in Coop/Wonca: -1.85 vs. -0.21 (NS)
Quality score* 10/11
Efficacy of Coblation nucleopasty for presumed discogenic low back pain or contained lumbar disc prolapse
We identified no relevant trials of Coblation nucleoplasty88. Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
For presumed discogenic back pain, one small, higher-quality trial found no differences between PIRFT and sham PIRFT (level of evidence: poor). For chronic low back pain with positive provocative discography, no trial of IDET versus PIRFT met inclusion criteria. One small, non-randomized study found IDET superior to PIRFT for improvement in pain, but differences were not statistically significant until 3 months after the procedure, and some differences in baseline pain scores were present (level of evidence: poor). There is insufficient data to judge harms associated with PIRFT. There is insufficient data (no trials) to judge efficacy of Coblation nucleoplasty.

The European COST guidelines found insufficient evidence to recommend PIRFT for nonspecific or discogenic chronic low back pain.
EVIDENCE REVIEW
Spinal cord stimulation

Spinal cord stimulation involves the placement of electrodes in the epidural space adjacent to the area of the spine presumed to be the source of pain85. An electric current is then applied to achieve sympatholytic and other neuromodulatory effects. The number and type of electrode leads and parameters of electrical stimulation can vary. Electrodes may be implanted percutaneously or by laminectomy, and power for the spinal cord stimulator is supplied by an implanted battery or transcutaneously through an external radiofrequency transmitter. Typically, a trial of spinal cord stimulation is attempted, with permanent implantation of the device only in patients who respond to the trial. Spinal cord stimulation has been most frequently studied in patients with failed back surgery syndrome (see Key Question 11), but is also used for chronic back pain no associated with prior surgery. Results of search: systematic reviews We identified one higher-quality systematic review (reported in two publications) on efficacy of spinal cord stimulation for chronic back and leg pain95, 96. Two other recent systematic reviews only included studies of spinal cord stimulation for failed back surgery syndrome (see Key Question 11)85, 98. We excluded one outdated systematic review192 and one review that did not use systematic methods195. Results of search: trials and observational studies We identified no trials of spinal cord stimulation for low back pain without failed back surgery. The systematic review included 72 case series (mean duration of pain 6.5 years), 27 of which evaluated spinal cord stimulation for chronic back and leg pain without failed back surgery syndrome (median quality score 1 on a 1 to 7 scale)95, 96. None of the studies prospectively studied consecutive patients using independently assessed and validated outcomes measures. The systematic review did not report results separately for patients with or without failed back surgery syndrome.
Efficacy of spinal cord stimulation for chronic low back pain with leg pain
Based on case series, the systematic review reported overall pooled estimates for the proportion of patients with greater than 50% pain relief of 62% (95% CI 56-69%) shortly following spinal cord stimulator implantation and 48% (95% CI 43-53%) during follow-up testing95, 96. The percentage of patients that achieved pain relief was 15% to 20% lower in studies rated higher quality (4 or higher on a 7 point scale), was reduced by 5% for every additional 10 months of follow-up, was increased by 10% for multicenter compared to single center studies, and was 20% higher in studies of patients with failed back surgery syndrome or chronic leg and back pain than in studies of patients with other conditions96. The proportion of patients that didnt require an analgesic after implantation was 53% (95% CI 48-56%), the proportion returned to work 40% (95% CI 28-50%), and the proportion satisfied with the intervention 70% (95% CI 62-85%). Harms Only 18 of the 72 studies reported usable harms data96. Overall, 43% (48/112) of patients with chronic back and leg pain or failed back surgery syndrome experienced at least one
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain complication with spinal cord stimulation. The most frequent complication was related to electrode or lead problems (27%). Other complications included infections (6%), generator problems (6%), extension cable problems (10%) and other issues (such as cerebrospinal fluid leak in 7%). No neurologic-related adverse events were reported. Costs We found no studies evaluating costs.
Summary of evidence
For chronic back and leg pain or failed back surgery syndrome, lower-quality evidence from multiple case series estimated that approximately half of patients experienced decreased pain after spinal cord stimulator implantation, and about 40% returned to work (level of evidence: poor). Spinal cord stimulation is associated with frequent complications, especially related to electrode or lead problems. Although most complications appear minor, infections (6% of complications) and cerebrospinal fluid leak (7%) have been reported (level of evidence: poor).

The European guidelines found insufficient evidence to recommend spinal cord stimulation for chronic nonspecific low back pain.
EVIDENCE REVIEW
Table 92. Systematic reviews on efficacy of interventional therapies for low back pain
Number of RCTs not included in any other relevant systematic review Not applicable
Author, year Gibson, 200781,
Number of included RCTs (number rated higherquality) * 19 (6)**
Median duration of follow-up (range) 1 year (3 months to 10 years)
Median sample sizes in included RCTs (range) 80 (29 to 173)
Interventions evaluated (number of RCTs) Chymopapain (17), collagenase (3)
Main conclusions Chemonucleolysis with chymopapain vs. placebo for lumbar disc prolapse Patient rated outcome no success at 1 year (2 RCTs): OR 0.24 (95% CI 0.12 to 0.49) Surgeon rated outcome no success at 3 to 12 months (4 RCTs): OR 0.40 (95% CI 0.21 to 0.75) Further disc surgery within 6 to 24 months (5 RCTs): OR 0.41 (95% CI 0.25 to 0.68) Chemonucleolysis with chymopapain vs. discectomy for lumbar disc prolapse Patient rated outcome unchanged or worse at 1 year (2 RCTs): OR=1.64 (95% CI 0.81 to 3.33) Surgeon rated poor outcome at 1 year (3 RCTs): OR=2.70 (95% CI 0.95 to 7.69) Further disc surgery within 1 year (4 RCTs): OR=14.3 (95% CI 5.56 to 50.0)
Chemonucleolysis (19 RCTs in one systematic review)

82
EVIDENCE REVIEW
Number of RCTs not included in any other relevant systematic review 10
Author, year Abdi, 200771
Number of included RCTs (number rated higherquality) * 23 (18)
Median duration of follow-up (range) 6 months (2 weeks to 5 years)
Interventions evaluated (number of RCTs) Interlaminar (10), caudal (8), transforaminal (5)
Main conclusions Lumbar interlaminar epidural injections for lumbar radicular pain/sciatica (11 RCTs): Strong evidence (8 of 11 positive RCTs) for short-term relief and limited evidence (2 of 11 positive RCTs) for long-term relief. Indeterminate evidence for axial low back pain and lumbar spinal stenosis Lumbar transforaminal epidural injections for lumbar radicular pain/sciatica (5 RCTs): Strong evidence for short-term (4 of 5 RCTs) and moderate evidence for long-term relief (4 of 5 RCTs) of lumbar radicular pain. Lumbar caudal epidural injections for lumbar radicular pain/sciatica (6 RCTs): Strong evidence (4 of 6 RCTs) for short-term relief and moderate evidence (4 of 6 RCTs) for long-term relief. Strong evidence (2 of 2 RCTs) for short-term relief of post lumbar laminectomy syndrome. Epidural steroid injection vs. control (4 higher-quality RCTs): Epidural steroid injections may improve radicular lumbosacral pain 2-6 weeks after injection vs. control (Level C, Class I-III evidence), but no difference with longerterm follow-up through 1 year. Average magnitude of effect is small. No effect on functional improvement or need for surgery.
Epidural steroid injection (33 unique RCTs in 8 systematic reviews)
Armon, 200774
Qualitative
6 (3)
12 months (3 months to 5 years)
64 (23 to 160)
Interlaminar or not specified (3), caudal (1), transforaminal (2)
EVIDENCE REVIEW
Author, year DePalma, 2005

77
Median duration of follow-up (range) 8 months (3 months to 5 years)
Interventions evaluated (number of RCTs) Interlaminar (3), caudal (1), transforaminal (5)
Main conclusions Transforaminal epidural corticosteroid injection vs. local anesthetic or saline placebo for lumbosacral radiculopathy (3 RCTs, 1 higher-quality): 1 RCT found steroid superior to anesthetic for proportion proceeding to surgery (71% vs. 33%), 1 RCT found steroid superior to saline for overall response (McNab criteria) at 3 months (54% vs. 40%), 1 RCT found no difference between steroid and saline at 12 months (65% response) Transforaminal epidural corticosteroid injection vs. interlaminar epidural steroid injection (2 RCTs, neither rated higherquality): 1 of 2 RCTs found transforaminal superior to interlaminar Transforaminal epidural corticosteroid steroid injection vs. trigger point injection (1 RCT, lower-quality): Epidural superior for successful outcome at 12 months (84% vs. 48%) (trial mis-classified as randomized)
EVIDENCE REVIEW
Author, year Luijsterburg2007

84
Median duration of follow-up (range) 6 months (2 weeks to 24 months)
Interventions evaluated (number of RCTs) Interlaminar or not specified (10), caudal (2), transforaminal (2)
Main conclusions Epidural corticosteroid vs. placebo injection for lumbosacral radiculopathy: Short-term pain (7 RCTs, 4 higher-quality): Conflicting evidence, with 5 RCTS (3 higherquality) showing no difference and 2 RCTs (1 higher-quality) finding epidural steroid superior Longer-term pain (3 RCTs, 2 higher-quality): No difference (strong evidence) Short-term overall improvement (5 RCTs, 3 higher-quality): Conflicting evidence, with 4 of 5 RCTs (3 higher-quality) showing no difference Long-term overall improvement (3 RCTs, 2 higher-quality): No difference (strong evidence) Disability and return-to-work (3 higherquality RCTs): No difference at short or longer-term follow-up (strong evidence) Epidural corticosteroid injection vs. placebo for low back pain with or without sciatica: RR=0.92 (95% CI 0.76 to 1.11) for pain relief >6 weeks after injection (3 RCTs); RR=0.93 (95% CI 0.79 to 1.09) for pain relief <6 weeks after injection Epidural corticosteroid injection versus various non-placebo comparators for low back pain with or without sciatica (6 RCTs): 4 of 6 trials found non-significant positive effect; 1 of 6 found significant short-term positive effect; 0 of 2 long-term trials reported significant differences
Nelemans, 86 2001
11 (5)
4.5 months (3 hours to 24 months)
30 (20 to 158)
Interlaminar or not specified (8), caudal (3)
EVIDENCE REVIEW
Author, year Resnick, 2005

92
Median duration of follow-up (range) Range 2 months to 6 months
Median sample sizes in included RCTs (range) Range 20 to 35
Interventions evaluated (number of RCTs) Interlaminar (3), caudal (1)
Main conclusions Epidural corticosteroid injection vs. control injection for chronic low back pain without significant radiculopathy (4 RCTs): No clear differences except epidural corticosteroid + morphine more effective than epidural corticosteroid + saline at 6 weeks in one RCT Epidural corticosteroid injection vs. epidural indomethacin, midazolam, or morphine for low back pain without radiculopathy (3 RCTs): No difference for each comparison. Two of three trials evaluated postlaminectomy patients. Note: Two placebo-controlled trials enrolled patients with sciatica, though purpose of review was to evaluate efficacy of epidural injections for non-radicular low back pain. Epidural corticosteroid injection vs. placebo or epidural local anesthetic for sciatica (6 RCTs): 4 of 6 studies found epidural corticosteroid injection superior for up to 12 weeks Epidural corticosteroid injection vs. placebo or epidural local anesthetic for LBP of mixed etiologies (3 RCTs): 2 of 3 RCTs found epidural corticosteroid injection superior Epidural corticosteroid injection vs. placebo injection for sciatica: OR 2.2 (95% CI 1.04.7) for improvement (4 RCTs); when including 8 excluded RCTs with <20 subjects in an arm, OR 2.0 (95% CI 1.1-3.7)
Staal, 200894
Qualitative
5 (2)
Range 1 week to 6 months
Median 35 (24 to 206)
Interlaminar (4) Caudal (1)
Tonkovich97 Quaranta, 2000
Qualitative
9 (not rated)
3 months (1 week to 24 months)
48 (20 to 100)
Vroomen, 2000100
Quantitative
4 (4)
Range 3 to 14 months
Range 51 to 158
EVIDENCE REVIEW
Author, year Boswell, 200775
Median duration of follow-up (range) 9 months (6 months to 2.5 years)
Interventions evaluated (number of RCTs) Facet joint injection (2), medial branch block (2)
Main conclusions Lumbar facet joint corticosteroid injection for presumed lumbar facet joint pain (2 higherquality RCTs): Moderate evidence (2 of 2 positive RCTs) for short-and long-term improvement relative to control or baseline Lumbar medial branch (facet joint nerve) block for presumed lumbar facet joint pain (2 higher-quality RCTs): Moderate evidence (2 of 2 positive RCTs, one unpublished) for short-term and long-term improvement of lumbar facet joint pain relative to control or baseline Lumbar facet joint corticosteroid injection vs. control injection for presumed facet joint pain (3 RCTs): No clear difference between interventions in 3 RCTs Lumbar facet joint corticosteroid injection vs. saline injection, extra-articular steroid, or medial branch block (3 RCTs, one higherquality): No clear differences Lumbar facet joint corticosteroid injection vs. placebo injection for presumed lumbar facet joint pain (1 higher-quality and 1 lowerquality RCT): In both trials, no differences in pain or functional status through 3 months; conflicting evidence on longer-term effects. Lumbar facet joint corticosteroid injection vs. facet joint injection without corticosteroid, medial branch block, exercise alone, or facet joint injection with hyaluronidase (4 RCTs, 1 higher-quality): No differences
Facet joint injection or medial branch block (6 unique RCTs in 4 systematic reviews)
Resnick, 200592
Qualitative
3 (0)
3 months (3 to 6 months)
101 (89 to 109)
Facet joint injection (3), medial branch block (1) Facet joint injection (3), medial branch block (1) Facet joint injection (5), medial branch block (1)
Slipman, 200393
Qualitative
3 (1)
3 months (3 to 6 months)
101 (89 to 109)
Staal, 200894
Qualitative
5 (2)
3 months (5 weeks to 6 months)
89 (60 to 109)
EVIDENCE REVIEW
Author, year Gibson, 200581,
Median duration of follow-up (range) 6 months (both RCTs)
Median sample sizes in included RCTs (range) 60 and 80
Interventions evaluated (number of RCTs) Intradiscal corticosteroid injection (2) Intradiscal electrothermal therapy (2)
Main conclusions Intradiscal corticosteroid injection vs. chemonucleolysis (2 RCTs): No differences in 2 RCTs IDET for presumed discogenic LBP: Median 51% (range 22% to 71%) improvement in pain score (15 studies, including observational data); median 65% (range 52% to 72%) achieved at least 2 point improvement in 10 point pain scale (5 studies, including observational data) IDET vs. sham for presumed discogenic LBP (2 RCTs): Conflicting results, with 1 of 2 RCTs reporting no differences IDET for presumed discogenic LBP: Mean 2.9 (95% CI 2.5 to 3.4) improvement in pain intensity on a 0 to 10 scale (13 studies, including observational data), mean=7.0 (95% CI 2.0 to 11.9) improvement in ODI in 3 studies (including observational data) IDET vs. sham for presumed discogenic LBP (1 RCT): IDET superior to sham in one of two RCTs for pain and disability through 2 years, but it evaluated a highly selected population (64 of potential cohort of 4253 randomized)
Intradiscal corticosteroid injection (2 RCTs in 1 systematic review)

82
Intradiscal electrothermal therapy (2 unique RCTs in 5 systematic reviews) Andersson, 72 2006 Quantitative 2 (2) 0 6 months (both RCTs) 57 and 64 2
Appleby, 200673
Quantitative
1 (not rated)
6 months
64
Intradiscal electrothermal therapy (1)
80
Gibson, 200579,
Qualitative
2 (2)**
6 months (both RCTs)
57 and 64
EVIDENCE REVIEW
Author, year NICE, 2004

87
Number of included RCTs (number rated higherquality) * 1 (not rated)
Median duration of follow-up (range) 6 months
Median sample sizes in included RCTs (range) 64
Interventions evaluated (number of RCTs) Intradiscal electrothermal therapy (1)
Main conclusions IDET vs. sham for presumed discogenic LBP (1 RCT): IDET superior for decrease in pain at 6 months (78% [25/32] vs. 46% [11/24]), but not for proportion with >50% pain relief (38% [12/32] vs. 33% [8/32]). IDET superior for improvement in ODI (11 vs. 4 points, p=0.050), but not for SF-36 bodily pain or physical function subscales. IDET vs. sham for presumed discogenic LBP (2 higher-quality RCTs): Inconsistent results, with no differences through 6 months in the highest quality RCT and small differences in favor of IDET for pain and disability in the other RCT Local injection vs. placebo for low back pain associated with degenerative disease (3 RCTs): Local injection superior to placebo for short-term symptoms Local injection vs. dry needle acupuncture stick for low back pain associated with degenerative disease (1 RCT): No difference (proportion responding 63% vs. 42%, p=0.09)
Urrutia, 200799
Qualitative
2 (2)
6 months (both RCTs)
57 and 64
Local injections (6 unique RCTS in 3 systematic reviews) Resnick, 200592 Qualitative 4 (0) 0 2 weeks (7 days to 2 months) 36 (15 to 63) Trigger point injection (2), iliac crest injection (1), iliolumbar injection (1) 2
EVIDENCE REVIEW
Author, year Staal, 2008

94
Median duration of follow-up (range) Range 2 weeks to 2 months
Interventions evaluated (number of RCTs) Trigger point injection (1) iliac crest injection (1) iliolumbar injection (1)
Main conclusions Iliolumbar ligament injection with corticosteroid plus local anesthetic versus placebo (1 lower-quality RCT): No difference in pain relief at 2 weeks, but higher self-reported improvement Iliac crest injection with local anesthetic versus saline (1 higher-quality RCT): Pain score significantly better in injection group at 2 weeks Trigger point injection with local anesthetic or local anesthetic + corticosteroid vs. single dry needlestick or ethyl chloride plus acupressure (1 higher-quality RCT): No differences between groups at two weeks in self-rated improvement.
Percutaeneous intradiscal radiofrequency thermocoagulation or Coblation nucleoplasty (2 unique RCTs in 5 systematic reviews)
80
Gibson, 200579,
Qualitative
1 (1)**
8 weeks
28
Percutaneous intradiscal radiofrequency thermo-coagulation (1) Not applicable
PIRFT vs. sham for presumed discogenic LBP (1 RCT): 1/13 vs. 2/15 judged a 'success' after eight weeks
NICE, 200488
Qualitative
Not applicable
Not applicable
Not applicable
Coblation therapy for presumed discogenic LBP: Case series data only, with mixed results (1 study reported no sustained pain relief at 12 months) PIRFT vs. sham for presumed discogenic LBP (1 RCT): No differences in pain,
NICE, 200487
Qualitative
2 (not rated)
8 weeks and 6 months
28 and 39
Percutaneous intradiscal
EVIDENCE REVIEW
Number of RCTs not included in any other relevant systematic review
Author, year
Type of systematic review
Number of included RCTs (number rated higherquality) *
Median duration of follow-up (range)
Median sample sizes in included RCTs (range)
Interventions evaluated (number of RCTs) radiofrequency thermocoagulation (2)
Main conclusions functional improvement, global effect 1 of 2 case series reported improvement after PIRFT (16/39 improved at 16 months) PIRFT vs. sham for presumed discogenic low back pain (1 RCT): Limited evidence that intradiscal radiofrequency thermocoagulation not effective PIRFT vs. sham for presumed discogenic LBP (one higher-quality RCT): No differences in pain, disability, quality of life, global effect, therapeutic success, and analgesic intake at 8 weeks PIRFT at 80 C for 120 seconds versus 360 seconds for presumed discogenic LBP (one higher-quality RCT): No differences in pain and disability at 6 months. Improvement in both groups at 1 month, but not at month 2 and beyond. PIRFT vs. IDET (one non-randomized study): IDET superior for pain at disability at 1 year
Overall quality using Oxman scale (1 to 7)
Niemisto, 200390,
91
Qualitative
1 (1)
12 weeks to 12 months
31 to 70
Percutaneous intradiscal radiofrequency thermocoagulation (3) Percutaneous intradiscal radiofrequency thermocoagulation (2)
Urrutia, 200799
Qualitative
2 (2)
8 weeks and 6 months
28 and 39
EVIDENCE REVIEW
Author, year Dagenais, 76 2007
Median duration of follow-up (range) 6 months (6 to 24 months)
Interventions evaluated (number of RCTs) Prolotherapy (5)
Main conclusions Prolotherapy vs. control injections: >50% pain relief: RR 1.88 (95% CI 0.57 to 6.19) at 3 months (1 RCT); RR 1.10 (95% CI 0.75 to 1.61) at 6 months (1 RCT); RR 0.94 (95% CI 0.47 to 1.85) at 12 months (1 RCT) >50% improvement in disability at 5 months (1 RCT): RR 1.50 (95% CI 0.94 to 2.40) Prolotherapy with co-interventions vs. control injection with co-interventions >50% improvement in pain or disability after 6 months (1 RCT with similar cointerventions): RR 1.47 (95% CI 1.04 to 2.06) at 6 months >50% improvement in disability (1 RCT with different co-interventions): RR 2.24 (95% CI 1.50 to 3.35) Radiofrequency denervation vs. sham for presumed facet joint pain (1 higher-quality RCT + 10 observational studies): Moderate evidence (1 positive RCT and 10 observational studies) for short- and longterm improvement in pain Radiofrequency denervation vs. sham for presumed facet joint pain (2 RCTs, 1 higherquality): 2 of 2 RCTs found radiofrequency superior to sham
Prolotherapy (5 RCTs in 1 systematic review)
Radiofrequency denervation (4 unique RCTs in 5 systematic reviews) Boswell, 200775 Qualitative 1 (1) 0 12 months 31 Radio-frequency denervation of lumbar medial branch nerve (1) 3
Geurts, 200178
Qualitative
3 (1)
3 to 12 months
Range 31 to 41
Radio-frequency denervation of lumbar medial branch nerve (3)
EVIDENCE REVIEW
Author, year Niemisto, 2003

91 90,
Median duration of follow-up (range) Range 3 to 12 months
Interventions evaluated (number of RCTs) Radio-frequency denervation of lumbar medial branch nerve (3)
Main conclusions Radiofrequency deneveration vs. sham for presumed facet joint pain (3 RCTs, 2 higherquality): Conflicting evidence of short-term effects (1 RCT positive, 1 neutral, 1 unclear)
Resnick, 200592
Qualitative
3 (2)
Range 31 to 70
Radio-frequency denervation of lumbar medial branch nerve (3) Radio-frequency denervation of lumbar medial branch nerve (3)
Radiofrequency denervation vs. sham for presumed facet joint pain (3 RCTs, two higher-quality): Mixed results, with radiofrequency denervation superior to sham in 2 of 3 RCTs Radiofrequency denervation vs. sham for presumed facet joint pain (3 RCTs): 3 RCTs reported a 'positive' response to radiofrequency denervation, but in 1 RCT there was no longer a significant difference at 12 weeks
Slipman, 200393
Qualitative
3 (not rated)
Range 31 to 70
Sacroiliac joint injection (0 RCTs in 1 systematic review) Hansen, 200783 Qualitative No RCTs or sacroiliac joint injections for sacroiliac pain not related to spondyloarthropathy Not applicable Not applicable Not applicable Not applicable Not applicable 5
EVIDENCE REVIEW
Author, year Taylor, 2005 and 95, 891 2006
Number of included RCTs (number rated higherquality) * No RCTs of spinal cord stimulation in patients without failed back surgery syndrome
Median duration of follow-up (range) Not applicable
Median sample sizes in included RCTs (range) Not applicable
Interventions evaluated (number of RCTs) Not applicable
Main conclusions Not applicable
Spinal cord stimulation (0 RCTs in 1 systematic review)
*Trials adequately meeting at least half of the quality rating criteria or rated as good or higher-quality if the number of criteria met was not reported **Trials adequately meeting criteria for adequate allocation concealment CI=confidence interval, ODI=Oswestry Disability Index, LBP=low back pain, OR=odds ratio, RCT=randomized controlled trial, RDQ=Roland-Morris Disability Questionnaire, RR=relative risk, WMD=weighted mean difference
EVIDENCE REVIEW
Table 93. Summary of evidence on interventional therapies for low back pain
PlaceboNumber of placebocontrolled Total controlled trials trials with number Net benefit (number rated higher-quality) 100 patients of trials vs. placebo* Inconsistency 1 (1) 0 1 Moderate (short-term only, one small trial) Unable to determine Not applicable
Intervention Non-spinal injections Botulinum toxin injection
Population Non-specific low back pain
Directness of evidence Direct
Comments
Local injections
Non-specific low back pain
3 (1)
No
Direct
Poor
Interventions and populations varied substantially between trials. No higher-quality trials, all trials had small sample sizes
Prolotherapy
5 (4)
No effect
No
Direct
Good
Intraspinal steroid injections and chemonucleolysis Chemonucleolysis Radicul-opathy with prolapsed lumbar disc Epidural steroid injection Radicul-opathy with prolapsed lumbar disc 6 (5) 2 22 Moderate No Direct Good Chemonucleolysis with chymopapain superior to placebo injection, but inferior to surgery Inconsistency between higher-quality trials could be due to use of epidural or non-epidural placebo injection In two of three trials, only a subgroup of patients had spinal stenosis No difference between epidural steroid and intrathecal midazolam injection in one small trial
21 (9)
34
Moderate (short-term only)
Yes
Direct
Fair
Spinal stenosis
3 (1)
Unable to determine No evidence
No
Direct
Poor
Not applicable
Not applicable
Direct
Poor
Epidural steroid injection
Failed back surgery syndrome
Not applicable
No evidence
No
Direct
Poor
EVIDENCE REVIEW
PlaceboNumber of placebocontrolled controlled trials Total trials with number Net benefit (number rated higher-quality) 100 patients of trials vs. placebo* Inconsistency 2 (1) 0 2 Not applicable 7 3 No effect No evidence No No
Intervention Facet joint steroid injection Intradiscal steroid injection
Population Presumed facet joint pain Radicul-opathy with prolapsed lumbar disc Presumed discogenic low back pain
Directness of evidence Direct Direct
Overall quality of evidence Fair Fair
Comments
No effect versus chemonucleolysis
3 (1)
No effect
No
Direct
Good
Medial branch Presumed facet block (therapeutic) joint pain Sacroiliac joint steroid injection Presumed sacroiliac joint pain Radiculopathy with prolapsed lumbar disc Presumed discogenic low back pain Intradiscal electrothermal therapy (IDET) Presumed discogenic low back pain
0 1 (1)
Not applicable 0
3 1
No evidence Substantial (one small trial)
No Not applicable
Direct Direct
Poor Poor The only available trial evaluated a periarticular corticosteroid injection
Radiofrequency denervation, intradiscal electrothermal therapy (IDET), and related procedures Coblation nucleoplasty 0 Not applicable 0 No evidence Not applicable Not applicable Not applicable Direct Not applicable
Not applicable
No evidence Not applicable
Not applicable
2 (2)
Unable to determine (two trials with inconsistent results) No benefit (one trial)
Yes
Poor
Percutaneous Presumed intradiscal discogenic low radiofrequency back pain thermocoagulation (PIRFT)
1 (1)
Not applicable
Direct
Poor
EVIDENCE REVIEW
PlaceboNumber of placebocontrolled controlled trials Total trials with number Net benefit (number rated higher-quality) 100 patients of trials vs. placebo* Inconsistency 1 (1) 0 1 No benefit (one trial) Unable to determine Not applicable
Intervention Radiofrequency denervation
Population Radiculopathy with prolapsed lumbar disc Presumed facet joint pain
Comments
6 (4)
Yes
Direct
Poor
1 higher-quality trial used an inadequate technique, another had large baseline differences in pain scores
Presumed discogenic low back pain Spinal cord stimulation Failed back surgery syndrome with persistent radiculopathy Non-specific low back pain, or radiculopathy with prolapsed lumbar disc
1 (0)
Unable to determine (one trial) Moderate (see comments)
Not applicable
Direct
Poor
1 (1)
No
Direct
Fair
Spinal cord stimulation superior to repeat surgery in one trial and superior to conventional medical management in a second trial
Not applicable No trials No evidence Not applicable
Not applicable
Not applicable
* Based on evidence showing intervention is more effective than placebo or sham therapy for one or more of the following outcomes: pain, functional status, overall improvement, or work status. Versus placebo, small benefit defined as 5-10 points on a 100-point Visual Analogue Scale (VAS) for pain (or equivalent), 1-2 points on the Roland Morris Disability Questionnaire (RDQ), 5-10 points on the Oswestry Disability Index (ODI), or a standardized mean difference (SMD) of 0.2-0.5. Moderate benefit defined as10-20 points on a VAS for pain, 2-5 points on the RDQ, 10-20 points on the ODI, or a SMD of 0.5-0.8. Large benefit defined as >20 points on a 100 point VAS for pain; >5 points on the RDQ, >20 points on the ODI, or a SMD of >0.8. Inconsistency defined as <75% of trials reaching consistent conclusions on efficacy (no effect vs. positive effect considered inconsistent). Quality of one small French-language trial not assessed.
EVIDENCE REVIEW
Key Question 9 How effective is surgery (and different surgical interventions) for non-radicular low back pain, radicular low back pain, or spinal stenosis, and under what circumstances? Surgery for non-radicular low back pain with common degenerative changes
The most common surgery for chronic, non-specific low back pain or degenerative disc disease is fusion, a surgical procedure that unites (fuses) two or more vertebra together. The goal of fusion surgery is to relieve symptoms by restricting motion at the source of spinal pain (usually presumed to be a degenerated intervertebral disc) after removing the disc. A variety of spinal fusion techniques are practiced. All involve placement of a bone graft between the vertebrae. fusion can be performed with or without the use of supplemental hardware (instrumentation) such as plates, screws, or cages that serve as an internal splint while the bone graft heals. Total disc replacement is a recently introduced alternative to fusion. A theoretical advantage of total disc replacement over fusion is that a prosthetic disc could help preserve normal range of motion and mechanics of the spine. This could reduce long-term degenerative changes in adjacent vertebral segments, which may be observed following fusion. Prosthetic discs approved by the Food and Drug Administration as of January 2007 are the Charite and ProDisc-L artificial discs. Results of search: systematic reviews We identified a total of 13 systematic reviews on efficacy of surgery for chronic, non-specific degenerative low back pain or degenerative disc disease with presumed discogenic low back pain: one higher-quality Cochrane review79, 80, three other higher-quality systematic reviews212, 215, 218 and nine lower-quality systematic reviews72, 210, 213, 214, 221, 224-226, 230. Four systematic reviews focused on efficacy or safety of vertebral disc replacement for degenerative disc disease with presumed discogenic low back pain212-214, 221, one systematic review evaluated both fusion and artificial disc replacement79, 80, and the remainder focused only on fusion. One other lower-quality systematic review of fusion focused only on harms230. We excluded two previous versions of the Cochrane review176, 255 and one other outdated systematic review260. Results of search: trials Twenty randomized trials evaluated surgery for non-radicular low back pain with common degenerative changes (usually degenerative disc disease with presumed discogenic low back pain)244-247, 252, 253, 892-905. All of the trials were included in at least one of 12 systematic reviews72, 79, 80, 210, 212, 214, 215, 218, 221, 224-226, 230 . Four trials244-247 compared surgery to non-surgical therapy and two trials252, 253 compared artificial disc replacement to fusion. We excluded one trial263 that evaluated surgery for foraminal stenosis due to degenerative disc disease and one trial265 that reported interim, single center results from a multicenter trial.
EVIDENCE REVIEW
Efficacy of fusion versus non-surgical management for non-radicular low back pain with common degenerative changes
Four higher-quality trials of fusion surgery versus non-surgical therapy enrolled patients with moderately severe pain (mean score 63 to 65 on a 0 to 100 scale244, 245, 247) or disability (mean ODI score=45246) for at least one year, unresponsive to standard non-surgical therapy. Positive results on provocative discography were not required for entry in any trial. Exclusion criteria included significant psychiatric or somatic illness, ongoing compensation issues or presence of other chronic pain conditions. Surgical techniques involved some type of fusion procedure, though specific methods varied (Table 94).
Table 94. Trials of fusion versus non-surgical therapy for non-radicular low back pain with common degenerative changes
Author, year Population evaluated Brox, 2003245 Chronic low back pain with degenerative disc disease at L4/L5 or L5/S1 (no prior discectomy) Brox, 2006244 Chronic low back pain with degenerative disc disease at L4/L5 or L5/S1 following discectomy Fairbank, 2005246 MRC Spine Stabilization Trial Chronic low back pain and considered a candidate for spinal fusion Fritzell, 2001247 Swedish Lumbar Spine study Chronic low back pain with degenerative disc disease at L4/L5 or L5/S1 Number of patients Duration of follow-up n=64 1 year
Surgical intervention Instrumented posterolateral fusion
Instrumented posterolateral fusion
n=60 1 year
Graf ligamentoplasty (15%) or fusion with technique left to discretion of surgeon (85%)
n=349 2 years
Noninstrumented posterolateral fusion (1/3), instrumented posterolateral fusion (1/3), or instrumented circumferential fusion (1/3)
n=294 2 years
Main results Surgery versus intensive rehabilitation with a cognitive-behavioral component ODI score, mean difference in change from baseline: 2.3 (-6.8 to 11.4) Back pain, mean difference in change from baseline: 8.6 (-3.0 to 20.1) Overall rating success: 71% vs. 63%, p=0.59 Surgery versus intensive rehabilitation with a cognitive-behavioral component ODI score, mean difference in change from baseline: -7.3 (-17.3 to 2.7) Back pain, mean difference in change from baseline: -5.2 (-18.0 to 7.6) Overall rating success: 50% vs. 48%, p=0.91 Surgery versus intensive rehabilitation with a cognitive-behavioral component ODI, mean difference in change from baseline: -4.1 (-8.1 to -0.1), p=0.045 SF-36 physical component score, mean difference in change from baseline: 2.0 (1.2 to 5.3) SF-36 mental component score, mean difference in change from baseline: -0.2 (-2.9 to 2.6) Surgery versus non-intensive physical therapy Back pain VAS score, mean change from baseline (0 to 100 scale): 21.0 vs. 4.3, p=0.0002 ODI score, mean change from baseline: 11.6 vs. 2.8, p=0.015 Overall rating better or much better: 63% vs. 29%, p<0.0001
Quality score* 8/9
8/9
6/9
7/9
*Excludes criteria involving blinding of patients and care providers, for maximum score of 9 (VAS): Visual Analogue, (RDQ): Roland Morris Disability Questionnaire, (ODI): Oswestry Disability Index, (SF-36):Shortform 36
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain The trials reported inconsistent results244-247. In the Swedish Lumbar Spine Study (n=294), independent assessors rated outcomes as excellent or good (no more than sporadic pain, slight restriction of function, and occasional analgesics) in 46% of those randomized to surgery versus 18% randomized to non-surgical therapy after 2 years (p<0.0001)247. More of the surgical patients rated results as better or much better (63% vs. 29%, p<0.0001). Patients randomized to surgery also experienced moderately greater improvements in pain (mean change from baseline on 0 to 100 VAS pain score 21.0 vs. 4.3, p=0.0002) and slightly greater improvements in ODI scores (mean change from baseline 11.6 vs. 2.8, p=0.015), and a higher proportion returned to work (36% vs. 13%, p=0.002). Two smaller (n=60 and 64) trials conducted by the same Norwegian investigators found no statistically significant differences between surgery versus non-surgical therapy on any of the main outcomes after one year among patients either with244 or without245 prior discectomy. In the latter trial surgery was associated with a trend towards slightly superior outcomes on the ODI (mean difference=-7.3, 95% CI -17.3 to +2.7) and back pain scores (mean difference=-5.2, 95% CI -18.0 to +7.6)244. The Medical Research Council (MRC) Spine Stabilization Trial (n=349) found surgery associated with statistically significant improvements in ODI scores after 24 months compared to non-surgical therapy, but the difference did not reach clinical significance (mean difference -4.1, 95% CI -9.1 to -0.1, p=0.045)246. There were no differences in other outcomes, including SF-36 scores and the shuttle walking test. The inconsistent results between trials could be related to differences in non-surgical comparator treatments. In the three trials that found clinically or statistically insignificant benefits following surgery, non-surgical treatment consisted of intensive rehabilitation incorporating cognitive behavioral therapy (75 hours over three weeks, with subsequent followup visits)244-246. In the one trial that showed surgery associated with clinically and statistically significant benefits, the non-surgical treatment intervention was less intensive (70 hours of supervised physical therapy over a 2 year period) and more heterogeneous (could be supplemented by other interventions such as transcutaneous electrical nerve stimulation [TENS], acupuncture, injections, advice, and cognitive therapy)247. In addition, one of the criteria for enrollment in this trial was inadequate response to non-surgical treatment, but patients randomized to the non-surgical arm may have continued to receive previously ineffective interventions. Two higher-quality systematic reviews also found inconsistent results for surgery versus no surgery that could be explained by the non-surgical comparator intervention79, 80, 218. Another higher-quality, quantitative systematic review found no difference between surgery and nonsurgical therapy when data from three trials245-247 were pooled (-4.13, 95% CI -9.08 to 0.82), but heterogeneity was present, in part because trials of intensive and standard rehabilitation were combined215. Two lower-quality systematic reviews estimated success rates of 67% to 79% following fusion, but pooled data across primarily uncontrolled observational studies72, 210. A third systematic review224 postulated that lack of efficacy observed in smaller (N<100) trials could have been due to small sample sizes and insufficient power to detect differences. However, even if statistically significant, point estimates from the smaller trials would either
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain favored non-surgical therapy (2.3 points on the ODI245) or would only slightly favor surgery (7.3 points on the ODI244).
Efficacy of different fusion techniques

There is insufficient evidence to determine optimal fusion methods. Instrumentation and electrical stimulation appear to enhance fusion rates, but effects on clinical outcomes are not established79, 80. Although pooled estimates in a higher-quality Cochrane review found instrumentation superior to no instrumentation (OR=0.49 for poor clinical outcome, 95% CI 0.28 to 0.84), results are sensitive to inclusion of two older, lower-quality outlier trials (one nonrandomized906) that reported unusually favorable results with surgery (83%906 and 93%905 success with instrumented fusion). A re-analysis that was limited to higher-quality trials published since 1997 found marginal and insignificant effects of fusion (74% vs. 68% pooled rates of clinical success). There are conflicting results from head-to-head trials regarding the relative effectiveness of various types of fusion (anterior, posterior, or combined)79, 80.
Efficacy of artificial disk replacement versus fusion

For chronic non-radicular back pain with single-level degenerative disc disease from L3 or L4 to S1, two trials (each funded by the manufacturer of the relevant artificial disc) found no clear differences between artificial disc replacement versus fusion through two years follow-up (Table 95)252, 253. One higher-quality trial252 (n=304) found the Charit Artificial Disc non-inferior to anterior lumbar interbody fusion using the BAK Interbody Fusion System (a technique no longer commonly used because of frequent poor outcomes907) on a composite outcome of clinical success (25% improvement in ODI, no device failure, no major complications, and no neurologic deterioration) at 24 months (57% vs. 46%, p<0.0001 for equivalence test; calculated RR 1.23, 95% CI 0.96 to 1.57)252. There were no differences in mean ODI (48.5 vs. 42.4, p=0.27 for difference), VAS pain scores (40.6 vs. 34.1, p=0.11) or rates of employment at 24 months, though disc replacement was slightly superior at earlier evaluations. One lower-quality trial (n=286) that compared Prodisc II artificial disc replacement to instrumented circumferential fusion was also designed as a non-inferiority trial, but results appeared to be reported using standard statistical tests for evaluating a superiority hypothesis253. It found the Prodisc II superior to circumferential fusion on a composite outcome of success (ODI improved >15 points, device success, neurologic success, SF-36 improved, and radiographic success) after 24 months (53% vs. 41%, p=0.044). However, there were no statistically significant differences on the ODI (mean scores or proportion with >15 point improvement), pain scores, or SF-36 composite mental and physical component scores.
EVIDENCE REVIEW
Table 95. Trials of artificial disc replacement versus fusion

Number of patients Duration of follow-up n=304 24 months Chronic low back pain with singlelevel degenerative disc disease between L4 and S1
Author, year Population evaluated Blumenthal, 2005252
Surgical intervention Charit artificial disc
Zigler, 2007253 Chronic low back pain with singlelevel degenerative disc disease between L3 and S1
Prodisc-L artificial disc
n=292 24 months
Main results Total disc replacement with Charit artificial disc vs. anterior lumbar interbody fusion with BAK cage Clinical success: 117/205 (57%) vs. 46/99 (46%), p<0.0001 for equivalence 25% improvement in ODI: 131/205 (64%) vs. 50/99 (50%) Length of hospitalization: 3.7 vs. 4.2 days, p=0.0039 ODI, mean improvement from baseline at 24 months: 49% vs. 42%, p<0.05 VAS for pain, mean improvement from baseline at 24 months (0 to 100 scale): 40.6 vs. 34.1, p<0.05 Patient satisfaction rated as 'satisfied': 74% vs. 53%, p=0.0011 'Would have same treatment again': 70% vs. 50%, p=0.0062 Use of opioids: 148/205 (72%) vs. 85/99 (86%), p=0.0083 Employed at 24 months (percent increase): 9.2% vs. 7.4%, NS Total disc replacement with Prodisc-L vs. circumferential fusion ODI (mean improvement at 24 months): 28.9 (46% improvement) vs. 22.9 (36% improvement) (p=0.055) ODI improved >15 points from baseline: 53% vs. 36% at week 6 (p=0.010), 60% vs. 45% at month 6 (p=0.029), 58% vs. 53% at month 12 (p=0.332), 68% vs. 55% at month 24 (p=0.045) SF-36 composite mental and physical component scores improved from baseline: 87% vs. 70% at month 3 (p=0.004), 81% vs. 77% at month 12 (p=0.302), 79% vs. 70% at month 24 (p=0.094) Overall success (ODI improved >15 points, device success, neurologic success, SF-36 improved, and radiographic success): 53% vs. 41% (p=0.044) VAS Pain (mean improvement at 24 months on 0 to 100 scale): 39 vs. 32 (p=0.08) VAS Patient satisfaction (0 to 100): 77 vs. 67 (p=0.015) Opioid use in persons achieving success: 39% vs. 31% (76% vs. 84% at baseline) Employed: 92% vs. 85% (p=0.048)
Quality score* 7/10
5/10
Selection of patients for surgery for non-specific low back pain

Patients enrolled in trials of surgery versus non-surgical treatment all had moderately severe chronic pain (mean pain score=62 to 65 on a 0 to 100 scale) or disability (mean ODI score=45) for at least one year, unresponsive to standard non-surgical therapy. Patients had moderate pain (mean scores 63 to 65244, 245, 247) or disability (mean ODI 45246). Exclusion criteria generally
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain included any significant psychiatric or somatic illness and often included ongoing compensation issues or other chronic pain conditions. Uncontrolled observational studies have shown poorer surgical outcomes in such patients254, 908, 909. In a recent randomized trial (the Swedish Lumbar Spine Study) of surgery versus non-surgical management of chronic low back pain, personality features and low disc height both predicted functional improvement after surgery, and lower age and short sick leave predicted return to work after surgery910. The presence of depressive symptoms predicted functional improvement after non-surgical treatment. Harms No operative deaths were reported in any randomized trial of fusion versus non-surgical therapy244-247. The pooled rate of early surgical complications from three trials245-247 was 16% (95% CI 12% to 20%)215. Major complications included deep wound infections, major bleeding during surgery, thrombosis, acute respiratory distress syndrome, pulmonary edema, and heart failure. One trial, which evaluated different fusion techniques, found higher risks of complications with more technically difficult procedures911. The total complication rate after two years was 12% with non-instrumented posterolateral fusion, 22% with instrumented posterolateral fusion, and 40% with circumferential fusion. A recent, large observational study based on the Nationwide Inpatient Sample reported <1% in-hospital mortality for all fusion procedures912. In systematic reviews that included observational studies, complication rates following fusion varied widely and were difficult to interpret due to differences in techniques, study populations, and methodological shortcomings72, 230. One systematic review found perioperative complications ranged from 2% to 54% in 31 studies of different fusion methods, with a trend towards higher complications with circumferential fusion72. Another systematic review found wide variation in estimates of common adverse events or undesirable outcomes following anterior or posterior lumbar interbody fusion with a stand-alone cage. Rates of nonunion ranged from 0% to 83% in 24 studies, rates of major vessel injury ranged from 0 to 12% in 12 studies, rates of neurologic complications ranged from 0 to 44% in 10 studies, and rates of dural injury ranged from 2% to 15% in 8 studies230. Higher rates of solid fusion were associated with potential author conflicts of interest, though there was no association between potential conflicts of interest and estimates for other outcomes. One shortcoming of this study is that other factors that could affect reported complication rates (such as study quality) were not assessed. In two trials of artificial disc replacement, one death was reported among 205 patients randomized to Charit total disc replacement252 and none in 161 patients randomized to Prodisc-L artificial disc replacement253. There were no major complications in the Prodisc-L trial, and in the Charit trial there were no differences between artificial disc replacement and fusion in rates of overall (p=0.6769) complications. Major complications occurred in 1% of patients in both groups. The rates of major (4.9% vs. 4%) and minor (9.8% vs. 8.1%) neurologic complications were similar for artificial disc replacement and fusion. Long-term data following artificial disc replacement are limited, but case reports and other uncontrolled observational studies have reported prosthesis migration or subsidence (settling or sinking into bone), adjacent level disc degeneration, and facet joint arthritis, with some patients undergoing
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain subsequent fusion or artificial disc removal913-916. One study found fewer complications and shorter length of hospitalization when Charite total disc replacement was performed by surgeons more experienced in the procedure917. Costs Two trials of surgery versus non-surgical management of chronic non-specific low back pain conducted cost-effectiveness analyses918, 919. One estimated an incremental cost-effectiveness ratio of 48,588/QALY (about $95,232 U.S./QALY) for surgery relative to intensive rehabilitation919. Estimates were sensitive to the proportion of patients in the rehabilitation group that required surgery in the future. The other found surgery associated with an incremental cost-effectiveness ratio relative to usual care of about $372 ($86-729) per case of improvement, $744 ($157-1,644) per one point improvement on a 100 point pain scale, $1,616 ($186-6,864) per one point improvement on the ODI score, and $586 ($14-3,060) per patient returned to work (converted from Swedish kroner)918. There were no differences in costs associated with three different fusion techniques (posterolateral fusion, instrumented posterolateral fusion, and circumferential fusion with solid autogenous bone grafts).
Summary of evidence
For chronic non-radicular low back pain with common degenerative changes, three higherquality trials found spinal fusion surgery no better or only slightly superior to intensive rehabilitation plus a cognitive intervention for improvement in pain or function, but a fourth trial found fusion surgery moderately superior to less intensive physical therapy supplemented by other non-invasive interventions (TENS, acupuncture, injections, advice, and/or cognitive therapy) for pain and slightly superior for functional status (level of evidence: fair). For mixed degenerative conditions (including degenerative spondylolisthesis), evidence on efficacy of instrumented versus non-instrumented fusion is inconsistent, though clinical outcomes are similar after excluding two lower-quality outlier trials and pooling data from the remaining six trials (level of evidence: fair). Evidence regarding efficacy of anterior, posterior, or combined fusion from four trials is inconsistent and does not permit reliable judgments about relative efficacy (level of evidence: fair). Electrical stimulation may improve fusion rates in non-instrumented (but not instrumented) fusion, but did not have a clear effect on clinical outcomes in three trials (level of evidence: fair). For degenerative disc disease, artificial disc replacement with the Charite artificial disc was non-inferior to anterior interbody fusion with a stand-alone cage for a combined measure of success at 24 months in one higher-quality trial, and artificial disc replacement with the ProDisc-L artificial disc was slightly superior to circumferential fusion for a combined measure of success at 24 months in another higher-quality trial. In both trials, there were no differences in pain relief or functional status at 24 months, though some earlier results favored artificial disc replacement (level of evidence: fair).
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Early complications following fusion occur in up to about 20% of patients. The rate of inhospital mortality is <1%. Rates of other complications vary widely between studies (level of evidence: fair). Complications from spinal fusion were more frequent with more technically difficult methods in one higher-quality trial (level of evidence: fair). Rates of complications were similar after artificial disc replacement and fusion in two higherquality trials that each evaluated a different artificial disc (level of evidence: fair). Trials of surgery versus non-surgical management generally included patients with moderate pain who failed to improve after 6 months to 2 years of non-surgical management, and had disease localized to L4-L5 and/or L5-S1.

The AHCPR guidelines recommend against spinal fusion for the treatment of low back problems during the first 3 months of symptoms (strength of evidence: C). The AHCPR guidelines recommend that spinal fusion be considered following decompression at a level of increased motion due to degenerative spondylolisthesis (strength of evidence: C). The European COST guidelines found insufficient evidence to recommend fusion surgery for chronic low back pain unless two years of all other recommended conservative treatments have failed and combined programs of cognitive interventions and exercises are not available in the given geographical area. It strongly recommends that only carefully selected patients with severe pain (and with maximum 2 affected levels) should be considered for fusion.
Surgery for isthmic spondylolisthesis

Isthmic spondylolisthesis refers to a condition in which a lytic defect in the pars interarticularis results in anterior subluxation of the affected vertebral body. The subluxation may also place stress on the adjacent intervertebral disc, resulting in degenerative disc disease. The most common site of isthmic spondylolisthesis is at L5, which can cause back pain or radicular symptoms due to tension or compression on the L5 nerve root. The most common surgical procedure for isthmic spondylolisthesis is fusion, with or without decompressive laminectomy. Results of search: systematic reviews We identified a higher-quality Cochrane review of surgery for degenerative conditions of the back, including isthmic spondylolisthesis79, 80. We identified one other lower-quality systematic review on efficacy of surgery for low-grade isthmic spondylolisthesis216. We excluded earlier versions of the Cochrane review176, 255. Results of search: trials Six unique trials (reported in ten publications233, 235, 920-927) evaluated surgery for isthmic spondylolisthesis (three in mixed populations of patients with isthmic or degenerative spondylolisthesis233, 235, 921, 926). All were included in previously published systematic reviews79, 80, 216 . Only one trial compared surgery to non-surgical therapy925. Long-term (9 years) results of
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain this trial have been reported922. The other trials evaluated different methods of surgery. Three trials met criteria for adequate allocation concealment in the Cochrane review920, 921, 924-926. We excluded one non-randomized study that compared different surgical techniques for unstable low-grade isthmic spondyolisthesis928.
Efficacy of surgery versus non-surgical treatment of isthmic spondylolisthesis

For lumbar isthmic spondylolisthesis of any grade with low back pain for at least one year and no radiologic disc prolapse or central canal stenosis, one lower-quality trial found posterolateral fusion associated with moderately decreased pain (mean score 37 vs. 56 on a 0 to 100 scale, p=0.002) and disability (mean Disability Rating Index 29 vs. 44 on a 0 to 100 scale, p=0.004) and superior patient-reported overall outcomes (74% vs. 43% better or much better) after 2 years compared to an exercise program, though there were no significant difference in workrelated outcomes (46% vs. 45% working) (Table 96)925. Nearly all patients enrolled in this trial (112 of 114) were categorized as having a Grade I or II slip. After an average of 9 years followup, differences were small and no longer significant for pain or function922. Relief of sciatica from nerve root compression (the major indication for surgery in patients with isthmic spondyolisthesis) was not reported.
Table 96. Trial of surgery versus non-surgical treatment for isthmic spondylolisthesis
925
Author, year Population evaluated Moller, 2000 922,
Surgical intervention
Chronic lumbar isthmic spondylolisthesis (any grade)
Posterolateral fusion with or without instrumentation
Main results Surgery versus exercise therapy Disability Rating Index, mean score (0 to 100 scale): 29 vs. 44 (p=0.004) at 2 years, 33 vs. 38 (NS) at 9 years Pain, mean score (0 to 100 scale): 37 vs. 56 (p=0.002) at 2 years, 40 vs. 49 at 9 years Proportion working: 46% vs. 45% (NS) at 2 years, 51% vs. 46% at 9 years Overall outcome much better or better: 74% vs. 43% at 2 years, 76% vs. 50% at 9 years
Quality score* 4/9
Efficacy of different surgery techniques for isthmic spondylolisthesis

For grade I or II isthmic L5/S1 spondylolisthesis without neurologic deficits, one small (n=42), higher-quality trial found fusion (with or without instrumentation) plus decompressive laminectomy associated with higher rates of pseudoarthrosis (22% vs. 0%, p=0.02) and unsatisfactory results (33% vs. 4%, p=0.01) compared to fusion alone (with or without instrumentation)920. Results may be somewhat confounded because only patients who smoked received instrumentation, though no benefits from decompression were observed in either smokers or non-smokers. In two lower-quality trials included in the Cochrane review, there were no differences in fusion rates923, 924 or clinical outcomes924 between patients randomized to instrumented posterolateral fusion versus non-instrumented fusion.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain A lower-quality systematic review found posterior fusion for isthmic spondyolisthesis associated with poorer success rates compared to anterior or combined approaches (75% versus 90% and 86%)216. Instrumented fusion was associated with higher success rates than non-instrumented fusion (85% vs. 64%, p<0.0001). There were no differences in success rates with fusion plus laminectomy versus fusion alone (74% vs. 80%, p=0.11). However, estimates may not be reliable because they are based on simple pooling of success rates across randomized and non-randomized studies, including lower-quality, uncontrolled surgical series. Results of trials233, 235, 921, 926, 927 that included mixed populations of patients with isthmic or degenerative spondylolisthesis are summarized in the section on surgery for spinal stenosis or degenerative spondylolisthesis (see below). Harms One trial of surgery versus non-surgical therapy for isthmic spondyolisthesis found posterolateral fusion associated with three major operative complications (two cases of permanent L5 root injury during instrumented fusion out of 37 subjects, and one case of permanent blindness)925. Another trial (n=42) reported two post-operative complications following fusion (with or without instrumentation and with or without laminectomy) that required operative revision, one case of transient palsy of the sacral nerve, and one dural tear920. A trial (n=27) of posterolateral fusion with or without instrumentation reported four wound hematomas, one screw breakage, one damaged nerve root, and one pedicle fracture with subsequent radiculopathy that required surgical exploration923. Costs We found no studies evaluating costs.
Summary of evidence
For isthmic spondylolisthesis with Grade I or II slip, posterolateral fusion was moderately superior to an exercise program for pain and disability after 2 years in one lower-quality trial, though differences were no longer significant after an average of 9 years (level of evidence: poor). For grade I or II isthmic L5/S1 spondylolisthesis without neurologic deficits, one small, higherquality trial found fusion (with or without instrumentation) plus laminectomy and decompression associated with higher rates of pseudoarthrosis and unsatisfactory results compared to fusion (with or without instrumentation) alone (level of evidence: poor). For mild isthmic spondylolisthesis, instrumented fusion was no better than non-instrumented fusion in two lower-quality trials (level of evidence: poor). For mild isthmic spondylolisthesis, pooled data from primarily lower-quality observational studies found fusion using the anterior or combined approach superior for success rates compared to fusion using the posterior approach, and instrumented fusion superior to noninstrumented fusion (level of evidence: poor). There is insufficient evidence to reliably judge safety of surgery for isthmic spondylolisthesis.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Evidence on efficacy of different surgical techniques in mixed populations of patients with isthmic or degenerative spondylolisthesis are summarized in the section on surgery for spinal stenosis or degenerative spondylolisthesis.

The other guidelines do not make specific recommendations for surgery in patients with isthmic spondylolisthesis.
Surgery for spinal stenosis with or without degenerative spondylolisthesis

Common causes of acquired spinal stenosis (narrowing of the spinal canal) include degenerative disc disease, degenerative spondyolisthesis, prolapsed intervertebral disc, and diffuse idiopathic skeletal hyperostosis. The most common surgery for spinal stenosis is decompressive laminectomy, or removal of the vertebral lamina in order to create more space and reduce pressure on the spinal column or nerve roots. Laminectomy can be performed with or without fusion or discectomy. Another surgical treatment for spinal stenosis is placement of a spacer device between the interspinous processes, which could theoretically improve postural symptoms of spinal stenosis by limiting extension or lordosis of the spine (which results in narrowing of the spinal canal) when standing. Degenerative spondylolisthesis is a condition characterized by degenerative changes at the facet joints, which leads to a loss of normal structural supports and subluxation (slippage) of the affected vertebral body. This can cause pain and neurologic deficits due to tension on nerve roots or spinal stenosis. The most common site for degenerative spondylolisthesis is L4. The most common surgical procedure for degenerative spondylolisthesis is decompressive laminectomy, often with an intertransverse process arthrodesis (fusion) using an autogenous bone graft. Results of search: systematic reviews We identified a higher-quality Cochrane review on efficacy of surgery for spinal stenosis, with or without degenerative spondylolisthesis79, 80, one other higher-quality systematic review217, and six lower-quality systematic reviews210, 223, 225, 226, 228, 229. We also identified one lower-quality systematic review on predictors of postoperative clinical outcomes in spinal stenosis254. We excluded earlier versions of the Cochrane review176, 255 and three other outdated systematic reviews257-259. Results of search: trials Nineteen trials evaluated surgery for spinal stenosis with or without degenerative spondylolisthesis231-233, 235, 236, 241, 243, 250, 251, 895, 906, 926, 929-935. Twelve trials were included in at least one of the eight systematic reviews79, 80, 210, 217, 223, 225, 226, 228, 229 and we identified seven additional trials231-233, 235, 236, 241, 250. Of six trials that compared surgery versus non-surgical therapy, four evaluated laminectomy236, 241, 243, 250 and two evaluated an interspinous spacer device231, 251. Previously published systematic reviews included only one trial of laminectomy243 and one trial of an interspinous spacer device251 versus non-surgical therapy.
EVIDENCE REVIEW
Efficacy of decompressive surgery versus non-surgical treatment for spinal stenosis with or without degenerative spondyolisthesis
Four higher-quality trials compared surgery to non-surgical therapy for spinal stenosis (Table 97)236, 241, 243, 250. One trial evaluated surgery for spinal stenosis without degenerative spondylolisthesis250, one trial evaluated surgery for spinal stenosis with degenerative spondylolisthesis241, and two trials evaluated surgery for spinal stenosis with or without degenerative spondylolisthesis236, 243. In three trials, baseline pain scores averaged 31 to 32 on the SF-36 bodily pain score241, 250 or 7 on a 0 to 10 pain scale236. The fourth trial did not report baseline severity or duration of pain243. Although two trials permitted enrollment of patients with as little as 12 weeks of symptoms, the majority of patients in all trials reported at least six months of symptoms at the time of enrollment.
EVIDENCE REVIEW
Table 97. Trials of decompressive surgery versus non-surgical treatment for spinal stenosis
Author, year Population evaluated Amundsen, 2000243 Chronic spinal stenosis with or without degenerative spondylolisthesis Malmivaara, 2007236 Chronic symptomatic spinal stenosis with or without degenerative spondylolisthesis Surgical intervention Laminectomy (without fusion) Number of patients Duration of follow-up n=31 10 years Laminectomy (with or without fusion) n=94 2 years Quality score* 6/9
Main results Laminectomy (without fusion) versus non-surgical treatment Proportion with good results: 9/13 (69%) vs. 6/18 (33%) at 1 year; 11/12 (92%) vs. 8/17 (47%) at 4 years; 10/11 (91%) vs. 12/17 (71%) at 10 years (p values not reported) Laminectomy (with or without fusion) versus non-surgical treatment (difference between groups, negative values favor surgery) ODI: -7.6 (95% CI -13.9 to -1.3) at 6 months, -11.3 (95% CI -18.4 to -4.3) at 12 months, -7.8 (95% CI -14.9 to -0.8) at 24 months Leg pain during walking (0 to 10 scale): -2.02 (95% CI -3.36 to -0.69) at 6 months, -1.51 (95% CI -2.77 to -0.25) at 24 months Low back pain during walking (0 to 10 scale): -2.64 (95% CI -3.88 to -1.40) at 6 months, -2.13 (95% CI -3.28 to -0.98) at 24 months Self-reported walking ability (m): No significant differences Laminectomy (with or without fusion) versus non-surgical treatment (positive SF-36 and negative ODI scores favor surgery) Intention-to-treat results, differences between interventions at 2 years SF-36 bodily pain: 1.5 (95% CI -4.2 to 7.3) SF-36 physical function: 1.9 (95 5CI -3.7 to +7.5) ODI: 2.2 (95% CI -2.3 to +6.8) As-treated results at 2 years, differences between interventions (randomized cohort only) SF-36 bodily pain: +17.8 (95% CI 12.5 to 23.0) SF-36 physical function: +16.7 (95% CI 11.4 to 22.1) ODI: -15.9 (95% CI -20.2 to -11.7) As-treated results at 2 years, differences between interventions (randomized and observational cohorts) SF-36 bodily pain: 18.1 (95% CI 14.5 to 21.7) SF-36 physical function: 18.3 (95% CI 14.6 to 21.9) ODI: -16.7 (95% CI -19.5 to -13.9) Very or somewhat satisfied with symptoms (%): 36.6 (95% CI 28.0 to 45.1) Self-rated major improvement in progress (%): 50.0 (955 CI 42.2 to 57.9)
6/9
Weinstein, 2007241 Spine Outcomes Research Trials Chronic symptomatic spinal stenosis with degenerative spondylolisthesis
Laminectomy (with or without fusion)
n=304 2 years
5/9
EVIDENCE REVIEW
Table 97. Trials of decompressive surgery versus non-surgical treatment for spinal stenosis
Author, year Population evaluated Weinstein, 2008250 Spine Outcomes Research Trials Chronic symptomatic spinal stenosis without degenerative spondylolisthesis Surgical intervention Laminectomy (with or without fusion) Number of patients Duration of follow-up n=289 2 years Quality score* 5/9
Main results Laminectomy versus non-surgical therapy (positive SF-36 and negative ODI scores favor surgery). Intention-to-treat results, differences between interventions SF-36 bodily pain: 2.4 (95% CI -4.2 to 9.1) at 3 months, 7.8 (95% CI 1.5 to 14.1) at 2 years SF-36 physical function: -4.2 (95% CI -10.9 to 2.6) at 3 months, 0.1 (95% CI -6.4 to +6.5) at 2 years ODI: 0.5 (95% CI -5.0 to 6.0) at 3 months, -3.5 (95% CI -8.7 to +1.7) at 2 years As-treated results at 2 years, differences between interventions (randomized cohort only) SF-36 bodily pain: 11.7 (95% CI 6.2 to 17.2) SF-36 physical function: 8.1 (95% CI 2.8 to 13.5) ODI: -8.7 (95% CI -13.3 to -4.0) As-treated results at 2 years, differences between interventions (randomized and observational cohorts) SF-36 bodily pain: 13.6 (95% CI 10.0 to 17.2) SF-36 physical function: 11.1 (955 CI 7.6 to 14.7) ODI: -11.2 (95% CI -14.1 to -8.3) Very or somewhat satisfied with symptoms (%): 38.7 (95% CI 30.0 to 47.3) Self-rated major improvement in condition (%): 34.1 (95% CI 25.6 to 42.6)
*Excludes criteria involving blinding of patients and care providers, for maximum score of 9 (ODI):Oswestry Disability Index
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For spinal stenosis either with or without degenerative spondylolisthesis, two trials found initial decompressive surgery superior to initial non-surgical therapy236, 243. One small (n=31) trial found initial decompressive surgery (without fusion) superior to non-surgical treatment (lumbar support and back school) for likelihood of experiencing an overall good outcome, though differences were somewhat decreased at longer follow-up (69% vs. 33% at 1 year, 92% vs. 47% at 4 years, 91% vs. 71% at 10 years)243. Interpretation of results is complicated by crossover from non-surgical therapy to surgery in10 of 18 patients. The second trial (n=94) found laminectomy with or without fusion moderately superior to non-surgical therapy on the ODI (mean difference 11.3 points) and for leg pain (mean difference 1.7 points on a 10 point scale) and substantially superior for low back pain (mean difference 2.3 points on a 10 point scale) at 1 year, but differences were diminished after 2 years (7.8 points on the ODI, 1.5 for leg pain, 2.1 for back pain)236. Two large multicenter trials (the Spine Outcomes Research Trials, or SPORT936) evaluated laminectomy with or without fusion versus non-surgical therapy for spinal stenosis specifically with241 or without250 degenerative spondylolisthesis. Although both trials found few differences between surgical versus non-surgical therapy through two years based on intention-to-treat analyses, results are difficult to interpret because nearly half of patients did not adhere to treatment assignments. In an on-treatment analysis of randomized patients adjusted for potential confounders, surgery was moderately superior (16 to 18 points on 100 point scales) to non-surgical therapy on the ODI and SF-36 bodily pain and functional scores after two years for spinal stenosis with degenerative spondylolisthesis241, and slightly to moderately superior (8 to 12 points) to non-surgical therapy for spinal stenosis without degenerative spondylolisthesis250. Analyses that combined on-treatment results of randomized patients with data from concurrent observational cohorts resulted in slightly higher estimates in favor of surgery241, 250. In both trials, average improvements from baseline on the ODI and SF-36 in patients who did not undergo surgery averaged about ten points. Results of a higher-quality, long-term (8 to 10 years) prospective observational study (n=148) of surgery versus non-surgical therapy for spinal stenosis (the Maine Lumbar Spine Study) are consistent with the randomized trials937. In general, benefits associated with surgery were statistically significant through 4 years, but attenuated or no longer present after 8 to10 years937939 . The proportion of patients with improvement in their predominant symptom was significantly greater with initial surgery compared to non-surgical therapy after 1 and 4 years (55% vs. 28%, p=0.003 and 70% vs. 52%, p=0.05, respectively), but not after 8 to 10 years (54% vs. 42%, p=0.3)937-939. Satisfaction with current status was also similar after 10 years (55% vs. 49%, p=0.5). Back-related functional status persistently moderately favored initial surgical treatment (mean change after 8 to 10 years -7.3 vs. -1.2 on modified RDQ scale, p=0.02). Among patients who initially had surgery, 23% underwent reoperation, and among patients who initially received nonsurgical treatment, 39% subsequently underwent surgery.
EVIDENCE REVIEW
Interspinous spacer device versus non-surgical therapy for spinal stenosis or degenerative spondylolisthesis
For chronic (>6 months) one or two level spinal stenosis with pain relieved with flexion, one higher-quality231 (n=75) and one lower-quality trial251, 940, 941 (n=200) both found the X STOP interspinous spacer device substantially superior to non-surgical treatment (epidural injection, NSAIDs, analgesics, physical therapy) for achieving an overall treatment success through two years based on the Zurich Claudication Questionnaire criteria (48% vs. 5% at 2 years940, 941) or a composite outcome for overall treatment success (63% vs. 13%231). At two years, the interspinous spacer device was superior to non-surgical therapy on the SF-36 bodily pain subscale in one trial reporting this outcome (mean difference in change from baseline about 19 points)941, but in both trials differences on the SF-36 mental and physical component subscales were small or not statistically significant. Effects on rates of subsequent laminectomy were mixed One trial941 found the interspinous spacer associated with lower rates of subsequent laminectomy compared to initial non-surgical therapy (6% vs. 22%), but the other trial231 found no difference in rates of laminectomy (12% vs. 12%). The device manufacturer funded both trials. No trial has compared an interspinous spacer device to standard decompressive surgery.
Table 98. Trials of interspinous spacer device versus non-surgical treatment for spinal stenosis
Author, year Population evaluated Anderson, 231 2006 Chronic symptomatic one- or twolevel spinal stenosis with symptoms relieved by forward flexion Zucherman, 2004251, 940, 941 Chronic symptomatic one- or twolevel spinal stenosis with symptoms relieved by forward flexion
Surgical intervention X-Stop interspinous spacer
X-Stop interspinous spacer
n=200 2 years
Main results Interspinous spacer device versus non-surgical treatment (results at 2 years) Zurich Claudication Questionnaire (0 to 100), mean improvement: -27.35 vs. -3.86 SF-36 Physical component subscale: +9.66 vs. -0.05 SF-36 Mental component subscale: +4.23 vs. -0.26 Patient satisfaction (0 to 5), mean score: 1.55 vs. 2.80 Clinical 'success' (>15 point improvement in Zurich Claudication Questionnaire score, <2.5 patient satisfaction score, and no further surgery): 63% vs. 13% Additional surgery: 12% (5/42) vs. 12% (4/33) Interspinous spacer device versus non-surgical treatment Treatment success (improvement in all three subscales of the Zurich Claudication Questionnaire): 59% vs. 12% at 1 year (p<0.05), 48% vs. 5% at 2 years SF-36 bodily pain, mean score: 56.1 vs. 36.9 at 1 year (p<0.05), 53.8 vs. 34.5 at 2 years (p<0.05) SF-36 physical function, mean score: 62.2 vs. 42.7 at 1 year (p<0.05), 59.3 vs. 41.1 at 2 years (p<0.05) SF-36 physical component subscale: 38.4 vs. 31.2 at 2 years (p<0.05) SF-36 mental component subscale: 54.3 vs. 52.5 (p>0.05) Underwent laminectomy by 2 years: 6% vs. 22%
Quality score* 5/9
2/9
*Excludes criteria involving blinding of patients and care providers, for maximum score of 9 (ODI):Oswestry Disability Index
EVIDENCE REVIEW
Efficacy of different surgical techniques for spinal stenosis or degenerative spondylolisthesis

There is insufficient evidence to determine optimal surgical methods for spinal stenosis with or without degenerative spondylolisthesis. Systematic reviews of randomized trials found inconsistent results or no clear differences between laminectomy plus posterolateral fusion versus laminectomy alone79, 80, 217, 228, 229. The Cochrane review included three small (total n=139) trials906, 931, 932 of patients with spinal stenosis or degenerative spondylolisthesis that found no significant differences between laminectomy and posterolateral fusion (with or without instrumentation) versus laminectomy alone for likelihood of experiencing a poor surgeon-rated outcome, though the trend favored laminectomy plus fusion (OR=0.44, 95% CI 0.13 to 1.48). A second higher-quality systematic review found laminectomy plus fusion superior to laminectomy alone for overall clinical success (RR 1.40, 95% CI 1.04 to 1.89, 7 studies)217. However, these results include data from 5 comparative observational studies and only 2 small (n=50 and n=43) randomized trials (RR 2.15, 95% CI 1.43 to 3.23)906, 932. One of these trials (rated lower-quality) may have skewed results because it reported an unusually high rate of successful clinical outcomes with fusion plus decompressive laminectomy (96%)932. Three systematic reviews evaluated evidence for instrumented versus non-instrumented fusion for mixed degenerative conditions, but did not report results separately for degenerative spondyolisthesis (see section on surgery for non-radicular low back pain)79, 80, 210, 226. A fourth (higher-quality) systematic review found no significant difference between instrumented versus non-instrumented fusion for degenerative spondyolisthesis (RR=1.19, 95% CI 0.92 to 1.54 based on 3 trials and 2 observational studies)217. For a mixed population of patients with degenerative or isthmic spondylolisthesis, one higherquality trial found instrumented circumferential fusion moderately to substantially superior to instrumented posterolateral fusion at long-term (5 to 9 years) follow-up on all four areas of the Dallas Pain Questionnaire, moderately superior on the ODI (mean score 28 vs. 40, p=0.004), and substantially superior on current back pain scores (mean score 3 vs. 6 on a 0 to 10 scale, p=0.021)927, 929. Differences on the SF-36 physical component and mental component summaries were small (5 to 6 points) but statistically significant. One small (n=25), lower-quality trial935 included in one lower-quality systematic review found no differences in clinical outcomes after 2 years between use of 1 versus 2 stand-alone posterolateral interbody fusion BAK cages for L4-5 degenerative spondyolisthesis, though use of 1 cage was associated with less operative blood loss and operation time225. The Cochrane review included one trial934 of laminectomy versus multiple laminotomy (partial laminectomy) for spinal stenosis that found no differences in clinical outcomes or spondylolisthesis progression, though confounding factors (including inconsistent surgical techniques and crossovers) may have affected results79, 80.
EVIDENCE REVIEW
Predictors of postoperative outcomes in lumbar spinal stenosis

For patients who underwent surgery for lumbar spinal stenosis, one lower-quality systematic review (21 studies, 8 higher-quality) found reported walking capacity (better preoperative walking capacity predicted better postoperative capacity) and depression (baseline depression predicted worse postoperative outcomes) to be significant predictors of postoperative outcomes254. Age and gender, the most frequently evaluated factors, only predicted outcomes in one of twelve studies. Harms No operative deaths were reported in four randomized trials of decompressive surgery versus non-surgical therapy for spinal stenosis236, 241, 243, 250. Dural tears were the most common operative complication, occurring in 7% to 11% of patients236, 241, 250. In two trials, neural injuries, vascular injury, and misplaced transpedicular screw were each reported in 1 patient undergoing surgery (total n=466)236, 241. In the observational Maine Lumbar Spine Study, neural injury occurred in 2.5% and dural tear in 10% of 81 operated patients938. Among 142 patients randomized to the X STOP interspinous spacer device, there was one case each of an incision complication requiring antibiotics, respiratory distress, pulmonary edema, and an ischemic coronary episode231, 251. Two malpositioned spacer devices, one implant dislodgement/migration, and one spinous process fracture were also reported. Costs A higher-quality decision analysis estimated incremental cost-effectiveness of $56,500/QALY for laminectomy with noninstrumented fusion versus laminectomy without fusion in patients with degenerative spondylolisthesis and spinal stenosis942. The cost-effectiveness ratio of instrumented fusion compared with noninstrumented fusion was $3,112,800/QALY. However, this estimate was sensitive to the proportion of patients experiencing symptom relief after surgery, and could be as low as $82,400/QALY if the proportion of patients experiencing symptom relief was 90% with instrumented fusion and 80% with noninstrumented fusion. For spinal stenosis, estimated costs of laminectomy alone and laminectomy plus noninstrumented or instrumented fusion were $12,615, $18,495, and $25,914 in a study published in 1997943.
Summary of evidence
For spinal stenosis with or without degenerative spondylisthesis, two small, higher-quality trials found standard decompressive surgery moderately to substantially superior to initial nonsurgical therapy for pain, function, or improved overall outcome at 1 year, but differences are attenuated with longer term follow-up. A well-designed, large observational study reported similar results, though surgery remained moderately superior for back-specific functional status through 10 years (level of evidence: fair). For spinal stenosis with degenerative spondylisthesis, a large, higher-quality trial found no differences between decompressive surgery and non-surgical therapy based on an intentionto-treat analysis, but results are difficult to interpret because of high rates of crossover in both
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain intervention groups. On-treatment analyses found decompressive surgery moderately superior to non-surgical therapy for both pain and function (level of evidence: fair). For spinal stenosis without degenerative spondylisthesis, a large, higher-quality trial found no differences between decompressive surgery and non-surgical therapy based on an intentionto-treat analysis, but results are difficult to interpret because of high rates of crossover in both intervention groups. On-treatment analyses found decompressive surgery slightly to moderately superior to non-surgical therapy for both pain and function (level of evidence: fair). For one- or two-level spinal stenosis relieved by flexion or sitting, two trials (one higherquality) found an interspinous spacer device moderately superior to non-surgical therapy for pain and function through two years (level of evidence: fair). For degenerative spondylolisthesis, there was a trend towards superior clinical outcomes following decompressive laminectomy plus posterolateral fusion compared to decompression alone in three small, lower-quality trials, but results may be skewed by a trial that reported unusually good results with laminectomy plus fusion (level of evidence: poor to fair). For degenerative spondylolisthesis, there was no difference between instrumented and noninstrumented fusion in three trials (level of evidence: fair). For mixed degenerative or isthmic spondyolisthesis, one higher-quality trial found circumferential instrumented fusion moderately superior to instrumented posterolateral fusion for function and substantially superior for pain through 5 to 9 years follow-up (level of evidence: fair). For spinal stenosis, one lower-quality trial found no differences between laminectomy versus multiple laminotomy (level of evidence: poor). For spinal stenosis with or without degenerative spondylolisthesis, decompressive surgery for spinal stenosis (with or without degenerative spondylolisthesis) was associated with no operative mortality four randomized trials. Neural injuries occur in up to 2.5% of operations and dural tear in about 10% (level of evidence: fair). For spinal stenosis with or without degenerative spondylolisthesis, placement of an interspinous spacer device was associated with a malpositioned spacer device in 1.4% of 142 cases, with other complications occurring in less than 1% of cases (level of evidence: fair).

The AHCPR guidelines found that elderly patients with spinal stenosis who can adequately function can be managed without surgery, and surgery should normally not be considered in the first three months of symptoms. Decisions on treatment should take into account patient preferences, lifestyle, surgical risk, and co-morbid medical problems, and should not be based solely on imaging tests, but take into account degree of neurogenic claudication symptoms, associated limitations, and detectable neurologic compromise (strength of evidence: D).
EVIDENCE REVIEW
Surgery for radiculopathy with herniated lumbar disc

The purpose of surgery for symptomatic lumbar disc prolapse is to relieve pressure on affected nerve roots by removing part of, or the entire, disc. Standard open discectomy involves removal of the disc via a standard surgical incision and surgery performed with direct visualization. It is often performed with laminectomy (removal of the vertebral lamina). Microdiscectomy, which is often also considered an open procedure, involves a small incision made in the back and use of an operating microscope to perform hemilaminotomy (removal of part of the lamina in order to adequately visualize the disc) and removal of the disk fragment compressing the affected nerves. It can be performed on an outpatient basis. A variety of minimally invasive techniques for performing discectomy are also available, including discectomy performed with endoscopic guidance and minimally invasive surgery with lasers to vaporize parts of the disc, automated percutaneous discectomy (using a pneumatically driven, suction-cutting probe), Coblation nucleoplasty (using a catheter emitting low-frequency radio waves to vaporize and heat parts of the nucleus), and the disc Dekompressor, a device for disc nucleus extraction that involves a rapidly rotating probe and autosuction. Results of search: systematic reviews We identified a higher-quality Cochrane review on efficacy of different surgical techniques for lumbar disc prolapse81, 82. We also identified five lower-quality systematic reviews that focused on efficacy of laser lumbar discectomy211, 219, endoscopic laser foraminoplasty220, automated percutaneous mechanical discectomy222, or fusion227 for treatment of symptomatic lumbar disc prolapse. We also identified one systematic review of Coblation nucleoplasty, but it identified no trials and is discussed in Key Question 8 because it focused on efficacy for degenerative disc disease88. We excluded earlier versions of the Cochrane review176, 177 and three other outdated systematic reviews190, 191, 256. Results of search: trials 35 trials evaluated surgery for radiculopathy with herniated lumbar disc110, 234, 237-240, 248, 249, 842, 869, 870, 873-878, 944-962 . Thirty of the trials were included in at least one of the six systematic reviews81, 82, 211, 219, 220, 222, 227 and we identified five additional trials (reported in six articles110, 234, 237-240). Four trials compared discectomy versus non-surgical therapy237-239, 248, 249; the remainder compared different surgical techniques. We excluded two trials published only as conference abstracts262, 264.
Efficacy of discectomy versus non-surgical treatment for radiculopathy with herniated lumbar disc
Four trials compared surgery to non-surgical therapy (Table 99)237-239, 248, 249. We rated three trials higher-quality237-239, 249. Each trial enrolled patients with sciatica present for at least six weeks. Baseline pain scores averaged about 20 points on the 0 to 100 SF-36 bodily pain score (lower scores indicate worse pain) in two trials238, 239, 249 and 60 on a 0 to 100 pain scale (higher scores indicate worse pain) in the third237. The fourth and oldest trial was rated lower-quality248. It enrolled patients unresponsive to two weeks of inpatient non-surgical treatment and did not report severity of baseline pain.
EVIDENCE REVIEW
Table 99. Trials of discectomy versus non-surgical therapy for radiculopathy with prolapsed lumbar disc
Author, year Population evaluated Osterman, 2006237 Radiculopathy for 6 to 12 weeks with imaging-confirmed lumbar disc prolapse Surgical intervention Microdiscectomy Number of patients Duration of follow-up n=58 2 years
Peul, 2007238, 239 Radiculopathy for 6 to 12 weeks with imaging-confirmed lumbar disc prolapse
Microdiscectomy
n=283 2 years
Main results Microdiscectomy vs. non-operative treatment (intention-to-treat, mean differences at 2 years, positive values favor microdiscectomy) Leg pain (0 to 100 scale): 9 (95% CI -1 to 20) Back pain (0 to 100 scale): 7 (95% CI -3 to 17) ODI (0 to 100 scale): 3 (95% CI -4 to 10) 15D Health-related quality of life (0 to 1.0 scale): 0.03 (-0.01 to 0.07) Subjective work ability (0 to 100 scale): 5 (95% CI -7 to 18) At 6 weeks, only leg pain superior in microdiscectomy group: mean score 12 vs. 25 On-treatment analyses (including 11 patients who crossed over to surgery): No differences for any outcomes Microdiscectomy vs. non-operative treatment (mean difference, negative values favor surgery except for SF-36 where positive values favor surgery) RDQ: -3.1 (95% CI -4.3 to -1.7) at 8 weeks, -0.8 (95% CI -2.1 to +0.5) at 26 weeks, 0.4 (95% CI -1.7 to +0.9) at 1 year, and -0.5 at 2 years (95% CI -1.8 to +0.8) VAS score for leg pain (0 to 100): -17.7 (95% CI -23.1 to -12.3) at 8 weeks, -6.1 (95% CI -10.0 to -2.2) at 26 weeks, 0 (95% CI -4.0 to +4.0) at 1 year, and +2 at 2 years (95% CI -2.0 to +6.0) VAS score for back pain (0 to 100): -11.3 (95% CI -17.4 to -5.6) at 8 weeks, -2.3 (95% CI -8.2 to +3.6) at 26 weeks, -2.3 (95% CI -8.2 to +3.6) at 1 year, and -1.4 (95% CI -6.3 to +4.5) at 2 years SF-36 Bodily Pain: +8.4 (95% CI 3.2 to 13.5) at 8 weeks, +3.3 (-1.8 to +8.4) at 26 weeks, +2.7 (95% CI -2.6 to +7.9) at 1 year, SF-36 Physical Functioning: +9.3 (95% CI +4.4 to +14.2) at 8 weeks, +1.5 (95% CI 3.4 to +6.4) at 26 weeks, +2.2 (95% CI -2.8 to +7.2) at 1 year, -1.3 (95% CI -6.3 to +3.7) at 2 years Recovery (defined as complete or nearly complete disappearance of symptoms as measured on a 7-point Likert scale): 81% vs. 36% at 8 weeks, 77% vs. 71% at 26 weeks, 86% vs. 82% at 1 year, 81% vs. 79% at 2 years (hazards ratio 1.97, 95% CI 1.7 to 2.2, at 1 year)
Quality* 6/9
7/9
EVIDENCE REVIEW
Table 99. Trials of discectomy versus non-surgical therapy for radiculopathy with prolapsed lumbar disc
Author, year Population evaluated Weber, 1983248 Radiculopathy unresponsive to two weeks of non-surgical inpatient treatment and with imaging-confirmed lumbar disc prolapse Surgical intervention Open discectomy Number of patients Duration of follow-up n=126 10 years
Weinstein, 2006249 Spine Outcomes Research Trials Radiculopathy for >6 weeks with imaging-confirmed lumbar disc prolapse
Open discectomy
n=501 2 years
Main results Discectomy versus initial non-surgical treatment 'Good' result (patient completely satisfied): 65% (39/60) vs. 36% (24/66) at 1 year, 67% (40/60) vs. 52% (34/66) after 4 years, 58% (35/60) vs. 56% (37/66) after ten years 'Poor' or 'bad' results: 8% (5/60) vs. 21% (14/66) at 1 year (OR=0.34, 95% CI 0.12 to 1.02), 14% (8/57) vs. 12% (8/66) after 4 years (OR=1.21, 95% CI 0.42 to 3.46), and 7% (4/55) vs. 6% (4/66) after 10 years (OR=1.22, 95% CI 0.29 to 5.10) Proportion with no low back pain: 60% (36/57) vs. 58% (38/66) at 4 years, 84% (43/51) vs. 79% (52/66) at 10 years Proportion with no radiating pain: 79% (45/57) vs. 68% (45/66) at 4 years, 98% (54/55) vs. 98% (65/66) at 10 years Standard open discectomy vs. non-operative treatment, intention-to-treat analyses (mean difference, negative values favor surgery) SF-36 bodily pain (0 to 100): -2.9 (95% CI -8.0 to 2.2) at 3 months; -3.2 (-8.4 to 2.0) at 2 years SF-36 physical function (0 to 100): -2.8 (95% CI -8.1 to 2.5) at 3 months; 0 (95% CI -5.5 to 5.4) at 2 years ODI: -4.7 (95% CI -9.3 to -0.2) at 3 months; -2.7 (95% CI -7.4 to 1.9) at 2 years Sciatica Bothersomeness Index (0 to 24): -2.1 (95% CI -3.4 to -0.9) at 3 months; 1.6 (95% CI -2.9 to -0.3) at 2 years Work status, satisfaction with symptoms, satisfaction with care: No significant differences SF-36 bodily pain scale: -15.0 (95% CI -19.2 to -10.9) at 1 year SF-36 physical function scale: -17.5 (95% CI -21.5 to -13.6) at year ODI: -15.0 (95% CI -18.3 to -11.7) Sciatica Bothersomeness Index: -3.2 (95% CI -3.2 to -2.1)
Quality* 4/9
6/9
*Excludes criteria involving blinding of patients and care providers, for maximum score of 9 VAS: Visual Analogue; RDQ: Roland Morris Disability Questionnaire; ODI: Oswestry Disability Index; SMD: standardized mean difference; SF-36:Short-form 36
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain For radiculopathy with concordant herniated lumbar disc on imaging, the first trial (n=128) to compare surgery to non-surgical treatment was published in 1983. It found standard open discectomy associated with a lower likelihood of poor results compared to non-surgical therapy after one year (OR 0.38, 95% CI 0.14 to 0.99), but not after four or ten years (OR 1.21, 95% CI 0.42 to 3.45 and OR 1.21, 95% CI 0.29 to 5.10, respectively)248. One quarter of patients randomized to initial non-surgical therapy eventually underwent surgery. This trial was rated lower-quality, in part because standards for reporting and design of randomized trials have become more stringent. Two higher-quality trials evaluated microdiscectomy versus non-surgical therapy237-239. The larger of the trials (n=283) found initial microdiscectomy moderately superior to initial nonoperative treatment on the RDQ (3 points) and leg and back pain scores (18 and 11 points, respectively, on 0 to 100 VAS scales) at 8 weeks. By 26 weeks, differences in pain scores were small (6 points for leg pain) or no longer present (back pain), and there were no differences in pain scores after 1 or 2 years238, 239. Early differences in SF-36 bodily pain and physical functioning scores were small (8 to 9 points) and did not reach statistical significance. By 26 weeks, both groups had improved by 40 to 50 points on both scores. Patients assigned to initial surgery reported a faster rate of perceived recovery at 1 year (hazard ratio 1.97, 95% CI 1.72 to 2.22), but differences in the proportion of patients who experienced recovery were only present at 8-week follow-up (81% vs. 36%). By 26 weeks, recovery rates were similar (79% vs. 78%). A smaller (n=58) trial found microdiscectomy moderately superior to non-surgical treatment (isometric exercises) for leg pain (but not back pain, the ODI, or other outcomes) at six weeks, but no difference on any outcome assessed at 3 months to 2 years237. In both trials, approximately 40% of patients assigned to initial non-surgical therapy underwent surgery, which could have attenuated benefits associated with surgery in intention-to-treat analyses. The large (n=501), multicenter, higher-quality Spine Outcomes Research Trial (SPORT)936 found no differences between standard open discectomy or microdiscectomy (technique left to discretion of the surgeon) versus non-surgical therapy based on an intention-to-treat analysis (the exception being a five point different in improvement in ODI scores at 3 months), but interpretation of these findings is complicated by low rates of adherence to treatment assignments249. At the end of the trial, only 60% of patients randomized to surgery had undergone discectomy, and 45% randomized to non-operative treatment had undergone surgery. In on-treatment analyses adjusted for potential confounders, surgery was moderately superior by about 15 points on ODI scores and SF-36 bodily pain and physical function scales after one year, and differences remained statistically significant through two years. Some significant baseline differences were present between those who crossed over and those who remained on their original treatment assignment. Regardless of treatment allocation, improvement averaged 30 to 40 points on the ODI and SF-36 bodily pain and physical function scales after 2 years. Results of a concurrent prospective cohort study were consistent with ontreatment analyses from the randomized trial963 and a combined analysis of the cohort study plus combined on-treatment results from the randomized trial found that benefits persisted through 4 years242.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain A higher-quality Cochrane review81, 82 also found discectomy superior to non-surgical therapy for short-term outcomes, but only included two248, 249 trials of surgery versus non-surgical therapy and one other trial262 only available as a conference abstract. The Maine Lumbar Spine Study, a well-designed, long-term (10 years follow-up) prospective cohort study (n=507) also found initial treatment with surgery associated with greater likelihood for improvement in the predominant symptom (either back or leg pain) at 1 year compared to initial non-surgical treatment (71% vs. 43%, p<0.001) for lumbar disc prolapse with radiculopathy, though differences were attenuated after 5 years (70% vs. 56%, p<0.001) and no longer significant after 10 years (69% and 61%, p=0.20)964-966. Patients initially treated surgically were also more likely to report long-term resolution of low back and leg pain (56% vs. 40%, p=0.006) and greater improvements in RDQ scores. Work and disability status were comparable between groups at all follow-up evaluations. About one-quarter of patients in either group underwent additional or subsequent back surgery. Another, lower-quality observational study (did not adjust for baseline differences or confounders) found that fewer patients (n=342) who initially underwent surgery reported their low back condition as unchanged or worse after 13 years compared to those who received initial non-surgical treatment (19% vs. 41%), though similar proportions reported sciatica (67% vs. 68%) and being disabled due to a back problem (20% vs. 20%)967. There were also no differences in long-term functional status.
Efficacy of discectomy versus chemonucleolysis for lumbar disc prolapse

Evidence on discectomy versus chemonucleolysis for lumbar disc prolapse is discussed in more detail in the section on chemonucleolysis (Key Question 8). Briefly, discectomy was generally superior to chemonucleolysis for patient-reported outcomes, surgeon-reported outcomes, and rates of additional or subsequent surgery in all trials included in the Cochrane review81, 82. However, differences were not always statistically significant.
Efficacy of discectomy versus epidural steroid injection for lumbar disc prolapse
Evidence on discectomy versus epidural steroid injection for lumbar disc prolapse is discussed in more detail in the section on epidural steroid injections (Key Question 8). Briefly, one higherquality trial842 included in the Cochrane review81, 82 found discectomy superior for short-term (one to three month) outcomes related to pain relief, functional status, motor deficits, and use of medications, though differences were no longer significant after 2-3 years of follow-up. Results are difficult to interpret because about one-third of the patients assigned to epidural steroids crossed over to surgery, and intention-to-treat results were not reported.
Efficacy of laser-assisted discectomy

The Cochrane review81, 82 included two trials of laser discectomy, neither of which compared laser discectomy versus non-surgical therapy, standard open discectomy, or microdiscectomy. One trial only reported in conference abstracts found chemonucleolysis superior to laser discectomy877, 968. The other trial only compared two type of lasers959. Three lower-quality systematic reviews of laser discectomy (with211 or without endoscopy219) and endoscopic laser foraminoplasty220 identified one additional trial264 that compared laser lumbar discectomy versus
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain epidural steroids, but it is also only published as a conference abstract. It found no differences between interventions on any outcome, and the trial was aborted before completion.
Efficacy of Coblation nucleoplasty or disc Dekompressor

We found no trials on efficacy of nucleopasty or the disc Dekompressor for lumbar disc prolapse with radiculopathy.
Efficacy of different surgical techniques for lumbar disc prolapse

There is insufficient evidence to determine optimal surgical methods for radiculopathy with prolapsed disc. Four trials found no clear differences between microdiscectomy and standard open discectomy234, 950, 956, 962. There are no published trials of Coblation nucleoplasty, disc Dekompressor, or laserassisted methods219, 220, and insufficient evidence from sparse, lower-quality, and primarily small (N<100) trials945, 949, 958, 961, 969, 970 to reliably evaluate sequestrectomy, automated percutaneous discectomy or percutaneous endoscopic discectomy211, 222. One larger trial (n=178) found endoscopic (interlaminar or transforaminal) discectomy superior to microdiscectomy for days of postoperative work disability (25 vs. 49, p<0.01) and postoperative pain (data not reported), with no differences in pain and ODI scores through 12 months, but it had a number of methodological shortcomings, including inadequate randomization method (alternate allocation) and lack of intention-to-treat analysis240. The Cochrane review also found that an inter-position gel covering the dura (five trials) and of fat (four trials) appear to reduce scar formation, but insufficient evidence to draw conclusions about effects on clinical outcomes81, 82. Harms No operative deaths were observed in randomized trials and large observational studies of standard open discectomy or microdiscectomy versus non-surgical therapy (total number of patients undergoing surgery about 1400)237, 238, 249, 963, 964. The most common complication associated with surgery for lumbar disc herniation was dural tear, which occurred in 1% to 4% of operations238, 249, 963, 964. Reoperation occurred in 3% to 7% of patients within 1 year238, 249, 963 of initial surgery and 9% within 2 years.249, 963 In the SPORT randomized trial and observational cohort, no complications were reported in 95% of open discectomies.249, 963 No cases of cauda equina syndrome occurred in patients randomized to non-operative treatment. Costs One study estimated a cost-effectiveness ratio of $12,000 to $33,900/QALY (depending on the cost of discectomy) for surgery for prolapsed disc relative to continued non-surgical management971. Another trial found similar costs for automated percutaneous lumbar discectomy and microdiscectomy for contained lumbar disc herniation (automated percutaneous lumbar discectomy associated with poorer outcomes)972.
EVIDENCE REVIEW
Summary of evidence
For lumbar disc prolapse with radiculopathy, two higher-quality RCTs and two well-designed observational studies found standard open discectomy associated with small to moderately improved outcomes at 3 months to 4 years compared to initial non-surgical therapy (or delayed surgery). Patients who received either initial surgery or non-surgical treatment both experienced moderate improvements in pain and functional status. In some studies, benefits of surgery were attenuated or no longer present at longer-term follow-up. Interpretation of results is complicated by high rates of nonadherence to assigned therapies in some trials (level of evidence: good). For lumbar disc prolapse with radiculopathy, two higher-quality trials found microdiscectomy moderately superior to initial non-surgical therapy for pain relief (2 trials) and function (1 trial) after 6 to 8 weeks, though differences were no longer present after 1 to 2 years (level of evidence: good). For lumbar disc prolapse with radiculopathy, chemonucleolysis was consistently associated with trends towards worse outcomes compared to standard discectomy in five lower-quality trials, and was associated with subsequent surgery in about 30% of cases (level of evidence: fair). For lumbar disc prolapse with radiculopathy, two lower-quality trials found inconsistent evidence on efficacy of automated percutaneous discectomy versus chymopapain chemonucleolysis, with one trial finding chemonucleolysis superior and the other finding no differences in functional status or rates of neurologic deficits (level of evidence: poor). One lower-quality trial found low-dose chymopapain chemonucleolysis plus transforaminal posterolateral endoscopic discectomy associated with a slightly lower rate of recurrent herniation compared to endoscopic discectomy alone, but there were no differences on other outcomes (level of evidence: poor). One trial found epidural steroids superior to discectomy for short-term but not longer-term outcomes, but results are difficult to interpret because crossover rates were high and intention-to-treat results not reported (level of evidence: poor). Four trials (three lower-quality) found no clear differences between standard open discectomy and microdiscectomy (level of evidence: fair). One lower-quality trial found no clear differences between percutaneous endoscopic discectomy (used modified forceps and an automated cutter with suction) versus microdiscectomy (level of evidence: poor). There is mixed evidence from two lower-quality trials on efficacy of automated percutaneous discectomy versus microdiscectomy, with one trial reporting similar outcomes and the other poorer outcomes with automated percutaneous discectomy (level of evidence: poor). There is insufficient evidence to judge efficacy of laser discectomy or foraminoplasty (level of evidence: poor). There are no randomized trials of Coblation nucleoplasty or disc Dekompressor.
EVIDENCE REVIEW
APS Clinical Guideline for the Evaluation and Management of Low Back Pain Use of inter-position membranes may reduce scar formation, but there is insufficient evidence from eight trials to determine whether they improve clinical outcomes (level of evidence: poor). In randomized trials and well-designed observational studies, open discectomy was associated with no operative mortality in over 1400 cases and no complications in 95% of operations. Dural tear was the most common complication. No cases of cauda equina syndrome were observed in patients that received initial non-surgical treatment (level of evidence: fair).

The AHCPR guidelines found that patients with acute low back pain who do not have findings suggestive of nerve root compression or positive red flags do not need surgical consultation for possible herniated lumbar disc (strength of evidence: D). The AHCPR guidelines recommend discussing further treatment options after 1 month of conservative therapy in patients with sciatica, and consider referral to a specialist when all of the following are met: 1) sciatica is both severe and disabling, 2) symptoms of sciatica persist without improvement or with progression, 3) there is clinical evidence of nerve root compromise (strength of evidence: B). The AHCPR guidelines found standard discectomy or microdiscectomy appropriate for selected patients with herniated discs and nerve root dysfunction (strength of evidence: B). The AHCPR guidelines recommend against percutaneous discectomy in patients with lumbar disc herniation because of poor efficacy relative to chymopapain (strength of evidence: C).
EVIDENCE REVIEW
Table 100. Systematic reviews on efficacy of surgery for low back pain
Author, year Andersson, 72 2006
Median duration of follow-up (range) 2 years (1 to 6 years)
Median sample size in included RCTs (range) 69 (11 to 279)
Interventions evaluated (number of RCTs) Fusion (9)
Main conclusions Lumbar fusion for presumed degenerative disc disease: Median=67% (range 17% to 100%) for proportion reporting good or excellent results after fusion (16 studies, including observational data) Lumbar fusion versus non-operative treatment for non-specific LBP (2 RCTs): conflicting results (no difference in one RCT and fusion superior in the other) Instrumented versus noninstrumented fusion for non-radicular low back pain with common degenerative findings (3 RCTs) or degenerative spondylolisthesis (3 RCTs) Proportion with good or excellent results: 75% vs. 79% (all studies, including observational data); instrumentation improved outcomes in one of three RCTs. Vertebral disc replacement for degenerative disc disease: Range 50% to 81% for good or excellent results (7 observational studies), range 7% to 46% for secondary surgery (3 observational studies) Vertebral disc replacement with Charite prosthetic disc vs. anterior interbody lumbar fusion with BAK cage for single level degenerative disc disease (1 completed RCT): 57% vs. 46% met all criteria for surgical success (p<0.0001 for equivalence test)
Surgery for non-radicular low back pain with common degenerative changes (20 unique RCTs in 11 systematic reviews)
Bono, 2004210
Quantitative
3 (not rated)
Range 1 to 2 years
Range 11 to 179
Fusion (3)
de Kleuver, 2003212
Qualitative
No RCTs
Not applicable
Not applicable
Not applicable
Not applicable
Freeman, 2006213
Quantitative
2 (not rated)
2 years
78 and 309
Fusion (2), prosthetic disc (2)
EVIDENCE REVIEW
Author, year Geisler, 2004214
Median duration of follow-up (range) Range 2 to 4 years
Median sample size in included RCTs (range) Range 46 to 304
Interventions evaluated (number of RCTs) Fusion (4), prosthetic disc (1)
Main conclusions 360 degree lumbar fusion via ALIF, PLIF, or TLIF vs. stand-alone ALIF or PLIF for non-specific LBP or degenerative disc disease Weighted mean change in VAS: -49.1% (13 studies) vs. -45.5% (7 studies). Weighted mean change in mean ODI: -20.6% (5 studies) vs. -27.9%(13 studies)
EVIDENCE REVIEW
Author, year Gibson, 200579, 80
Median duration of follow-up (range) 14 months (6 to 48 months)
Interventions evaluated (number of RCTs) Fusion (14), prosthetic disc (2)
Main conclusions Fusion vs. non-surgical treatment for nonspecific LBP (2 RCTs): Surgery superior for back to work and patient rating at 2 years in 1 RCT (vs. physical therapy), but no differences for 1 year outcomes in other RCT (vs. multidisciplinary rehab) Prosthetic vertebral disc vs. fusion (2 RCTs): Small numbers, but no statistically significant differences between interventions Fusion with instrumentation vs. fusion without instrumentation (8 RCTs): OR 0.43 (95% CI 0.21 to 0.91) for no fusion (8 RCTs), OR 0.64 (95% CI 0.35 to 1.17) for poor clinical outcome (8 RCTs). Comparisons of anterior, posterior, and combined fusion (4 RCTs): Conflicting results; unable to draw conclusions about comparative effectiveness of anterior, posterior, or circumferential fusion Electrical stimulation vs. no electrical stimulation: OR 0.38 (95% CI 0.22 to 0.64) for no fusion after non-instrumented fusion (3 RCTs); OR 0.59 (95% CI 0.15 to 2.30) for no fusion after instrumented fusion. No significant effect on clinical outcomes in 2 RCTs Lumbar fusion vs. non-operative treatment for non-specific LBP ODI (mean difference between interventions in improvement from baseline, negative values favor fusion, 3 RCTs): -4.13 (95% CI -9.08 to 0.82)
Ibrahim, 215 2008
Quantitative
3 (3)
Range 1 to 2 years
Range 60 to 349
Fusion (3)
EVIDENCE REVIEW
Author, year Mirza, 2007218
Main conclusions Lumbar fusion vs. non-operative treatment for non-specific LBP: Range -8.8 to +3.9 for mean difference in improvement on ODI (4 RCTs); range -2.3 to +3.9 for fusion vs. intensive rehabilitation (3 RCTs) Vertebral disc replacement vs. anterior interbody lumbar fusion with BAK cage for single-level degenerative disc disease (1 RCT): Total disc replacement superior for proportion of patients with improvement in ODI (at least 25% improvement): 62% vs. 49%, p=0.04 Lumbar fusion vs. non-operative-treatment for non-specific LBP (2 RCTs): Fusion superior to standard non-operative treatments in 1 RCT (n=294), fusion no better than intensive rehabilitation in 1 RCT (n=64) Interbody fusion associated with higher fusion rates compared with posterolateral fusion for back pain due to degenerative disc disease limited to 1 or 2 levels Conflicting evidence on effects of interbody fusion on functional outcomes. No clear differences between different interbody fusion techniques
NICE, 2004221
Qualitative
1 (not rated)
2 years
304
Fusion (1), prosthetic disc (1)
Resnick, 2005224
Qualitative
2 (1)
1 and 2 years
61 and 264
Fusion (2)
Resnick, 2005225
Qualitative
2 (2)
2 years and 35 months
53 and 264
Fusion (2)
EVIDENCE REVIEW
Median duration of follow-up (range) 2 years (16 to 26 months)
Main conclusions Lumbar fusion with pedicle screw fixation to PLF for non-specific low back pain or degenerative disc disease (5 RCTs) or degenerative spondylolisthesis (3 RCTs) increases radiologic fusion success when assessed by plain x-ray with dynamic imaging (supported by all Class I and the majority of Class II and Class III evidence). No convincing clinical correlation between radiographic fusion and clinical outcome. Lumbar fusion with pedicle screw fixation: conflicting evidence on clinical outcomes (primarily Class II and III evidence). The largest contemporary RCT did not show a benefit with pedicle screw fixation
EVIDENCE REVIEW
Author, year Gibson, 79, 80 2005
Type of systematic review Quantitative and qualitative
Interventions evaluated (number of RCTs) Fusion (4), laminectomy (1)
Main conclusions Posterolateral fusion vs. non-surgical treatment for isthmic spondylolisthesis (1 higher-quality RCT): Surgery superior for pain and disability though not occupational outcomes at 2 years (no data on relief of sciatica) Fusion (with or without instrumentation) + laminectomy vs. fusion alone (with or without instrumentation) for isthmic spondylolisthesis (1 higher-quality RCT): Fusion + laminectomy inferior for rates of pseudoarthrosis (22% vs. 0%, p=0.02) and unsatisfactory results (33% vs. 4%, p=0.01) Instrumented vs. non-instrumented posterolateral fusion for isthmic spondylolisthesis (1 higher-quality RCT): No differences Combined vs. posterior fusion for isthmic spondylolisthesis: 86% vs. 75% (p=0.0045) for successful clinical results (includes observational data) Combined vs. anterior fusion for isthmic spondylolisthesis: 86% vs. 90% (p=0.65) for successful clinical results (includes observational data) Posterior vs. anterior fusion for isthmic spondylolisthesis: 75% vs. 90% (p=0.0047) for successful clinical results (includes observational data)
Surgery for isthmic spondylo-listhesis (4 unique RCTs in 2 systematic reviews)
Kwon, 2005216
Quantitative
4 (not rated)
Range 2 to 5 years
Range 27 to 111
Fusion (4), laminectomy (1)
EVIDENCE REVIEW
Author, year Bono, 2004210
Main conclusions Instrumented versus noninstrumented fusion for non-radicular low back pain with common degenerative findings (3 RCTs) or degenerative spondylolisthesis (3 RCTs) Proportion with good or excellent results: 75% vs. 79% (all studies, including observational data); instrumentation improved outcomes in one of three RCTs Surgical decompression vs. non-surgical therapy for spinal stenosis (1 RCT): OR 0.09 (95% CI 0.01 to 0.89) for secondary surgery by 4 years; OR 2.43 (95% CI 0.09 to 57.58) for 'bad result' after 10 years. Laminectomy vs. multiple laminotomy for spinal stenosis (1 RCT): no differences. Interspinous spacer device vs. non-surgical therapy (including epidural steroid) for spinal stenosis (1 RCT): 1 year pain and claudication results superior with spacer device Laminectomy plus fusion vs. laminectomy alone for spinal stenosis with or without degenerative spondylolisthesis: OR 0.44 (95% CI 0.13 to 1.48) for poor result (surgeon rated) at 18 to 24 months (3 RCTs); OR 4.69 (95% CI 0.51 to 42.83) for re-operation after 2-4 years (2 RCTs)
Surgery for spinal stenosis with or without degenerative spondylo-listhesis (12 unique RCTs in 8 systematic reviews) Gibson, 79, 80 2005 Qualitative and quantitative 10 (4) 2 2 years (1 to 10 years) 66 (31 to 200) Fusion (7), laminectomy (5), interspinous spacer (1) 6
EVIDENCE REVIEW
Author, year Martin, 2007217
Main conclusions Fusion versus decompression alone for degenerative lumbar spondylolisthesis: RR 2.15 (95% CI 1.43 to 3.23) for satisfactory clinical outcome (2 lowerquality RCTs); RR 1.40 (95% CI 1.04 to 1.89) when pooled with observational studies Instrumented fusion versus noninstrumented fusion for degenerative lumbar spondylolisthesis: RR 1.58 (95% CI 0.60 to 4.12) for satisfactory clinical outcome (3 RCTs, 1 higher-quality); RR 1.19 (95% CI 0.92 to 1.54) when pooled with observational studies; RR 1.96 (95% CI 1.35 to 2.84) for achieving solid fusion (2 RCTs, 1 higher-quality); RR 1.37 (95% CI 1.07 to 1.75) when pooled with observational studies Interspinous spacer implant vs. nonsurgical therapy for lumbar spinal stenosis with neurogenic claudication exacerbated in extension and relieved with flexion (1 RCT): 45% vs. 7% (p<0.001) improvement in symptom severity from baseline at 1 year, 44% vs. 0.4% (p<0.001) improvement in physical function scores at 1 year, 48% vs. 5% fulfilled all Zurich Claudication Questionnaire criteria at 2 years
NICE, 2005223
Qualitative
1 (not rated)
2 years
200
Interspinous spacer device (1)
EVIDENCE REVIEW
Main conclusions Laminectomy + fusion versus laminectomy alone for spinal stenosis with degenerative spondylolisthesis (1 RCT): Fusion superior for excellent or good outcome at 3 years (96% vs. 44%), also for leg and back pain Laminectomy + fusion versus laminectomy alone for spinal stenosis with or without degenerative spondylolisthesis (3 RCTs): No evidence that fusion (with or without instrumentation) provides any benefit over laminectomy alone for lumbar stenosis without evidence of preoperative deformity or instability Circumferential instrumented fusion versus posterolateral fusion (1 lower-quality RCT): Lower re-operation rate through 2 years (7% vs. 22%), leg pain at 1 year (p<0.03), and peak back pain at 2 years (p<0.04); no difference in functional status (2 years) One stand-alone posterolateral interbody fusion BAK cage versus two stand-alone posterolateral interbody BAK cages for L4L5 degenerative Grade I spondylolisthesis (1 lower-quality RCT): No differences
Resnick, 229 2005
Qualitative
3 (0)
Range 2 to 3 years
Range 44 to 50
Fusion (3), laminectomy (3)
Resnick, 225 2005
Qualitative
2 (0)
2 years
25 and 147
Fusion (2)
EVIDENCE REVIEW
Median duration of follow-up (range) 2 years
Main conclusions Lumbar fusion with pedicle screw fixation to PLF for non-specific low back pain or degenerative disc disease (5 RCTs) or degenerative spondylolisthesis (3 RCTs) increases radiologic fusion success when assessed by plain x-ray with dynamic imaging (supported by all Class I and the majority of Class II and Class III evidence) No convincing clinical correlation between radiographic fusion and clinical outcome Lumbar fusion with pedicle screw fixation: conflicting evidence on clinical outcomes (primarily Class II and III). The largest contemporary RCT did not show a benefit with pedicle screw fixation Information about percutaneous endoscopic laser discectomy is very limited and the information available is of poor quality The safety and/or efficacy of the procedure cannot be determined due to an incomplete and/or poor quality evidence base
Surgery for radiculopathy with herniated lumbar disc (30 unique RCTs in 6 systematic reviews) Boult, 2000211 Qualitative No RCTs Not applicable Not applicable Not applicable Not applicable 4
EVIDENCE REVIEW
Author, year Gibson, 200781, 82
Median duration of follow-up (range) 1 year (6 weeks to 10 years)
Interventions evaluated (number of RCTs) Standard discectomy or not specified (12), microdiscectomy (10), laminotomy (2), minimally invasive, laser, or automated percutaneous mechanical discectomy (9)
Main conclusions Discectomy (standard or micro-) versus initial non-surgical therapy Poor/bad result (surgeon-rated): OR 0.38 (95% CI 0.14 to 0.99) at 1 year, OR 1.21 (95% CI 0.42 to 3.45) at 4 years, OR 1.21 (95% CI 0.29 to 5.10) at 10 years (1 RCT); qualitatively, discectomy superior for shortterm outcomes in all 3 RCTs, but differences attenuated at longer follow-up in 2 of the RCTs Microdiscectomy versus standard open discectomy for lumbar disc prolapse (4 RCTs): Outcomes broadly similar (data couldn't be pooled) Automated percutaneous discectomy vs. microdiscectomy (2 RCTs) or chemonucleolysis (2 RCTs) for lumbar disc prolapse: Automated percutaneous discectomy similar to microdiscectomy in 1 trial, inferior for satisfactory results in another (29% vs. 80%); and inferior to chemonucleolysis in 1 trial Percutaneous endoscopic discectomy (cannula inserted into the central disc) vs. microdiscectomy for lumbar disc prolapse (1 RCT): No differences Use of interposition membranes (8 RCTs): Effects on clinical outcomes inconsistent.
EVIDENCE REVIEW
Author, year NICE, 2005222
Median duration of follow-up (range) Range 6 months to 1 year
Interventions evaluated (number of RCTs) Automated perctuaneous mechanical lumbar discectomy (3)
Main conclusions Automated percutaneous lumbar discectomy vs. standard open discectomy for lumbar disc prolapse (1 RCT): 41% (7/17) vs. 40% (4/10) for excellent or good outcome Automated percutaneous lumbar discectomy vs. microdiscectomy for lumbar disc prolapse (1 RCT): 29% (9/31) vs. 80% (32/40) for successful outcome (p<0.001) Automated percutaneous lumbar discectomy vs. chemonucleolysis for lumbar disc prolapse (1 RCT): 44% (30/69) vs. 61% (44/72) for successful outcome (p<0.05) Laser lumbar discectomy vs. epidural corticosteroid injection for lumbar disc prolapse (1 RCT): No difference between groups on ODI or modified MacNab score 3 comparative observational studies only compared results of endoscopic laser foraminoplasty in different populations, or complications only (no efficacy data) of endoscopic laser foraminoplasty vs. historical controls Lumbar fusion for lumbar disc prolapse with radiculopathy: No convincing evidence to support routine use of lumbar fusion at the time of primary lumbar disc excision
NICE, 2003219
Qualitative
1 abstract (not rated)
1 to 26 months
29
Laser discectomy (1)
NICE, 2003220
Qualitative
No RCTs
Not applicable
Not applicable
Not applicable
Not applicable
Resnick, 2005227
Qualitative
No RCTs
Not applicable
Not applicable
Not applicable
Not applicable
*Trials adequately meeting at least half of the quality rating criteria or rated as good or higher-quality if the number of criteria met was not reported **Trials adequately meeting criteria for adequate allocation concealment
EVIDENCE REVIEW
Table 101. Summary of evidence on surgery for low back pain

Number of trials of surgery versus nonsurgical therapy (number rated Population higher-quality) Lumbar interbody fusion Non-radicular low 4 (4) back pain with common degenerative changes Number of trials of surgery vs. non-surgical therapy with >100 patients 2
Total number of trials 18
Net benefit* Small to moderate versus standard physical therapy supplemented by other nonsurgical therapies, no benefit versus intensive rehabilitation No difference versus fusion
Effective vs. nonsurgical therapy Yes versus standard physical therapy (1 trial), no versus intensive rehabilitation (3 trials)
Inconsistency Some inconsistency (see comments)
Comments Inconsistency between trials may be related to use of different comparator interventions
Artificial disc replacement Non-radicular low back pain with single-level degenerative disc disease Posterolateral fusion Isthmic spondylolisthesis Lumbar disc prolapse with radiculopathy 1 (0) 1 6 Moderate Yes (1 trial) Not applicable Direct Poor 2 (1) 2 2 No trials No Direct Fair One trial of the Prodisc II and one trial of the Charit Artificial Disc
Standard open discectomy or micro-discectomy 4 (4) 3 35 Moderate Yes (4 trials) No Direct Good Benefits associated with surgery diminish or no longer present after 3 months follow-up
EVIDENCE REVIEW
Table 101. Summary of evidence on surgery for low back pain

Number of trials of surgery versus nonsurgical therapy (number rated higher-quality) 4 (4) Number of trials of surgery vs. non-surgical therapy with >100 patients 2
Population Spinal stenosis with or without degenerative spondylo-listhesis
Total number of trials 17
Effective vs. nonsurgical therapy Yes (4 trials)
Inconsistency No
Comments Benefits associated with surgery present through 1 to 2 years follow-up Two trials of the X STOP interspinous spacer device
Laminectomy (with or without fusion)
Interspinous spacer device One- or two-level spinal stenosis with symptoms relieved by forward flexion 2 (1) 1 2 Moderate to substantial (pain relief) slight to moderate (function) Yes (2 trials) No Direct Fair
* Based on evidence showing intervention is more effective than placebo or sham therapy for one or more of the following outcomes: pain, functional status, overall improvement, or work status. Versus placebo, small benefit defined as 5-10 points on a 100-point Visual Analogue Scale (VAS) for pain (or equivalent), 1-2 points on the Roland Morris Disability Questionnaire (RDQ), 5-10 points on the Oswestry Disability Index (ODI), or a standardized mean difference (SMD) of 0.2-0.5. Moderate benefit defined as10-20 points on a VAS for pain, 2-5 points on the RDQ, 10-20 points on the ODI, or a SMD of 0.5-0.8. Large benefit defined as >20 points on a 100 point VAS for pain; >5 points on the RDQ, >20 points on the ODI, or a SMD of >0.8. Inconsistency defined as <75% of trials reaching consistent conclusions on efficacy (no effect vs. positive effect considered inconsistent). Trials of artificial disc replacement versus fusion.
FINAL DRAFT EVIDENCE REVIEW

Key Question 10 How effective are combinations of therapies for acute and chronic low back pain?
This section focuses on studies that compared dual therapy with two non-invasive interventions to monotherapy with one of the interventions. Most of the systematic reviews and trials included in this section are described in more detail in the relevant sections of Key Questions 3 and 4. We did not include invasive interventions in this section because they are generally only considered after failure of non-invasive therapies.
Combinations of medications
Results of search: systematic reviews A Cochrane review included five trials (four higher-quality) on efficacy of skeletal muscle relaxants plus an NSAID or acetaminophen versus an NSAID or acetaminophen alone488, 489. We found no other systematic reviews on efficacy of one drug added to another relative to one of the drugs alone. Results of search: trials We identified one additional lower-quality trial on efficacy of opioids plus an NSAID versus an NSAID alone517.
Efficacy of a muscle relaxant plus an analgesic versus an analgesic alone

For acute low back pain, the Cochrane review of muscle relaxants488, 489 included three higherquality trials449, 973, 974 that consistently found tizanidine plus acetaminophen or NSAIDs superior to placebo plus acetaminophen or NSAIDs for short-term (up to one week) pain relief and decrease of muscle spasm. Another higher-quality trial included in the Cochrane review found no differences in global efficacy between orphenadrine plus acetaminophen compared to placebo plus acetaminophen, but the combination was associated with significantly fewer disability days975. One lower-quality trial found no differences in pain intensity or global efficacy between cyclobenzaprine plus an NSAID versus an NSAID alone, though effects on muscle spasm were superior976.
Efficacy of an opioid plus an NSAID versus an NSAID alone

Naproxen was inferior to set-dose or titrated-dose opioid plus naproxen in one small n=36) lowerquality trial (Table 102)517. However, results are difficult to interpret because the naproxen dose was not specified and average doses not reported.

Table 102. Trials of an opioid versus an NSAID or acetaminophen

Author, year Jamison, 1998517
Main results Sustained-release morphine + immediate-release oxycodone (titrated dose) + naproxen versus immediaterelease oxycodone (set dose) + naproxen versus naproxen alone (mean scores over 16 weeks, all outcomes on 0 to 100 scales) Average pain: 54.9 vs. 59.8 vs. 65.5 Anxiety: 11.2 vs. 15.0 vs. 31.6 Depression: 10.8 vs. 16.4 vs. 26.9 Level of activity: 49.3 vs. 49.3 vs. 51.5 Hours of sleep (means): 5.9 vs. 5.9 vs. 6.1
Quality score 3/11
Harms The Cochrane review found a higher risk of central nervous system adverse effects with the combination of a muscle relaxant plus an analgesic (4 trials, RR=2.44, 95% CI 1.05 to 5.63)488, 489 . For overall adverse effects, there was no significant difference (RR=1.34, 95% CI 0.67 to 2.67). Costs We found no studies evaluating costs.
Summary of evidence
For acute low back pain, there is consistent evidence from three higher-quality trials that tizanidine combined with acetaminophen or an NSAID is associated with greater short-term pain relief and decrease of muscle spasm compared to acetaminophen or an NSAID alone (level of evidence: good). For acute low back pain, one higher-quality trial found no benefits from adding orphenadrine to acetaminophen, though the combination was associated with fewer disability days (level of evidence: fair). For acute low back pain, one lower-quality trial found no benefits from adding cyclobenzaprine to an NSAID (level of evidence: poor). There is insufficient evidence from one trial (doses unclear) to judge efficacy of opioids plus an NSAID versus an NSAID alone (level of evidence: poor). Adding a muscle relaxant to acetaminophen or an NSAID is associated with an increased risk of central nervous system adverse effects (level of evidence: good).

The AHCPR guidelines found no additional benefit from using muscle relaxants plus NSAIDs over using NSAIDs alone.

APS Clinical Guideline for the Evaluation and Management of Low Back Pain The European COST guidelines recommend adding a short course of muscle relaxants on its own or added to NSAIDs in patients with acute low back pain, if acetaminophen or NSAIDs failed to reduce pain.
Self-care advice combined with other interventions

Results of search: systematic reviews We found no relevant systematic reviews. Results of search: trials We identified five trials (three rated higher-quality368, 661, 977 that compared a self-care book plus another intervention to a self-care book alone363, 978.
Efficacy of a self-care book combined with other interventions

For low back pain of less than 6 weeks duration, one higher-quality trial found a self-care book plus advice and immediate exercise therapy using a biopsychosocial approach associated with more rapid improvements in function than a self-care book plus advice and waiting for 6 weeks to initiate exercise therapy (Table 103)977. For patients off work for less than one year due to low back pain, a lower-quality trial found addition of a brief exercise intervention to a self-care book and advice associated with quicker return to work (20 versus 13 days, p=0.034) and greater improvement in pain scores through two months compared to a self-care book and advice without the exercise intervention978.
Table 103. Trials of a self-care book + exercise versus a self-care book alone
Author, year Wright, 2005978]
Wand, 2004977
n=102 6 months
Main results Self-care book plus advice plus usual care vs. self-care book plus advice plus brief exercise therapy McGill Pain Questionnaire, VAS (0 to 100): 34.9 vs. 23.6 at 1 month (p=0.047), 30.9 vs. 18.4 at 2 months (p=0.023) McGill Pain Questionnaire, PPI (0 to 10): 1.75 vs. 1.13 at 1 month (p=0.039), 1.53 vs. 1.09 at 2 months (p=0.087) SF-12, physical subscale: 14.6 vs. 16.4 at 2 months (NS) SF-12, mental subscale: 20.8 vs. 22.1 at 2 months (NS) Return to work, median number of days: 20 vs. 13, (p=0.034) Self-care book + advice + immediate exercise therapy with biopsychosocial assessment vs. self-care book + advice + delayed therapy Roland Disability score (0 to 24), mean: 4.5 vs. 6.3 at 6 weeks (p=0.02), 3.9 vs. 4.4 at 6 months (p=0.94)Pain (0 to 10): 2.4 vs. 3.3 at 6 weeks (p=0.22), 2.1 vs. 2.4 at 6 months (p=0.61) SF-36 bodily pain (0 to 100): 65 vs. 54 at 6 weeks (p=0.06), 73 vs. 65 at 6 months (p=0.32) No differences on other SF-36 subscales at 3 or 6 months, though immediate therapy superior at 6 weeks on vitality, social functioning, and mental health
Quality score* 3/9
6/9

APS Clinical Guideline for the Evaluation and Management of Low Back Pain A higher-quality trial found addition of interferential therapy to a self-care book associated with greater improvement in functional status at three months compared to the self-care book alone, but baseline differences may invalidate results (Table 104)661.
Table 104. Trials of a self-care book + exercise versus a self-care book alone
Main results Interferential therapy applied to painful area + selfcare book versus interferential therapy applied to area of spinal nerve + self-care book versus self-care book alone (difference in median scores from baseline to 3 months) McGill Pain Questionnaire Pain Rating Index (0 to 78): +2.2 vs. -2.5 vs. -9.7 RDQ Score (0 to 24): -3.5 vs. -8.0 vs. -4.0 EQ-5D: No difference RDQ Score, median score at 3 months: 2.0 vs. 1.0 vs. 1.0
Quality score* 5/9
Two other trials evaluated efficacy of a self-care book plus face-to-face advice with a self-care book alone. One higher-quality trial found a brief nurse-led educational intervention plus a selfcare book associated with a higher proportion of patients exercising and greater patient satisfaction than a self-care book alone (Table 105)368. However, there were no differences in pain or functional status. A lower-quality trial found outcomes no better with the combination of a self-care book and advice to exercise compared to the self-care book alone363.
Table 105. Trials of a self-care book + another intervention versus a self-care book alone
Author, year Duration of LBP Little, 2001363 Acute or subacute (<3 months) Cherkin, 1996368 Not specified
n=300 1 year
Main results Self-care book vs. exercise advice vs. both vs. neither (control) (mean changes versus control) Pain/function scale (0 to 100): -8.7 vs. -7.9 vs. -0.1 at 1 week, 6.3 vs. -1.4 vs. -4.0 at 3 weeks (NS) Aberdeen pain and function scale (0 to 100): -3.8 vs. -5.3 vs. -1.9 at 1 week (NS) Self-care book vs. nurse education + self-care book vs. usual care (mean change from baseline) RDQ score (0 to 24 scale): -5.4 vs. -5.2 vs. -5.3 (NS) at 1 week Symptom bothersomeness score (0 to 10 scale): -3.3 vs. -3.3 vs. -3.6 (NS) at 1 week Health care visits for low back pain: 45% vs. 46% vs. 47% in first 7 weeks after intervention (NS) Work loss days: 24% vs. 36% vs. 29% in first 7 weeks after intervention (NS)
Quality score* 4/9
6/9

Summary of evidence
Two trials (one higher-quality) found a self-care book plus advice plus exercise therapy superior to the self-care book and advice alone. One trial evaluated patients with back pain for less than 6 weeks and the other evaluated patients off work less than one year due to back pain (level of evidence: fair). Two trials (one higher-quality) found the addition of face-to-face advice to a self-care book did not improve clinical outcomes, though one of the trials found self-reported exercise and patient satisfaction higher (level of evidence: fair). For subacute low back pain, one higher-quality trial compared interferential therapy plus a selfcare book to a self-care book alone and found the combined intervention improved functional status at 3 months, but differences could be due to baseline differences between groups (level of evidence: fair).

The other guidelines do not address self-care.
Exercise combined with other interventions

Results of search: systematic reviews We identified one higher-quality Cochrane review613, 614 and an associated meta-regression615 that evaluated efficacy of exercise therapy plus other non-invasive interventions relative to exercise therapy alone for chronic low back pain. Results of search: trials One recent, large, lower-quality trial (UK BEAM) not included in the Cochrane review evaluated effects of exercise plus spinal manipulation versus spinal manipulation alone for subacute or chronic low back pain629. A higher-quality trial not included in the Cochrane review evaluated exercise plus self-care advice versus exercise alone362.
Efficacy of exercise therapy plus other non-invasive treatments versus exercise alone
The meta-regression performed in conjunction with the Cochrane review analyzed 36 groups that received exercise plus another intervention and 36 groups that received exercise alone615. In multivariate analyses, adding other non-invasive interventions had a small average additional effect compared to exercise therapy alone of 5.1 points (95% CI 3.6 to 7.1) for pain and 2.1 points (95% CI 0.7 to 3.7) for function (each on 100 point scales). Results of the recent, large (n=1334) UK BEAM trial were consistent with these findings (Table 106)629. At 12 months, the combination of exercise and manipulation was associated with small net improvements in RDQ scores compared to manipulation alone (net improvement relative to usual care 1.30 points on a 0 to 24 scale, 95% CI 0.54 to 2.07 and 1.01 points, 95% CI 0.22 to

APS Clinical Guideline for the Evaluation and Management of Low Back Pain 1.81, respectively). The difference between combination therapy and manipulation was similar at three months. There were also no significant differences on the modified Von Korff scale or SF36, though results on the back beliefs and fear avoidance questionnaires favored combination therapy. A higher-quality trial found exercise plus advice to remain active slightly more effective than exercise alone for subacute (6 to 12 weeks) low back pain362. However, statistical significance was not reported for this comparison. Differences averaged about 0.3 points on a 0 to 10 pain scale and 0.6 points on the RDQ.
Table 106. Trials of exercise therapy plus other non-invasive interventions versus exercise therapy alone not included in published systematic reviews
Author, year Pengel, 2007362 (nonspecific low back pain)
UK BEAM Trial, 2004629
n=1334 12 months
Main results Exercise plus advice versus sham advice, sham ultrasound and sham diathermy (mean change reported for all results) Pain: -1.5 (95% CI -2.2 to -0.7) at 6 weeks, -0.8(95% CI -1.7 to +0.1) at 12 months Patient-specific functional scale: +1.1 (95% CI +0.3 to +1.9) at 6 weeks, +1.1 (95% CI +0.3 to +1.8) at 12 months Global perceived effect: +1.3 (95% CI +0.7 to +1.9) at 6 weeks, +0.8 (95% CI 0.0 to +1.6) at 12 months RDQ: -1.3 (95% CI -2.7 to +0.2) at 6 weeks, -0.9 (95% CI -2.7 to +0.8) at 12 months Depression Anxiety Stress Scale: +0.2 (95% CI -2.5 to +2.8) at 6 weeks, -0.4 (95% CI -3.1 to +2.3) at 12 months Exercise versus sham ultrasound plus sham diathermy (mean change reported for all results) Pain: -0.8 (95% CI -1.3 to -0.3) at 6 weeks, -0.5 (95% CI -1.1 to +0.2) at 12 months Patient-specific functional scale: +0.4 (95% CI -0.2 to +1.0) at 6 weeks, +0.5 (95% CI -0.1 to +1.0) at 12 months Global perceived effect: +0.5 (95% CI +0.1 to +1.0) at 6 weeks, +0.4 (95% CI -0.1 to +1.0) at 12 months RDQ: -0.8 (95% CI -1.8 to +0.3) at 6 weeks, -0.3 (95% CI -1.6 to +0.9) at 12 months Depression Anxiety Stress Scale: -0.7 (95% CI -2.5 to +1.2) at 6 weeks, -0.6 (95% CI -2.6 to +1.3) at 12 months Manipulation + exercise versus manipulation versus exercise versus usual care (all results are net benefit relative to usual care at 12 months) RDQ Questionnaire (0 to 24 scale): 1.30 (95% CI 0.54 to 2.07) vs. 1.01 (95% CI 0.22 to 1.81) vs. 0.39 (95% CI -0.41 to 1.19) Modified Von Korff pain score (0 to 100 scale): 6.71 (95% CI 2.47 to 10.95) vs. 5.87 (95% CI 1.58 to 10.17) vs. 4.90 (95% CI 0.30 to 9.50) Modified Von Korff disability score (0 to 100 scale): 6.71 (95% CI 2.62 to 10.80) vs. 5.65 (95% CI 1.57 to 9.72) vs. 4.56 (95% CI 0.34 to 8.78)
Quality score* 8/9
2/9

Summary of evidence
The addition of other non-invasive interventions to exercise was associated with small improvements in pain (about 5 points on a 100 point scale) and no clinically significant improvement in function (about 2 points on a 100 point scale) in a meta-regression of 36 comparison groups. Results of two additional trials not included in the meta-regression are consistent with these findings (level of evidence: good).

The other guidelines do not address exercise combined with other interventions.
Acupuncture combined with other non-invasive interventions

Results of search: systematic reviews We identified one higher-quality Cochrane review of acupuncture for low back pain that evaluated efficacy of acupuncture added to other non-invasive interventions for acute (one lowerquality trial) or chronic (four higher-quality trials) low back pain69, 70. Results of search: trials We did not search for additional trials
Efficacy of acupuncture plus other non-invasive treatments versus the other treatment alone
For acute low back pain, the Cochrane review included one lower-quality trial (n=100) that found the combination of acupuncture and moxibustion plus Chinese herbal medicine superior to Chinese herbal medicine alone for pain and function at long-term follow-up979. For chronic low back pain, the Cochrane review also included four higher-quality trials (n=289) that found addition of acupuncture to another intervention more effective than the other intervention alone (co-interventions included exercises, NSAIDs, aspirin, non-opioid analgesics, mud packs, infrared heat therapy, back care education, ergonomics, or behavioral modifications). In pooled analyses, the addition of acupuncture was associated with moderate improvements in pain (two trials, SMD -0.76, 95% CI -1.14 to -0.38) and function (three trials, SMD -0.55, 955 CI -0.92 to -0.18) that persisted through 3 to 12 months of follow-up69, 70. Despite the evaluation of different co-interventions, there was no between-study heterogeneity in the pooled analyses. Harms No trial reported adverse events.

Summary of evidence
For acute low back pain, there is insufficient evidence to judge effects of acupuncture added to other interventions (one lower-quality trial) (level of evidence: poor). For chronic low back pain, the addition of acupuncture to a variety of other non-invasive interventions was associated with consistent, moderate beneficial effects compared to the other intervention alone on pain and function through 3 to12 months (four higher-quality trials) (level of evidence: good).

The other guidelines do not address acupuncture combined with other non-invasive interventions.
Spinal manipulation combined with other interventions

Results of search: systematic reviews A higher-quality Cochrane review of spinal manipulation did not evaluate additive benefits of spinal manipulation to other non-invasive interventions66, 67. Results of search: trials One recent, large trial evaluated exercise therapy plus manipulation versus manipulation alone and exercise therapy alone629.
Efficacy of spinal manipulation plus exercise versus exercise alone

For acute or subacute low back pain, the UK BEAM trial found RDQ scores improved an average of 1.30 points (95% CI 0.54 to 2.07) with manipulation plus exercise compared to 0.39 (-0.41 to 1.19) with exercise alone (Table 106)629. The small difference in average effect (about one point) was not statistically significant. There were also no significant differences on other outcome measures including the Von Korff scale, the back beliefs questionnaire, the fear avoidance beliefs questionnaire, or SF-36. Another higher-quality trial found the combination of manipulation and exercise and a brief educational intervention (physician consultation) slightly superior to physician consultation alone for long-term pain scores (average 6.3 point difference on a 100 point pain scale at 12 months and 2.4 point difference after 24 months) compared to physician consultation alone in patients with chronic low back pain (Table 41)612. There were no differences on the ODI score or health-related quality of life.
Efficacy of chiropractic care plus physical modalities versus chiropractic care alone
The higher-quality UCLA Low Back Pain Study found chiropractic care plus physical modalities (heat or cold, ultrasound, or electrical muscular stimulation) no better than chiropractic care alone for pain or functional status (Table 107)780, 781.

Table 107. Results of the UCLA Low Back Pain Study

Author, year Hurwitz, 2002780, 781 UCLA Low Back Pain Study
Main results Chiropractic care + physical modalities vs. chiropractic care alone Most severe pain (0 to 10 scale): -0.15 (95% CI -0.85 to 0.55) at 6 months, +0.25 (-0.49 to 0.98) at 18 months Average pain (0 to 10 scale): -0.26 (95% CI -0.81 to 0.29) at 6 months, +0.12 (-0.46 to 0.71) at 18 months RDQ score (0 to 24 scale): +0.12 (95% CI -1.15 to +1.38) at 6 months, -0.01 (95% CI -1.35 to +1.32) at 18 months
Quality score* 7/9
Harms No trial reported adverse events. Costs The UK BEAM Trial estimated a cost-effectiveness ratio of 3800/QALY (about $7,448 U.S./QALY) for manipulation plus exercise relative to best care alone629. The cost-effectiveness of the combined treatment was superior to either manipulation or exercise alone (4,800/QALY or about $9,408 U.S./QALY and 8,300/QALY or about $16,268 U.S./QALY respectively, each relative to best care alone). Although the UCLA Low Back Pain Study found the addition of physical modalities to chiropractic care associated with negligible additional average cost ($579 vs. $560), there were also no differences in outcomes787.
Summary of evidence
For subacute or chronic low back pain, spinal manipulation plus exercise was not associated with significant benefits compared to exercise alone in a recent, large, lower-quality trial (level of evidence: poor). For back pain of unspecified duration, adding physical modalities to chiropractic care did not improve outcomes compared to chiropractic care alone in one higher-quality trial (level of evidence: fair). For chronic low back pain, the combination of spinal manipulation plus exercise and a brief educational intervention (physician consultation) was slightly superior for long-term pain but not function compared to physician consultation alone in one higher-quality trial (level of evidence: fair).

The other guidelines do not address spinal intervention combined with other interventions.

Massage combined with other interventions

Results of search: systematic reviews We identified one recent, higher-quality Cochrane review of massage therapy700, 701 that included one higher-quality trial736 of massage plus exercise and education versus exercise and education without massage. Results of search: trials We did not search for additional trials.
Efficacy of massage plus exercise and education versus exercise and education without massage
For subacute low back pain, one trial found combined treatment with massage, exercise and education moderately superior to exercise and education without massage for pain (McGill Present Pain Intensity) and disability (RDQ score) at one-month follow-up736. Mean Present Pain Intensity scores (0 to 5 scale) were 0.42 (95% CI 0.17 to 0.66) for the combination versus 1.33 (0.97 to 1.7) for exercise and education alone, and mean RDQ scores (0 to 24 scale) 1.54 (95% CI 0.69 to 2.4) versus 5.71 (95% CI 3.5 to 7.9). Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
For subacute low back pain, the addition of massage therapy to exercise and education was moderately superior to exercise and education alone for short-term pain and disability in one higher-quality trial (level of evidence: fair).

The other guidelines do not address massage combined with other interventions.
Psychological therapies combined with other interventions

Results of search: systematic reviews We identified one recent, higher-quality Cochrane review that included six lower-quality trials comparing psychological therapies in addition to another treatment versus the other treatment alone for chronic low back pain301. We considered combination therapy including psychological therapy separately from interdisciplinary rehabilitation, which we defined as a more integrated (and usually more intensive) intervention that often involves three or more different interventions (see Key Question 4). Results of search: trials We did not search for additional trials

Efficacy of psychological therapies in addition to another intervention versus the other intervention alone
The Cochrane review included six trials that compared psychological therapies combined with exercise and back education, multidisciplinary treatment, inpatient pain management, various forms of medical treatment (pain medication, nerve blocks, or physical therapy), and exercise therapy301. In pooled analyses, adding psychological therapies to other interventions was not associated with beneficial effects on long-term pain intensity (SMD=-0.24, 95% CI -0.64 to 0.16), functional status (SMD=0.26, 95% CI -0.06 to 0.57), or behavioral outcomes (SMD=0.32, 95% CI -0.06 to 0.71). Despite the evaluation of different co-interventions, little between-study heterogeneity was present. Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
For chronic low back pain, psychological therapies did not improve outcomes when added to a variety of other interventions in six lower-quality trials. Diversity in both the psychological and non-psychological interventions may limit generalizability of these findings (level of evidence: fair).

The other guidelines do not address psychological therapy combined with other interventions.
Traction combined with other interventions

Results of search: systematic reviews We identified a recent, higher-quality Cochrane review676, 677 that included one lower-quality trial980 comparing traction plus physical therapy to physical therapy alone. Results of search: trials We did not search for additional trials
Efficacy of traction plus physical therapy versus physical therapy alone

For chronic low back pain with or without sciatica, one small (n=42) trial included in the Cochrane review found no statistically significant differences between traction plus physical therapy (exercise, hot packs, and ultrasound) and physical therapy alone for pain, functional status, global recovery, or satisfaction980. Harms No trial reported adverse events.

Summary of evidence
For chronic low back pain with or without sciatica, traction plus physical therapy was no better than physical therapy alone in one small, lower-quality trial (level of evidence: poor).

The other guidelines do not address traction combined with other interventions.
Combination therapy for spinal stenosis

Results of search: systematic reviews We found no relevant systematic reviews. Results of search: trials We identified one higher-quality trial on efficacy of two different physical therapy-based interventions for spinal stenosis630.
Efficacy of combined physical therapy interventions for spinal stenosis

For chronic spinal stenosis, one higher-quality trial found the combination of manual therapy (manipulation and mobilization), tailored exercises, and body-weight supported treadmill ambulation associated with a higher likelihood of perceived recovery compared to lumbar flexion exercises, standard treadmill walking, and subtherapeutic ultrasound (79% vs. 41% at 6 weeks, 38% vs. 21% at 29 months), though there were no differences on other outcomes including pain scores and the ODI630.
Table 108. Trial of physical therapy-based interventions for spinal stenosis
Number of patients Duration of follow-up n=60 Mean 29 months
Author, year Whitman, 2006630
Main results Manual therapy, tailored exercises, and body-weight supported treadmill ambulation program vs. lumbar flexion exercises, standard treadmill walking, and subtherapeutic ultrasound Perceived recovery (global rating of change +3 or higher): 79% vs. 41% at 6 weeks, 62% vs. 41% at 1 year, 38% vs. 21% at mean 29 months ODI, between group differences (positive values favor manipulation/mobilization group): 3.93 (95% CI -2.07 to 9.93) at 6 weeks, 2.10 (95% CI -8.50 to 4.32) at 1 year Spinal Stenosis Scale Satisfaction Subscale (1 to 4), between group differences: 0.26 (95% CI -0.09 to 0.62) at 1 year Numeric Pain Rating Scale for lower extremity symptoms (0 to 10), between group differences: 0.47 (95% CI -1.23 to 2.18) at 1 year Treadmill walking distance: No differences
Quality score* 8/9

Summary of evidence
For chronic spinal stenosis, one higher-quality trial found manual therapy, tailored exercises, and body-weight supported treadmill ambulation moderately superior for perceived recovery compared to standardized lumbar flexion exercises, standard treadmill walking, and subtherapeutic ultrasound through two years, but found no differences on other outcomes (level of evidence: fair).

The other guidelines do not address combination therapy for spinal stenosis.
Key Question 11 How effective are different treatment strategies for failed back surgery syndrome?
The term failed back surgery syndrome is commonly used to refer to a heterogeneous group of conditions characterized by chronic disabling low back pain with or without leg pain following one or more spinal surgeries. Because success rates of a second spinal operation are substantially lower than with initial surgery (and continue to decline with subsequent operations)981, effective non-surgical treatment alternatives have been sought for patients with failed back surgery syndrome.
Adhesiolysis and forceful epidural injection

Adhesiolysis (also referred to as lysis of epidural adhesions, epidural neurolysis, and epidural neuroplasty) is a relatively new procedure. The goal of adhesiolysis is to facilitate application of drugs to target nerves and other tissues by removing scars and adhesions in the epidural space. Adhesiolysis can be performed percutaneously or with endoscopic guidance and involves the injection of isotonic saline, hypertonic saline, or hyaluronidase (intended to facilitate lysis of adhesions). It is typically reserved for patients with back pain refractory to other treatments, often in the post-surgical setting. The purpose of forceful epidural injections is also to disrupt epidural adhesions or fibrosis following lumbar surgery by injecting large volumes (20 mls or more) of saline, with or without a corticosteroid. In published trials of forceful epidural injections, the epidural space has been accessed via the sacroccygeal hiatus, which may help avoid areas of fibrosis that can impede placement of medications using other approaches.

APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: systematic reviews We identified one lower-quality systematic review on efficacy of adhesiolysis982. We excluded an earlier version of this review983. We also included one lower-quality systematic review of endoscopic division of epidural adhesions984. Results of search: trials The systematic review of adhesiolysis included four randomized trials115, 985-987. However, one trial classified as a randomized trial was clearly not randomized (patients were allocated to the no adhesiolysis group if the insurer or patient refused the treatment)986. In addition, the systematic review did not report quality ratings for included trials. We independently abstracted and analyzed the three randomized trials included in the systematic review115, 985, 987. One was rated higher-quality115. It compared adhesiolysis to epidural steroid injection without adhesiolysis. Another trial compared adhesiolysis to a poorly defined physical therapy intervention987. The third trial compared different adhesiolysis methods (hypertonic versus normal saline, with or without hyaluronidase)985. We also identified two lowerquality trials evaluating forceful epidural injections for persistent post-operative back pain or sciatica848, 850. The systematic review of endoscopic division of epidural adhesions identified no trials that met inclusion criteria984. From 472 potentially relevant citations, we identified one higher-quality trial on efficacy of targeted steroid placement with epidural endoscopy with adhesiolysis if adhesions were observed at the target nerve, versus caudal epidural steroid without endoscopy843.
Efficacy of adhesiolysis with or without hypertonic saline versus other interventions

The systematic review of adhesiolysis had important methodological limitations, including misclassification of a non-randomized trial as randomized986, failure to report quality ratings for included trials, and classification of trials as positive (showing efficacy for adhesiolysis) if patients improved compared to baseline, regardless of whether any differences were observed versus a control group. We independently rated the quality of adhesiolysis trials (Table 109). For chronic low back pain that failed to respond to standard treatments (including epidural steroids) and had negative facet joint block testing (about 70% with previous back surgery), one higher quality trial (n=75) found adhesiolysis with or without hypertonic saline associated with significantly greater likelihood for >50% pain relief compared to epidural steroid alone (72% and 60% vs. 0%, p<0.001) after 12 months115. However, even though patients enrolled in this trial had failed a previous epidural injection, the 0% response rate with epidural steroids is much lower than in other trials. For example, in a higher-quality trial of epidural steroids versus saline placebo, rates of improvement in pain were approximately 70% in both groups130. For chronic low back pain with sciatica (about 15% with previous back surgery), a second, lowerquality randomized trial found adhesiolysis substantially superior to physical therapy for leg and

APS Clinical Guideline for the Evaluation and Management of Low Back Pain back pain and moderately superior on the ODI at 6 months, though differences were no longer significant at 6 or 12 months987. The physical therapy intervention was not described in this trial, there was high loss to follow-up in the physical therapy arm, and intention-to-treat results were not reported. The third study reported itself as a randomized trial but was actually a nonrandomized comparative study986. Adhesiolysis was superior to usual care on most measured outcomes including pain, measures of functional status, and opioid intake.
Table 109. Studies of adhesiolysis versus other interventions
Author, year Veihelmann, 2006987 (randomized controlled trial)
Manchikanti, 2004115 (randomized controlled trial)
n=75 12 months
Manchikanti, 2001986 (non-randomized comparative study)
n=45 18 months to 3 years
Main results Adhesiolysis vs. physiotherapy (improvement from baseline) VAS leg pain (0 to 10): -4.8 vs. -1.1 at 3 months (p<0.05), -4.4 vs. -0.8 (NS) VAS back pain (0 to 10): -4.7 vs. -0.6 at 3 months (p<0.05), -4.2 vs. -0.3 (NS) ODI (0 to 100):-12.5 vs. +0.2 at 3 months (p<0.05), -11.5 vs. +0.2 (NS) Adhesiolysis with hypertonic saline vs. adhesiolysis with isotonic saline vs. epidural steroid Proportion with >50% pain relief at 12 months: 72% vs. 60% vs. 0% (p<0.001) ODI disability index score at 12 months: 23 vs. 24 vs. 32 (p<0.001) VAS pain score (0 to 10) at 12 months: 4.6 vs. 5.2 vs. 7.7 Taking opioids: 52% vs. 16% vs. 16% vs. 52% (p<0.001) Adhesiolysis vs. usual care Average pain (0 to 10): 3.9 vs. 6.9 (p<0.06) Functional status (0 to 10): 5.3 vs. 4.3 (p<0.05) Opioid intake moderate or heavy: 74% vs. 80% Employed: 17% vs. 20%
Quality 2/9*
8/11
3/11
Efficacy of adhesiolysis with hypertonic saline versus hyaluronidase versus isotonic saline
One higher-quality trial found no significant differences in pain relief between patients randomized to adhesiolysis with hypertonic saline compared to adhesiolysis with isotonic saline (Table 110)115. One lower-quality trial found no significant differences between patients who underwent adhesiolysis with hypertonic saline alone versus adhesiolysis with hyaluronidase, hyaluronidase alone, or isotonic saline for pain relief or in the proportion of patients requiring additional treatments (Table 110)985.

Table 110. Trials of adhesiolysis with hypertonic saline versus hyaluronidase versus isotonic saline
Author, year Manchikanti, 2004115 (randomized controlled trial)
Heavner, 1999985 (randomized controlled trial)
n=83 12 months
Main results Adhesiolysis with hypertonic saline vs. adhesiolysis with isotonic saline vs. epidural steroid Proportion with >50% pain relief at 12 months: 72% vs. 60% vs. 0% (p<0.001) ODI disability index score at 12 months: 23 vs. 24 vs. 32 (p<0.001) VAS pain score (0 to 10) at 12 months: 4.6 vs. 5.2 vs. 7.7 Taking opioids: 52% vs. 16% vs. 16% vs. 52% (p<0.001) Adhesiolysis with hypertonic saline vs. hypertonic saline + hyaluronidase vs. isotonic saline vs. isotonic saline + hyaluronidase No significant differences on McGill Questionnaire, VAS pain score, and percentage requiring additional treatments through 1 year (data only reported in graphs, raw data not provided)
Quality 8/11
2/11
Efficacy of forceful epidural injection

For persistent post-operative low back pain with sciatica and epidural fibrosis by CT, one lowerquality trial evaluated efficacy of repeated (two injections in first 48 hours, then once a month for four months) forceful epidural injection via the sacrococcygeal hiatus with a total of 40 ml of saline plus 125 mg of prednisolone (5 ml) versus standard epidural injection with 125 mg of prednisone alone (Table 111)850. It found greater rates of treatment success at 6 months with forceful epidural injection with corticosteroid for sciatica (45% vs. 19%, p=0.025) and for low back pain (24% vs. 6%, p=0.002), though differences were only significant for low back pain at 18 months (31% vs. 19%, p<0.05). Forceful epidural steroid injection was slightly superior on low back pain scores (average difference about 6 points on a 100 point scale) but not on sciatica pain scores. There were no differences in functional status or work status. For persistent post-discectomy sciatica, a second lower-quality trial found repeated (once a month for three months) forceful epidural injection with 20 ml of saline (with or without 125 mg prednisolone) inferior to standard epidural injection with 125 mg prednisolone alone on VAS pain scores through the first thirty days (p=0.01)848. However, there were no significant differences between the three groups one month after the third injection. There was also no difference in the proportion of patients achieving >15% improvement in pain score (p=0.30).

Table 111. Trials of forceful epidural steroid injection

Author, year Meadeb, 2001848 (randomized controlled trial)
Revel, 1996850 (randomized controlled trial)
n=60 18 months
Main results Forceful saline + corticosteroid vs. forceful saline alone vs. epidural corticosteroid alone (all via sacrococcygeal hiatus) Improvement in pain (0 to 100 VAS): -1.9 vs. -10.7 vs. -10.1 (p=0.08 for B vs. A or C) Proportion with >15% improvement in pain score: 20% vs. 44% vs. 25% (p=0.30 for B vs. A or C) Dallas activity of daily living score: No differences Forceful epidural saline + corticosteroid injection vs. epidural corticosteroids alone (via sacrococcygeal hiatus) Treatment 'success' for sciatica (improved or considerably improved on global efficacy evaluation): 45% vs. 19% at 6 months (p=0.025) and 34% vs. 29% at mean 18 months (p>0.05) Treatment 'success' for low back pain: 24% vs. 6% at 6 months (p=0.002) and 31% vs. 19% at 18 months (p<0.05) Change in low back pain score (0 to 100): 7 vs. 1 (p=0.015) Change in sciatica pain score (0 to 100): 3 vs. -10 (p=0.08) No differences in medication use, functional index, return to work or return to leisure activities
Quality score 3/11
5/11
Efficacy of endoscopic lysis of epidural adhesions

A systematic review found no trials of endoscopic lysis of epidural adhesions984. The only nonrandomized comparative study evaluated endoscopic lysis of epidural adhesions in pregnant women not necessarily with low back pain. For patients with persistent (>6 months) sciatica without previous spinal surgery, one trial not included in the systematic review found no differences in pain, anxiety, or depression scores between targeted steroid placement during spinal endoscopy with adhesiolysis performed if scar tissue was observed, versus caudal epidural steroid injection843. However, only 3 of 27 patients randomized to spinal endoscopy required endoscopic lysis of epidural adhesions. Harms One higher-quality trial of adhesiolysis reported one subarachnoid block among 50 patients115. One lower-quality trial reported no adverse effects among 59 patients undergoing adhesiolysis985, but another lower-quality trial reported transient sensory deficits in about onethird of patients and catheter problems in about 9%987. A non-randomized comparative study reported one suspected infection and minor complications (such as rash and itching) in 10% of patients986. In other observational studies, subarachnoid puncture was reported in up to 9% of procedures988, suspected infection in up to 10%988, and post dural headache in 14%989. Two trials both found greater rates of pain (73% vs. 52%)848 or drop-outs due to pain (14% vs. 3.2%)850 associated with forceful epidural injection compared to standard epidural injection (Table 112).

Table 112. Trials of forceful epidural injection vs. standard epidural injection
Author, year Meadeb, 2001848
Revel, 1996850
n=60 18 months
Main results Forceful saline + corticosteroid vs. forceful saline alone vs. epidural corticosteroid alone (all via sacrococcygeal hiatus) Improvement in pain (0 to 100 VAS): -1.9 vs. -10.7 vs. -10.1 (p=0.08 for B vs. A or C) Proportion with >15% improvement in pain score: 20% vs. 44% vs. 25% (p=0.30 for B vs. A or C) Dallas activity of daily living score: No differences Forceful saline + corticosteroid vs. epidural corticosteroids alone (via sacrococcygeal hiatus) Treatment 'success' for sciatica (improved or considerably improved on global efficacy evaluation): 45% vs. 19% at 6 months (p=0.025) and 34% vs. 29% at mean 18 months (p>0.05) Treatment 'success' for low back pain: 24% vs. 6% at 6 months (p=0.002) and 31% vs. 19% at 18 months (p<0.05) Change in low back pain score (0 to 100): 7 vs. 1 (p=0.015) Change in sciatica pain score (0 to 100): 3 vs. -10 (p=0.08) No differences in medication use, functional index, return to work or return to leisure activities
Quality score 3/11
5/11
Costs We found no studies evaluating costs.
Summary of evidence
For primarily post-surgical patients with refractory back pain who failed a previous epidural steroid injection, one small, higher-quality trial found adhesiolysis markedly superior to epidural steroid injection for pain relief. However, confirmation of results by other trials is necessary because of the extremely low (0%) response rate in the epidural steroid group (level of evidence: poor). For patients with chronic back pain and sciatica (with or without prior surgery), one lowerquality trial found adhesiolysis substantially superior to physical therapy for pain and functional status at 3 months, but the physical therapy intervention was not described, loss to follow-up was high in the physical therapy arm, differences were no longer significant after 6 months, and intention-to-treat results were not reported (level of evidence: poor). There is no clear evidence that adhesiolysis with hypertonic saline or hyaluronidase improves outcomes compared to adhesiolysis with isotonic saline alone (level of evidence: fair). Adverse events other than transient sensory deficits were infrequent and usually minor in trials of adhesiolysis, but were more common in observational studies and included suspected infection, subarachnoid puncture, and post-dural headache in up to 9-14% of patients (level of evidence: fair). For persistent post-surgical sciatica, one of two lower-quality trials found forceful epidural steroid injection superior to non-forceful epidural steroid injection (level of evidence: poor).

APS Clinical Guideline for the Evaluation and Management of Low Back Pain For persistent sciatica without previous spinal surgery, one small, higher-quality trial found no difference between targeted placement of epidural steroids during spinal endoscopy compared to caudal epidural steroid injection without endoscopy, but few patients randomized to endoscopy required lysis of epidural adhesions (level of evidence: poor).

The other guidelines do not address adhesiolysis and forceful epidural injections.
Intrathecal therapy
Intrathecal therapy involves the delivery of pain medication (usually an opioid) via a catheter directly into the intrathecal space. It is reserved for patients who do not respond to less invasive interventions990, including patients with failed back surgery syndrome. Before placing patients on long-term intrathecal therapy, a short-term trial is performed to determine responsiveness of pain and tolerability to treatment. For long-term, continuous use, intrathecal therapy is usually delivered using an implanted pump. Ziconotide is a non-opioid analgesic that acts by blocking neuron-specific calcium channels. It is the synthetic equivalent of venom from the marine snail Conus magus. It approved in December 2004 by the FDA for treatment of chronic severe pain in patients in whom intrathecal therapy is indicated. Results of search: systematic reviews We found no systematic reviews evaluating efficacy of intrathecal delivery of opioids or other drugs in patients with low back pain. Results of search: trials From 207 potentially relevant citations, we found no randomized trials meeting inclusion criteria. Two recent trials of ziconotide did not include991 or did not report results separately for patients with low back pain992.
Efficacy of intrathecal therapy

For failed back surgery syndrome, the only comparative observational study (n=67) found implantation with an intrathecal pump and administration of opioids associated with improvement in the ODI in 27% of patients after 5 years, compared to 12% of patients receiving usual care993. For chronic low back pain, one prospective study (n=136, 76% with prior back surgery) found pain scores had dropped by more than 47% at 12-month follow-up994. In addition, more than 65% of implanted patients had improvements in ODI scores. Other data on efficacy of intrathecal opioid therapy primarily comes from small case series of patients with cancer or noncancer pain, with the proportion of patients with good or excellent results ranging from 50% to close to 100%995. A recent trial of ziconotide for refractory pain due to various conditions found greater short-term (three-week) improvements in pain scores with ziconotide (14.7% improvement) compared to

APS Clinical Guideline for the Evaluation and Management of Low Back Pain placebo (7.2% improvement, p=0.036)990. Results specifically in patients with chronic low back pain were not reported. Harms Complications following intrathecal pump implantation are common. In one study, there was an average of 0.77 mean complications per implant (n=23)993. The most common complication was catheter-related and occurred in 26% (6/23) of patients. Other complications include pump flipping (22%) and infection (22%). One patient required pump explanation, and another developed late-onset meningitis after catheter replacement. In another study, adverse events occurred in 23 of 136 (17%) patients after intrathecal pump implantation, with 21 (15%) requiring surgical correction994. Adverse events included infection (12%), dislodgement or migration (1.5%), and cerebrospinal fluid leak (0.7%). Recently, a number of inflammatory masses of the catheter tip (granulomas) in patients receiving intrathecal opioids have also been reported990. These masses appear associated with high doses of certain opioids (such as hydromorphone), and can become large enough to cause neurological injury. Ziconotide was associated with significantly greater incidences of dizziness (47% vs. 13%), confusion (18% vs. 5%), ataxia (16% vs. 1.9%), abnormal gait (15% vs. 1.9%), and memory impairment (11.6% vs. 0.9%) compared to placebo in a recent trial of patients with chronic pain due to a variety of underlying conditions990. A trial of patients with refractory cancer of AIDsrelated pain also reported increased fever, postural hypotension, nausea and vomiting, somnolence, and urinary retention with ziconotide compared to placebo991. Most adverse events with ziconotide resolve upon discontinuation of the medication. Costs We identified two cost studies996, 997. Both estimated fewer costs with intrathecal morphine relative to medical management, but used poor-quality observational data for key parameters.
Summary of evidence
In patients with failed back surgery syndrome, there is insufficient evidence to judge efficacy of intrathecal opioid therapy (data from generally lower-quality observational studies only) (level of evidence: poor). Adverse events with intrathecal opioid therapy appear frequently and often require surgery (level of evidence: poor). There is insufficient data to judge efficacy of intrathecal ziconotide for low back pain (no trials). Intrathecal ziconotide is associated with a number of side effects including ataxia, dizziness, somnolence, confusion, nausea and vomiting, postural hypotension, and urinary retention (level of evidence: fair).

The other guidelines do not address intrathecal therapy.

Non-invasive interventions
Results of search: systematic reviews We found no systematic reviews of non-invasive interventions for failed back surgery syndrome. Results of search: trials From 472 potentially relevant citations, we identified one lower-quality trial that compared lowtech exercise, high-tech exercise, physical agents, manipulation, and no treatment for chronic low back pain following L5 laminectomy998.
Efficacy of non-invasive interventions for failed back surgery syndrome

The trial (n=250) found no significant differences in ODI scores at the end of an 8-week course of treatment of high-tech exercise (using specialized exercise equipment), low-tech exercise (using McKenzie and spinal stabilization training exercises), physical agents (hot packs, ultrasound, TENS), joint manipulation, or control, though trends favored the two exercise groups (Table 113)998.
Table 113. Trial of efficacy of non-invasive interventions for failed back surgery syndrome
Number of patients Duration of follow-up n=250 At end of 8 week course of treatment Quality score* 2/9
Author, year Timm, 1994998
Main results Low-tech exercise vs. high-tech exercise vs. physical agents vs. manipulation vs. no treatment (at end of 8 week treatment session) ODI (0 to 100), mean improvement: -20.5 vs. -18.1 vs. -0.14 vs. -3.8 vs. -0.18
One comparative observational study found interdisciplinary rehabilitation to be moderately more effective at reducing self-reported pain and slightly more effective at improving ODI scores in patients with failed back surgery syndrome than in patients with chronic low back pain without previous back surgery999. Harms No trial reported adverse events. Costs We found no studies evaluating costs.
Summary of evidence
For chronic low back pain following L5 laminectomy, one lower-quality trial found no significant differences in immediate post-treatment ODI scores between exercise, physical agents, manipulation, and no treatment (level of evidence: poor). Interdisciplinary rehabilitation was moderately more effective for reducing pain and slightly more effective for improving functional status in patients with failed back surgery syndrome compared to those without previous surgery in one comparative observational study (level of evidence: poor).


The other guidelines do not address non-invasive interventions.
Spinal cord stimulation

Results of search: systematic reviews We identified three systematic reviews (all rated higher-quality) on efficacy of spinal cord stimulation for failed back surgery syndrome85, 95, 96, 98. We excluded one outdated systematic review192 and one review that did not use systematic methods195. Results of search: trials Two randomized trials evaluated spinal cord stimulation for failed back surgery syndrome113, 153. Interim or final results of one trial153 were included in three higher-quality systematic reviews85, 95, 96, 98 and we identified one additional trial113. The systematic reviews also included one lower-quality, controlled observational study1000. Seventy-two other case series of spinal cord stimulation for chronic back and leg pain or failed back surgery syndrome were included in one of the reviews95 and are reviewed in Key Question 9.
Efficacy of spinal cord stimulation

For failed back surgery syndrome with persistent radiculopathy, one higher-quality trial (n=50) found spinal cord stimulation associated with a greater likelihood for >50% pain relief compared to re-operation after a mean of 2.9 years (38% or 9/24 vs. 12% or 3/26, p=0.048) (Table 114)153. Spinal cord stimulation was also associated with a lower rate of increased use of opioids (13% vs. 42%), and fewer patients allocated to spinal cord stimulation subsequently underwent surgery (21% or 5/24) compared to those allocated to surgery who later received spinal cord stimulation (54% or 14/26). Three-year results were similar. A second, higher-quality trial (n=100) of patients with persistent radicular pain following anatomically successful surgery for herniated disc found spinal cord stimulation associated with greater likelihood of experiencing >50% pain relief after six months compared to conventional medical management (48% versus 9%, p<0.001)113. Spinal cord stimulation was also moderately superior (by 10 to 20 points) on 7 of 8 SF-36 subscales and the ODI. The trial was designed so that patients randomized to spinal cord stimulation would undergo device implantation only if they experienced greater than 50% pain relief or 80% paresthesia coverage following a screening trial. 92% (48/52) of patients randomized to spinal cord stimulation underwent implantation, including five patients who did not meet criteria for a positive trial.

Table 114. Randomized trials of spinal cord stimulation for failed back surgery syndrome
Sample size Type of LBP Duration of symptoms n=100 Failed back surgery syndrome with persistent radiculopathy Chronic
Author, year Kumar, 2007113
North, 2005153
n=60 Failed back surgery syndrome with persistent radiculopathy Chronic
2 years
Main results Spinal cord stimulation vs. conventional medical management 50% pain relief at 6 months: 48% (24/50) vs. 9% (4/44) (p<0.001) >50% pain relief at 6 months: 34% vs. 7% (crossover at 6 months considered failures) (p=0.005) SF-36 Physical function: 38.1 vs. 21.8 (p<0.001) SF-36 Role physical: 17.5 vs. 8.0 (p=0.12) SF-36 Bodily pain: 33.0 vs. 19.5 (p<0.001) SF-36 General health: 52.8 vs. 41.3 (p<0.001) SF-36 Vitality: 41.3 vs. 31.1 (p=0.01) SF-36 Social functioning: 49.3 vs. 33.5 (p=0.002) SF-36 Role-emotional: 51.3 vs. 29.5 (p=0.02) SF-36 Mental health: 62.6 vs. 50.1 (p=0.002) ODI: 44.9 vs. 56.1 (p<0.001) Opioids use: 56% vs. 70% (p=0.20) Satisfied with pain relief: 66% vs. 18% (p<0.001) Return to work: 11% vs. 3% (p=0.36) Spinal cord stimulation vs. recurrent lumbosacral spine surgery >50% pain relief and satisfied with treatment: 38% (9/24) vs. 12% (3/26) (p=0.04) Crossed over: 21% (5/24) vs. 54% (14/26) (p=0.02) Opioid use stable or decreased: 87% (20/23) vs. 58% (15/26) (p=NS) Opioid use increased: 13% (3/23/) vs. 42% (11/26) Activities of daily living, neurologic status, ability to work: Differences not significant
Quality score 6/11
6/11
Harms In the randomized trials, 26% to 32% of patients experienced a complication following spinal cord stimulator implantation, including electrode migration, infection or wound breakdown, generator pocket-related complications, and lead problems113, 153. Long-term complications included one infection, two implantation generator pocket-related complications, and one defective lead. Evidence on harms from non-randomized studies of spinal cord stimulation is discussed in Key Question 8. Costs One of the randomized trials153 that evaluated clinical outcomes also collected economic data1001. It estimated a 78% likelihood that the additional cost of spinal cord stimulation is less than $40,000/QALY compared to repeat surgery, based on data from 40 of the 50 patients originally enrolled in this trial. A decision analysis found that spinal cord stimulation dominated continued medical management over the lifetime of a patient with failed back surgery syndrome891. This study is difficult to interpret because it used potentially unreliable cost data from an observational study993. In addition, it assumed that the rate of pain relief in the nonspinal cord stimulator group could be estimated from a trial of patients with non-radicular low back pain randomized to surgery versus nonsurgical therapy247, even though 80% of enrollees in that trial had never

APS Clinical Guideline for the Evaluation and Management of Low Back Pain previously undergone back surgery. No sensitivity analyses were performed on key costs, utilities, or rates of pain relief.
Summary of evidence
For failed back surgery syndrome, one small, higher-quality trial found spinal cord stimulation associated with a higher likelihood of pain relief, lower likelihood of increase in opioid use, and lower likelihood of crossing over to reoperation (versus crossing over to spinal cord stimulation) compared to initial reoperation through 3 years and one small, higher-quality trial found spinal cord stimulation associated with moderately superior pain and functional outcomes compared to conventional medical management through 6 months (level of evidence: fair). About one-quarter of patients experience complications that usually not serious following spinal cord stimulator implantation. Most complications are related to infection and generator or leadassociated problems (level of evidence: fair).

The European COST guidelines found insufficient evidence to recommend spinal cord stimulation for patients with chronic low back pain.
Key Question 12 How effective are different methods of integrating or coordinating low back pain care?
Integration or coordination of care usually refers to a broad intervention that aims to help meet patient health care needs by enhancing information sharing across providers; encouraging use of evidence-based testing and interventions; insuring appropriate follow-up of referrals, testing, and interventions; and promoting goal-setting and patient self-management. Although interdisciplinary rehabilitation (see Key Question 4) may be considered a type of coordinated care intervention, it does not necessarily address the same broad framework as a formal coordination of care intervention. Results of search: systematic reviews We identified no systematic reviews on efficacy of different methods for integrating or coordinating care in patients with low back pain. Results of search: trials From 79 potentially relevant citations, we identified one lower-quality trial on efficacy of coordination of care relative to usual care in patients with back-pain associated disability1002. One other lower-quality trial evaluated efficacy of integrated care between primary care and neurology via a psychiatrist liaison versus usual care in patients with back pain of unspecified duration1003.
Efficacy of coordinated or integrated care versus usual care

For workers receiving compensation for low back pain for 4 to 8 weeks, one trial found coordination of primary health care slightly superior to usual care for functional status after 6

APS Clinical Guideline for the Evaluation and Management of Low Back Pain months on the ODI scale (average 9 point difference, p=0.02) and moderately superior on the Quebec Back Pain Disability Scale (average twelve point difference on a 100 point scale, p=0.01) (Table 115)1002. Coordination of care was also associated with slightly quicker return to work (6.6 days, not significant). Patients randomized to coordinated care used three times fewer specialized imaging tests (p<0.01) and exercised twice as much (p<0.05) as controls. Two primary care physicians and a nurse performed the coordination of care intervention, which involved a complete examination, recommendations to the treating physician for clinical management consistent with guidelines, and support to carry out the recommendations. For low back pain of unspecified duration, another lower-quality trial found integration of care between a neurologist and primary care physician via a psychiatrist did not improve patient outcomes, satisfaction of general practitioners, or affect utilization of healthcare services compared to usual care1003. The protocol called for the psychiatrist, who did not see the patient, to facilitate communication between the primary care physician and neurologist through structured telephone communication, weekly information sharing, and development of a treatment plan of care. However, the protocol was only fully implemented in about one-quarter of the 50 patients randomized to the intervention group.
Table 115. Trials of integration or coordination of care
Author, year Rossignol, 20001002
Meeuwesen, 19961003
n=104 6 months
Main results Coordination of care versus usual care Return to work by 6 months: 78% vs. 73% Time to return to work: average difference 6.6 days (NS) Pain, mean difference from baseline to 6 months: 22.9 vs. 12.8, p=0.1 Quebec Back Pain Disability Scale, mean difference from baseline to 6 months (0 to 100 scale): 20.9 vs. 9.1, p=0.01 ODI, mean difference from baseline to 6 months: 17.2 vs. 7.8, p=0.02 Dallas Pain Questionnaire, mean difference from baseline to 6 months: 25.9 vs. 11.7 (p=0.01) Coordination of care versus usual care SCL-90 subscales, DSM-III-R somatoform disorders (DSM-SOM) scale: No differences between interventions Functional impairment scale (FBI), mean difference from baseline to 6 months: 1.6 vs. 0.9 (NS) General Health Questionnaire-28, mean difference from baseline to 6 months: 2.0 vs. 1.7 (NS) Satisfaction of general practitioners: no differences between interventions Medication use: no differences between interventions Diagnostic imaging: no differences between interventions
Quality score* 4/9
2/9

Summary of evidence
In workers receiving short-term (4 to 8 weeks) compensation for low back pain, coordination of low back pain care was superior to usual care for improving functional status and pain after 6 months and reduced use of specialized imaging tests in one lower-quality trial (level of evidence: poor). There is insufficient evidence to judge efficacy of coordination or integration of low back pain care in other (primary care) settings (one low quality trial) (level of evidence: poor). Interdisciplinary rehabilitation is reviewed in Key Question 4.

The other guidelines do not address integrating or coordinating care in improving outcomes.
Key Question 13 How effective are interventions for secondary prevention of low back pain in patients who have had an episode of acute low back pain, or for prevention of flares of low back pain in patients with chronic low back pain? Back schools
Results of search: systematic reviews A recent, higher-quality Cochrane review of back schools (19 trials) included five trials (three higher-quality600, 604, 605) that reported recurrent low back pain episodes (or sick leave due to low back pain) as an outcome586, 587. Another recent, lower-quality systematic review of back schools did not include any additional trials1004. Results of search: trials We did not search for additional trials.
Efficacy of back schools versus no back school, usual care, or placebo for preventing recurrent episodes of low back pain
The Cochrane review included five trials that compared back schools to no treatment or usual care586, 587. Four trials were conducted in occupational settings and the fifth600 in a mixed setting. Longer-term follow-up603, 1005 is available from two higher-quality trials600, 604. For subacute low back pain, one trial found no difference between mini back school and usual care in the proportion of patients with one or more sick-leave recurrences randomized through five years of follow-up (72% or 142/198 versus 74% or 118/160), though the proportion with two or more recurrences was lower in the back school group (35% or 69/198 vs. 46% or 74/160)603, 604 . In patients no longer on sick leave, the other longer-term trial found that the mean number of low back pain recurrences decreased more with an intensive back school program than with no back school through three years (mean decrease 0.9 vs. 0.3 episodes/year, p<0.05)600, 1005. On

APS Clinical Guideline for the Evaluation and Management of Low Back Pain the other hand, three trials (one higher-quality) with shorter duration of follow-up (one year) reported no difference in low back pain recurrences with back school relative to usual care, placebo treatment, or wait-list control149, 598, 601. Two of the trials enrolled patients with back pain for less than three months, and the third601 enrolled patients with at least three episodes of low back pain annually.
Efficacy of back schools versus exercise for preventing recurrent episodes of low back pain
In workers with frequent (at least three annually) low back pain episodes, the Cochrane review included one lower-quality trial601 that found back school associated with a higher incidence of low back pain episodes than biweekly calisthenics through 12 months in workers with frequent (mean number of painful months 7.3 vs. 4.5, p<0.05).
Summary of evidence
Evidence on efficacy of back schools for preventing recurrent episodes of low back pain is mixed, which may be due in part to diversity among populations and interventions evaluated. One higher-quality trial found that an intensive back school intervention decreased recurrent episodes of low back pain more than no back school through three years of follow-up, but another trial that evaluated mini back school found no clear effect. Three shorter-term (1 year) trials (one higher-quality) also found no effect on recurrences (level of evidence: fair). One lower-quality trial found back school inferior to callisthenic exercises for reducing low back pain episodes through 12 months (level of evidence: poor).

The VA/DoD guidelines found inconclusive evidence on the long term benefit of back schools (strength of evidence: A to B). The UK RCGP guidelines found that group education based on the Swedish back school approach may be effective in occupational settings (strength of evidence: **). The UK RCGP guidelines found that the efficacy of back schools in non-occupational settings has not been demonstrated (strength of evidence: *). The European COST guidelines recommend against back schools for acute low back pain. The European COST guidelines recommend consideration of back schools where information is consistent with evidence-based recommendations for short-term (<6 weeks) pain relief and improvements in functional status, but recommend against back schools for chronic low back pain when aiming for long-term effects (>12 months).
Exercise
Results of search: systematic reviews A recent, higher-quality Cochrane review of exercise for low back pain did not include recurrences as an outcome613, 614. Only one systematic review reported low back pain recurrences, but was rated lower-quality1004.

APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials We identified four lower-quality trials that reported effects of exercise on recurrences of low back pain364, 1006-1008.
Efficacy of exercise versus no exercise for preventing recurrent episodes of low back pain
One trial found that a weekly, ongoing exercise program reduced the average number of low back pain episodes over a 1 year period by 0.27, compared to an average increase of 0.19 episodes in the no exercise group (Table 116)1007. However, this study had numerous methodologic shortcomings including unclear randomization and allocation concealment methods, unclear use of blinded outcomes assessment, and lack of intention-to-treat analysis with high loss to follow-up. Another small (n=39), lower-quality trial found medical management (advice and medications) plus an exercise program aimed at strengthening the multifidus muscle associated with a lower number of low back pain recurrences after 1 year (30% vs. 84%) and 2-3 years (35% vs. 70%) compared to medical management alone1006.
Table 116. Trial of exercise versus no exercise for preventing recurrent episodes of low back pain
Number of patients Duration of follow-up n=125 1.5 years
Author, year Kellett, 19911007
Hides, 20011006
n=39 3 years
Main results Exercise versus no exercise Mean episodes of low back pain in 1.5 years prior to intervention minus episodes during 1.5 years during intervention: 0.27 vs. -0.19 (p<0.05) Mean sick days in 1.5 years prior to intervention minus episodes during 1.5 years during intervention: 2.86 vs. 1.63 (p<0.02) Exercise (strengthening of multifidus) plus advice and medications versus advice and medications alone Rate of low back pain recurrences in 1st year: 6/20 (30%) vs. 16/19 (84%) in year 1 (p<0.05) Rate of continuing recurrences in years 2 and 3: 7/20 (35%) vs. 12/16 (75%) (p<0.05)
Quality score* 1/9
4/9
Efficacy of exercise versus education only for preventing recurrent episodes of low back pain
Two lower-quality trials both found that exercise reduced the number of back pain recurrences (Table 117)364, 1008. In one trial of patients with a back pain episode who had completed treatment and sick leave, a course of McKenzie extension exercises was associated with fewer low back pain recurrences than back education only through one year follow-up (44% vs. 74%)364. The benefit persisted from one to five years follow-up (proportion of patients with recurrences 64% vs. 88%, p<0.01). In the other trial, a 13-week course of a Mensendieck exercise program (incorporating exercises and education) was associated with fewer recurrences compared to information about the exercise program only during 12 months of follow-up (32% versus 57%, p<0.05)1008.

Table 117. Trials of exercise versus education for preventing recurrent episodes of low back pain
Number of patients Duration of follow-up n=77 12 months Stankovic, 1995364 n=100 5 years
Author, year Soukup, 19991008
Main results Mensendieck exercise program versus education only

Low back pain recurrences during 12 month follow-up: 32% (11/34) vs. 57% (20/35) (p<0.05) Sick leave (mean days): 30 vs. 38 (NS) Pain, 0 to 100 scale: 26 vs. 32 (p=0.22)
Quality score* 3/9
McKenzie exercise versus back education Recurrences: 44% (22/50) vs. 74% (37/50) after 1 year; 64% (30/47) vs. 88% (37/42) between 1 and 5 years (p<0.01) Sick leave between 1 and 5 years: 51% (24/47) vs. 74% (31/42) (p<0.03)
3/9
Efficacy of exercise versus other interventions for preventing recurrences of low back pain
One higher-quality trial (reviewed in detail in the section on self-care books) found that approximately 50% of subjects randomized to exercise, manipulation, or a self-care book experienced a recurrence of low back pain during the first year after the intervention, and 70% during the second year367. There were no differences in the proportion of patients who sought care for back pain in the second year (20% vs. 29% vs. 24%, p=0.29).
Summary of evidence
There is consistent evidence from two lower-quality trials that an exercise program is superior to education only for reducing long-term low back pain recurrences (level of evidence: fair). There is insufficient evidence (single lower-quality trials) to judge efficacy of an ongoing exercise program or an exercise program aimed at strengthening the multifidus muscles for reducing future episodes of low back pain (level of evidence: poor).

The other guidelines do not address exercise.
Lumbar supports
Results of search: systematic reviews One recent, higher-quality Cochrane review included no trials evaluating efficacy of lumbar supports for secondary prevention of low back pain385. Results of search: trials We found no additional trials.

Efficacy of lumbar supports for preventing recurrent episodes of low back pain
There are no trials on efficacy of lumbar supports for prevention of low back pain recurrences. The Cochrane review found moderate evidence that lumbar supports are not more effective than other interventions or no treatment for primary prevention of low back pain385.
Summary of evidence
No trials have evaluated the efficacy of lumbar supports for secondary prevention.

The European COST guidelines recommend against lumbar supports for prevention of low back pain.
Advice to stay active

Results of search: systematic reviews One recent, higher-quality Cochrane review of advice to stay active included no trials reporting low back pain recurrences as an outcome360. Results of search: trials We identified one lower-quality trial on effects of a multidisciplinary examination and advice to stay active on recurrent sick leave due to low back pain (Table 118)608. It was excluded from the Cochrane review360 because it didnt evaluate advice to stay active as a single intervention.
Efficacy of advice to stay active for preventing recurrent episodes of low back pain
One trial of patients on sick leave for 8 to 12 weeks due to low back pain found that a single visit to a spine clinic with examination by a physiatrist and physical therapist and advice on remaining active was associated with similar rates of recurrent episodes of low back pain compared to usual care through three years (62% vs. 61%, NS)360. There were also no differences in the proportion off sick leave at 3 years, though the intervention group was superior at 1 year followup (OR=1.60, 95% CI 1.08 to 2.39).
Table 118. Trial of spine clinic exam and advice to stay active versus usual care for preventing recurrent episodes of low back pain
Author, year Molde Hagen, 2003608
Main results Spine clinic exam and advice to stay active versus usual care New episodes of sick leave due to LBP (through 3 years): 62% (147/237) vs. 61% (135/220) (NS) LBP still present at 1 year: 47% vs. 52% (NS) On sick leave at 3 years: 64% vs. 62% (NS)
Quality score* 4/9

Summary of evidence
In patients on sick leave for low back pain for 8 to 12 weeks, one lower-quality trial found no difference in long-term (through 3 years) low back pain recurrences following randomization to a single spine clinic exam and advice to stay active versus usual care (level of evidence: poor).

Recommendations from other guidelines for advice are summarized in Key Question 3.
Early occupational medicine intervention

Results of search: systematic reviews We identified no systematic reviews evaluating the effects of an early occupational medicine intervention for preventing future episodes of low back pain. Results of search: trials We identified one lower-quality trial evaluating early evaluation by an occupational physician in workers with low back pain1009.
Efficacy of an early occupational medicine intervention versus usual care for preventing recurrent episodes of low back pain
One trial of hospital workers on sick leave for at least 10 days due to low back pain found that early, routine management by occupational physicians trained in recent guidelines was associated with a greater likelihood of recurrent sick leave due to low back pain than usual management by the workers supervisor for the first three months (52% vs. 25%, hazard ratio=2.4, 95% CI 1.2 to 4.7) (Table 119)1009. However, there were no differences in the amount of time until return to work (hazard ratio=1.3, 95% CI 0.90 to 1.90) or other outcomes. A high rate of crossovers (24%) in the usual care group and some deviation from the guidelines by the occupational medicine physicians could have affected results.
Table 119. Trial of early occupational medicine intervention versus usual care for preventing recurrent episodes of low back pain
Author, year Verbeek, 20021009
Main results Early intervention by an occupational physician versus no early intervention Time to return to work: 51 vs. 62 days (NS) Recurrence of sick leave in 1 year: 51% (26/51) vs. 25% (12/48) (p<0.05) Pain intensity (mean at 12 months, VAS 0 to 100): 24 vs. 30 (p=0.18) RDQ score (0 to 100): 20 vs. 21 (p=0.57)
Quality score* 4/9

Summary of evidence
In workers on sick leave for at least 10 days, an early occupational medicine intervention was associated with a greater likelihood of lower back pain recurrences than no early intervention in one lower-quality trial (level of evidence: poor).

The other guidelines do not address early occupational medicine intervention.
Psychological therapies, interdisciplinary rehabilitation, spinal manipulation, acupuncture, patient information or education
Results of search: systematic reviews Recent, higher-quality Cochrane reviews of psychological therapies301, interdisciplinary rehabilitation299, 300, 643, 644, and acupuncture69, 70 included no trials that reported rates of low back pain recurrences. One trial of spinal manipulation was discussed in the section on exercise therapy367. We found no systematic reviews on effects of patient information or education on recurrent low back pain. Results of search: trials We found no additional relevant trials for any of these interventions.
Summary of evidence
There is no evidence on effects of psychological therapies, interdisciplinary rehabilitation, and acupuncture on recurrent back pain episodes.

The other guidelines do not address psychological therapies, interdisciplinary rehabilitation, spinal manipulation, acupuncture, patient information or education.
Key Question 14 How effective are interventions for managing low back pain during pregnancy and post-partum?
We considered low back pain during pregnancy as separate from pelvic girdle pain (defined as pain experienced between the posterior iliac crest and the gluteal fold, particularly in the vicinity of the sacroiliac joints). The AHCPR, VA/DoD, UK RCGP, and European COST guidelines do not address low back pain in pregnancy, though the latter has developed a guideline on diagnosis and treatment of pelvic girdle pain1010. We excluded trials on management of back pain during labor.
Acupuncture during pregnancy

Results of search: systematic reviews A recent, higher-quality Cochrane review69, 70 of acupuncture (reviewed earlier in this report) included one lower-quality trial1011 of acupuncture versus exercise in pregnant women. This trial was also included in a systematic review of physical therapy for pregnancy-related back pain1012.

APS Clinical Guideline for the Evaluation and Management of Low Back Pain Results of search: trials From 373 potentially relevant citations, we identified two lower-quality trials1013, 1014 5 of acupuncture during pregnancy that were not included in the Cochrane review. Both compared acupuncture to usual care.
Efficacy of acupuncture versus usual care

Acupuncture was superior to usual care for pain relief in pregnant women in two lower-quality trials (Table 120)1013, 1014. One found a higher proportion of women reported >50% decrease in average pain intensity in the acupuncture group relative to usual care (78% vs. 15%, p<0.0001)1013. The other found decreased pain intensity in 60% of patients with acupuncture versus 14% with usual care (p<0.01)1014. Both trials also found acupuncture associated with increased capacity to perform general activities (p=0.01)1013 or decreased pain with activity (p<0.05)1014. One trial reported less use of other therapies with acupuncture compared to usual care (p<0.01)1014
Table 120. Trials of acupuncture versus usual care for low back pain during pregnancy
Author, year Guerreiro da Silva, 20041013
Kvorning, 20041014
n=72 From third trimester to birth
Main results Acupuncture vs. usual care Average pain (0 to 10), mean difference relative to baseline: -4.8 vs. +0.3 (p<0.0001) Average pain intensity decrease by > 50%: 78% (21/27) vs. 15% (5/34) (p<0.0001) Medication use, median number of daily doses between initial and final interviews: 0.0 vs. 2.0 (p=0.005) General activities functional status (0 to 10), median difference relative to baseline: -1.0 vs. 0.0 (p=0.01) Ability to perform work (0 to 10): 0.0 vs. +1.0 (p<0.001) Ability to walk (0 to 10): 0.0 vs. +2.0 (p<0.001). Acupuncture vs. usual care Pain intensity decreased: 60% vs. 14% (p<0.01) Decreased pain with activity: 43% vs. 9% (p<0.01) Analgesic drug use: 0% (0/37) vs. 15% (5/34) (p<0.05)
Quality score* 4/10
5/10
Efficacy of acupuncture versus exercise

Both systematic reviews included one lower-quality trial that found acupuncture moderately to substantially superior to exercise for mean pain scores after treatment (difference of about 1.5 to 3 points on a 10 point VAS pain scale)1011. Acupuncture was also more effective than exercise for improving functional status for various activities as measured by the Disability Rating Index, and a higher proportion of patients reported excellent or good pain relief with acupuncture (96% or 27/28 versus 78% or 14/18). However, there was a high drop-out rate in the exercise group (12/30), and drop-outs did not appear to be included in the data analysis. Harms None of the trials reported serious adverse effects in mothers or their infants following acupuncture1011, 1013, 1014. In one trial, two women had small bruises, three reported ecchymosis at

APS Clinical Guideline for the Evaluation and Management of Low Back Pain one or two points and one experienced a higher level of pain for a few hours after the first session1013. In another trial, symptoms were reported in 38% of 37 patients including local pain (n=6), heat or sweating (n=5), local hematoma (n=2), tiredness (n=2), nausea (n=2) and weakness (n=1)1014. Costs We found no studies evaluating costs
Summary of evidence
For low back pain during pregnancy, three lower-quality trials found acupuncture more effective than usual care (2 trials) or exercise (1 trial) for improving pain and function (level of evidence: fair).

The other guidelines do not address acupuncture during pregnancy
Physical therapy during pregnancy

Results of search: systematic reviews We identified one higher-quality systematic review of exercise or other physical therapy interventions for back pain during pregnancy that included five trials1012. Only one, a trial comparing water gymnastics to usual care, was rated higher-quality1015. Results of search: trials From 373 potentially relevant citations, we identified one additional trial (lower-quality) of a sitting pelvic tilt intervention1016
Efficacy of physical therapy versus usual care

The systematic review included five trials of physical therapy (exercise, education, advice, or combination of these interventions) for back pain compared to usual care1012. It did not attempt to pool trials because of diversity in the populations and interventions studied. In the only higher-quality trial, water gymnastics was associated with decreased pain relative to usual care1015. Pain intensity was lower in the water gymnastics group relative to the usual care group in the first postpartum week (p=0.034, data not reported). In addition, the water gymnastics group reported less absence from work after 32 weeks of pregnancy (OR 0.38, 95% CI 0.16-0.88). Two trials1017, 1018 included in the systematic review1012 found individualized exercise superior to usual care on measures of pain intensity. Group education or therapy, however, was superior to usual care in only one1019 of three1018, 1020 trials. All of these trials were rated lower-quality. The only trial with long-term (six years) follow-up found back pain during pregnancy appeared to return to baseline levels soon after pregnancy1021.

APS Clinical Guideline for the Evaluation and Management of Low Back Pain One lower-quality trial not included in the systematic review compared a sitting pelvic tilt exercise to no exercise (Table 121)1016. Those in the pelvic tilt intervention group had less pain at day 56 versus usual care (2.03 vs. 7.49 on a 10 point VAS, p<0.001).
Table 121. Trial of a sitting pelvic tilt interventions versus usual care for low back pain during pregnancy
Author, year Suputtitada, 20021016
Main results Sitting pelvic tilt exercise versus no exercise Pain (0 to 10), mean on day 56: 2.03 vs. 7.49 (p<0.05) Labor onset at 37-38 weeks: 56% vs. 20% (p<0.05) Birth weight, mean: 3009g vs. 3192g (p=0.018)
Quality score* 3/9
Efficacy of physical therapy versus other interventions

One systematic review included a trial that compared exercise therapy and acupuncture1011. We reviewed this trial in the section on efficacy of acupuncture versus other interventions. Harms In the trial of a sitting pelvic tilt exercise, labor onset was slightly earlier and birth weight slightly lower for those randomized to the pelvic tilt intervention, although there were no cases of preterm labor or low birth weight in either group1016. Other trials did not report adverse events associated with exercise in pregnancy. Costs We found no studies evaluating costs.
Summary of evidence
For back pain during pregnancy, water gymnastics was superior to usual care in one higherquality trial (level of evidence: fair). For back pain during pregnancy, individualized physiotherapy was superior to usual care in two lower-quality trials (level of evidence: fair). For back pain during pregnancy, evidence on efficacy of group education and exercise was mixed, with group education and exercise superior to usual care in one of three lower-quality trials (level of evidence: poor).
For back pain during pregnancy, a pelvic tilt exercise was associated with decreased pain in
one lower-quality trial, but also with lower birth weight and earlier (though full-term) onset of labor (level of evidence: poor)

The other guidelines do not address physical therapy during pregnancy

Massage during pregnancy

Results of search: systematic reviews One higher-quality systematic review of physical therapy interventions1012 included one lowerquality trial of massage therapy versus progressive relaxation therapy1022. Results of search: trials From 373 potentially relevant citations, we identified one lower-quality trial not included in the systematic review of massage versus progressive relaxation or usual care in depressed pregnant women1023.
Efficacy of massage versus usual care

In depressed pregnant women (n=84), mean back pain intensity was significantly lower with massage than with usual care immediately before the last of 32 treatment sessions (over 16 weeks), but the difference was only 0.3 points on a 10-point scale (Table 122)1023. Statistical significance of between-group differences was not reported. Outcomes at later follow-up were not reported.
Table 122. Trial of massage versus progressive relaxation and usual care for low back pain in depressed pregnant women
Number of patients Duration of follow-up n=84
Author, year Field, 20041023
Main results Massage vs. progressive relaxation vs. usual care (mean scores immediately before last treatment) 16 weeks Back pain (0 to 10): 2.9 vs. 4.0 vs. 2.6 (between group differences not reported) Anxiety (0 to 80): 42 vs. 45 vs. 39 (between group differences not reported) Mood (0 to 60): 8.2 vs. 9.6 vs. 8.7 (between group differences not reported) *Excludes criteria involving blinding of patients and care providers, for maximum score of 9
Quality score* 1/9
Efficacy of massage versus progressive relaxation therapy

In a small (n=26) trial of non-depressed women included in the systematic review, back pain intensity was lower in the massage therapy group compared to progressive relaxation after treatment (4.6 vs. 3.8 on a 10-point scale), but statistical significance of differences compared to baseline (3.8 vs. 3.2) were not reported1022. In a separate trial by the same investigator of depressed pregnant women, mean back pain intensity was moderately lower with massage compared to progressive relaxation immediately before the last of 32 treatment sessions (over 16 weeks), with a difference averaging 1.1 points on a 10-point scale1023. Statistical significance of between-group differences and longer-term follow-up results were not reported, and there were baseline differences between groups in pain and other baseline scores.

APS Clinical Guideline for the Evaluation and Management of Low Back Pain Harms In a trial of depressed pregnant women, scores on the Obstetric Complications Scale were higher (superior) in the massage group relative to the relaxation group (102.1 vs. 91.2), primarily related to decreased prematurity and low birth weight in the massage group1023. Costs We found no studies evaluating costs.
Summary of evidence
For back pain during pregnancy, two lower-quality trials found that massage therapy decreased pain scores, but effects appeared small, were only assessed during treatment, and it was not clear if the differences were significant relative to usual care or progressive relaxation therapy (level of evidence: poor).

The other guidelines do not address massage during pregnancy.
Supportive devices during pregnancy

Results of search: systematic reviews A Cochrane review of interventions during pregnancy1024 included one lower-quality crossover trial (unclear if randomized) of the Ozzlo pillow (a wedge-shaped pillow designed to give support to pregnant women while lying on their side in bed) versus a standard pillow1025. Results of search: trials From 373 potentially relevant citations, we found no additional trials.
Efficacy of supportive devices versus usual care

The Ozzlo pillow was superior to a standard pillow for pain at night (median score 14 vs. 19, p=0.002) and during the day (19 vs. 25, p=0.02), though there was no effect on sleeping scores1025. The pillow was rated as at least moderately useful by 47 of 92 women using it versus 31 of 92 using the standard pillow (OR 0.32, 95% CI 0.18-0.58). Harms No adverse events from the Ozzlo pillow were described. Costs We found no studies evaluating costs.
Summary of evidence
For back pain during pregnancy, there is insufficient evidence from one lower-quality trial to determine efficacy of the Ozzlo pillow versus standard pillows (level of evidence: poor).

The other guidelines do not address supportive devices during pregnancy.

Key Question 15 What is the cost-effectiveness associated with different interventions or management strategies (such as care provided by different types of providers) for managing low back pain?
We identified five recent systematic reviews of cost-effectiveness studies of different interventions or management strategies for low back pain1026-1030. All found few full costeffectiveness or cost-utility analyses and important methodological deficiencies in the available cost studies, including inadequate methods for identifying, valuing, and analyzing costs, and lack of sensitivity analyses for evaluating robustness of conclusions. In one systematic review, 12 of 17 included studies did not mention using the societal perspective to analyze costs1030. All of the systematic reviews concluded that current economic analyses are insufficient for determining the most cost-effectiveness interventions. Costs Individual cost studies are summarized separately elsewhere for each of the interventions reviewed in this report.
Systematic reviews published too recently to be included in this evidence review

Systematic reviews of non-invasive interventions that were published too recently to be included in this evidence review are shown in Appendix 9.
SUMMARY AND DISCUSSION

Specific findings from this evidence review are reported in the executive summary. We identified several key research gaps:
Nearly all trials are efficacy trials conducted in ideal setting and selected populations, usually with short-term follow-up. More effectiveness studies assessing long-term outcomes in more generalizable populations are needed to determine the effectiveness of interventions in realworld settings. For most interventions, data on harms are sparse, with disproportionate attention paid to benefits. Better assessment and reporting of harms (adhering to CONSORT recommendations1031) would help provide a more balanced assessment of the balance of benefits to harms associated with different back pain interventions. More research is needed on effective interventions for identification and treatment of yellow flags in order to prevent the development of chronic disabling low back pain. The optimal use of combinations of medications has not been well studied. In addition, emerging data on potential cardiovascular risks with non-selective NSAIDs may alter riskbenefit assessments. There is also little evidence on opioids specifically for low back pain. In particular, evidence on long-term use of opioids and risks of abuse and addiction remains sparse.

APS Clinical Guideline for the Evaluation and Management of Low Back Pain Decision tools or classification schemes for matching patients to interventions (such as manipulation, specific exercise regimens or other interventions) that they are more likely to benefit from are promising, but require additional validation. In addition, currently available tools include assessment of physical exam findings that many primary care clinicians are unfamiliar with or that have uncertain reliability and reproducibility. More research on decision tools or classification schemes that can be reliably used by most clinicians need to be developed and tested in clinical settings. The diagnostic value of provocative discography remains uncertain, and the use of discography to select patients for surgery or other invasive procedures has not been proven to improve clinical outcomes compared to selecting patients based on non-invasive testing. Clinical trials addressing this issue would be very helpful for resolving this long-standing controversy. Additional long-term trials with adequate follow-up and appropriate comparison interventions are needed to further clarify the role of fusion in patients with chronic non-specific low back pain. Confirmatory trials and trials that evaluate long-term outcomes associated with vertebral disc replacement are needed to help clarify its role as an option for surgical management. There is no evidence on optimal sequencing of interventions, and limited evidence on optimal combinations of interventions. In many cases, combinations of interventions were not much more effective than monotherapy, but more research is needed to clarify when and how treatments should be combined. High quality research on management of failed back surgery syndrome and back pain during pregnancy is lacking and provides little guidance for appropriate management in these populations. Few trials of medications and non-invasive interventions specifically evaluated patients with spinal stenosis or sciatica, and it remains unclear if optimal non-surgical treatments for this condition are different than for patients with non-specific low back pain without spinal stenosis or sciatica. Many interventions for low back pain appear to have similar effects on patient outcomes. Higher quality studies of cost-effectiveness could help clarify optimal choices between such interventions.

BIBLIOGRAPHY
1. Chou R, Qaseem A, Snow V, et al. Diagnosis and treatment of low back pain: A joint clinical practice guideline from the American College of Physicians and the American Pain Society. Ann Intern Med. 2007;147:478-491. Chou R, Loeser JD, Owens DK, et al. Interventional therapies, surgery and interdisciplinary rehabiltiation for low back pain: An evidence-based clnical practice guideline from the American Pain Society. Spine. 2009;In press. Walker BF. The prevalence of low back pain: a systematic review of the literature from 1966 to 1998. J Spinal Disord. 2000;13(3):205-217. Hart LG, Deyo RA, Cherkin DC. Physician office visits for low back pain: Frequency, clinical evaluation, and treatment patterns from a U.S. national survey. Spine. 1995;20:11-19. Deyo RA, Mirza SK, Martin BI. Back pain prevalence and visit rates. Spine. 2006;31(23):2724-2727. Carey TS, Evans AT, Hadler NM, et al. Acute severe low back pain: A population-based study of prevalence and care-seeking. Spine. 1996;221:339-344. Luo X, Pietrobon R, Sun SX, Liu GG, Hey L. Estimates and patterns of direct health care expenditures among individuals with back pain in the United States. Spine. 2004;29: 79-86. Straus BN. Chronic pain of spinal origin: The costs of intervention. Spine. 2002;27: 2614-2619. Andersson GBJ. Epidemiological features of chronic low-back pain. Lancet. 1999;354: 581-585. Frymoyer JW, Cats-Baril WL. An overview of the incidences and costs of low back pain. Orthop Clin North Am. 1991;22:263-271. Pengel LHM, Herbert RD, Maher CG, Refshauge KM. Acute low back pain: systematic review of its prognosis. BMJ. 2003;327:323-327. Von Korff M, Saunders K. The course of back pain in primary care. Spine. 1996;21: 2833-2839. Hestbaek L, Leboeuf-Yde C, Manniche C. Low back pain: what is the long-term course? A review of studies of general patient populations. Eur Spine J. 2003;12:149-165. 14. Gatchel RJ, Polatin PB, Mayer TG. The dominant role of psychosocial risk factors in the development of chronic low back pain disability. Spine. 1995;20:2702-2709. Cherkin DC, Deyo RA, Wheeler K, Ciol MA. Physician variation in diagnostic testing for low back pain. Who you see is what you get. Arth Rheum. 1994;37:15-22. Cherkin DC, Deyo RA, Wheeler K, Ciol MA. Physician views about treating low back pain: the results of a national survey. Spine. 1995;20:1-10. Cherkin DC, Deyo RA, Loeser JD, Bush T, Waddell G. An international comparison of back surgery rates. Spine. 1994;19: 1201-1206. Volinn E, Mayer J, Diehr P, Van Koevering D, Connell FA, Loeser JD. Small area analysis of surgery for low-back pain. Spine. 1992;17:575-579. Carey TS, Garrett J, Jackman A, McLaughlin C, Fryer J, Smucker DR. The outcomes and costs of care for acute low back pain among patients seen by primary care practitioners, chiropractors, and orthopedic surgeons. N Engl J Med. 1995;333:913-917. Shekelle PG, Markovich M, Louie R. Comparing the costs between provider types of episodes of back pain care. Spine. 1995;20:221-227. Deyo RA. Fads in the treatment of low back pain. N Engl J Med. 1991;325(14): 1039-1040. Deyo RA. Treatments for back pain: can we get past trivial effects? Ann Intern Med. 2004;141:957-958. Spitzer W. Scientific approach to the assessment and management of activityrelated spinal disorders. A monograph for clinicians. Report of the Quebec Task Force on Spinal Disorders. Spine. List 4/20/05 1987;12(Suppl 1):S1-S59. Bigos SJ, Bowyer R, Braen R, et al. Clinical Practice Guideline 14. Acute Low Back Problems in Adults: AHCPR 1994. Deyo RA. The messenger under attack-intimidation of researchers by special-interest groups. N Engl J Med. 1997;336:1176-1180. Koes B, van Tulder M, Ostelo R, Kim Burton A, Waddell G. Clinical guidelines for the management of low back pain in primary care: an international comparison. Spine.
15.
2.
16.
3.
17.
4.
18.
5.
19.
6.
20.
7.
21.
8.
22.
9.
23.
10.
11.
24.
12.
25.
13.
26.

2001;26(22):2504-2513, discussion 2513-2504. Clancy CM, Cronin K. Evidence-based decision making: global evidence, local decisions. Health Affairs. 2005;24:151-162. Arnau JM, Vallano A, Lopez A, Pellise F, Delgado MJ, Prat N. A critical review of guidelines for low back pain treatment. Eur Spine J. 2006;15:543-553. van Tulder MW, Tuut MK, Pennick V, Bombardier C, Assendelft WJJ. Quality of primary care guidelines for acute low back pain. Spine. 2004;29:E357-E362. McGuirk B, King W, Govind J, Lowry J, Bogduk M. Safety, efficacy, and cost effectiveness of evidence-based guidelines for the management of acute low back pain in primary care. Spine. 2001;26:2615=2622. Buchbinder R, Jolley D, Wyatt M. Population base intervention to change back pain beliefs and disability: three part evaluation. BMJ. 2001;322:1516-1520. Buchbinder R, Jolley D. Improvements in general practitioner beliefs and stated management of back pain persist 4.5 years after the cessation of a public health media campaign. Spine. 2007;32:E156-E162. Goldberg HI, Deyo RA, Taylor VM, et al. Can evidence change the rate of back surgery? A randomized trial of community-based education. Effective Clinical Practice. 2001;4:95-104. VA/DoD Evidence-Based Clinical Practice Guideline Working Group. Low Back Pain or Sciatica in the Primary Care Setting Veterans Health Administration, Department of Veterans Affairs and Health Affairs, Department of Defense; 1999. Waddell G, McIntosh A, Hutchinson A, Feder G, Lewis M. Clinical guidelines for the management of acute low back pain Evidence Review. Royal College of General Practitioners London, England: RCGP; 1996. Waddell G, McIntosh A, Hutchinson A, Feder G, Lewis M. Clinical guidelines for the management of acute low back pain. Royal College of General Practitioners London, England: RCGP; 1996. Airaksinen O, Brox J, Cedraschi C, et al. Chapter 4 - European guidelines for the management of chronic nonspecific low back pain. Eur Spine J. 2006;15(Suppl 2): s192-s300. 38. Burton A, Balagu F, Cardon G, et al. Chapter 2 - European guidelines for prevention in low back pain. European Spine Journal. 2006;15(Suppl 2):s136-s168. van Tulder M, Becker A, Bekkering T, et al. Chapter 3 - European guidelines for the management of acute nonspecific low back pain in primary care. European Spine Journal. 2006;15(Suppl 2):s169-s191. Michel LJ, Wood R. Back pain in y oung athletes: significant differences from adults in causes and patterns. Arch Pediatr Adolesc Med. 1995;149:15-18. Combs JA, Caskey PM. Back pain in children and adolescents: a retrospective review of 648 patients. South Med J. 1997;90:789-792. Bombardier C. Outcome assessments in the evaluation of treatment of spinal disorders: summary and general recommendations. Spine. 2000;25(24):3100-3103. Deyo RA, Battie M, Beurskens AJHM, et al. Outcome measures for low back pain research: a proposal for standardized use. Spine. 1998;23(18):2003-2013. Fairbank JCT, Pynsent PB. The Oswestry Disability Index. Spine. 2000;25:2940-2953. Bombardier C, Hayden JA, Beaton DE. Minimal clinically important difference. Low back pain: outcome measures. J Rheumatol. 2001;28:431-438. Ware JE. SF-36 health survey update. Spine. 2000;25:3130-3139. Von Korff M, Jensen MP, Karoly P. Assessing global pain severity by self-report in clinical and health services research. Spine. 2000;25:3140-3151. Salaffi F, Stancati A, Silvestri CA, Ciapetti A, Grassi W. Minimal clinically important changes in chronic musculoskeletal pain intensity measured on a numerical rating scale. Eur J Pain. 2004;8:283-291. Drummond MF, Richardson WS, O'Brien BJ, Levine M, Heyland D. Users' guides to the medical iterature. XIII. How to use an article on economic analysis of clinical practice. A. Are the results of the study valid? JAMA. 1997;277:19. Oxman AD, Guyatt GH. Validation of an index of the quality of review articles. J Clin Epidemiol. 1991;44(11):1271-1278. Furlan AD, Clarke J, Esmail R, Sinclair S, Irvin E, Bombardier C. A critical review of reviews on the treatment of chronic low back pain. Spine. 2001;26(7):E155-E162.
27.
39.
28.
29.
40.
30.
41.
42.
31.
43.
32.
44. 45.
33.
46. 47.
34.
48.
35.
49.
36.
50.
37.
51.

52. Jadad AR, McQuay HJ. Meta-analyses to evaluate analgesic interventions: a systematic qualitative review of their methodology. J Clin Epidemiol. 1996;49: 235-243. Canter PH, Ernst E. Sources of bias in reviews of spinal manipulation for back pain. Wien Klin wochenschr. 2005;117(9-10): 333-341. Bombardier C, Esmail R, Nachemson AL, the Back Review Group Editorial Board. The Cochrane Collaboration Back Review Group of Spinal Disorders. Spine. 1997;22:837-840. Bouter LM, Pennick V, Bombardier C, the Editorial Board of the Back Review Group. Cochrane Back Review Group. Spine. 2003;28:1215-1218. van Tulder M, Furlan AD, Bombardier C, Bouter L, the Editorial Board of the Cochrane Collaboration Back Review Group. Updated method guidelines for systematic reviews in the Cochrane Collaboration Back Review Group. Spine. 2003;28(12):1290-1299. Deeks J, Dinnes J, D'Amico R, et al. Evaluating non-randomized intervention studies. Health Technol Assess. 2003;7(27):1-192. Harris RP, Helfand M, Woolf SH, et al. Current methods of the U.S. Preventive Services Task Force. Am J Prevent Med. 2001;20(3S):21-35. Lijmer JG, Mol BW, Heisterkamp S, et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA. 1999;282:1061-1066. Whiting P, Rutjes AWS, Reitsma JB, Glas AS, Bossuyt PMM, Kleijnen J. Sources of variation and bias in studies of diagnostic accuracy: a systematic review. Ann Intern Med. 2004;140:189-202. Malmivaara A, Koes BW, Bouter LM, van Tulder MW. Applicability and clinical relevance of results in randomzied controlled trials. The Cochrane review on exercise therapy for low back pain as an example. Spine. 2006;31:1405-1409. Song F, Altman D, Glenny A-M, Deeks J. Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses. BMJ. March 1, 2003;326(7387):472-. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Earlbaum Associates; 1988. 64. Hagen KB, Hilde G, Jamtvedt G, Winnem M. Bed rest for acute low-back pain and sciatica. Cochrane Database of Systematic Reviews. 2004(4):Art. No.: CD001254. Hagen KB, Jamtvedt G, Hilde G, Winnem MF. The updated Cochrane review of bed rest for low back pain and sciatica. Spine. 2005;30(5):542-546. Assendelft WJJ, Morton SC, Yu EI, Suttorp MJ, Shekelle PG. Spinal manipulative therapy for low back pain. A meta-analysis of effectiveness relative to other therapies Ann Intern Med. 2003;138(11):871-881. Assendelft WJJ, Morton SC, Yu EI, Suttorp MJ, Shekelle PG. Spinal manipulative therapy for low-back pain Cochrane Database of Systematic Reviews. 2004(1):Art. No.: CD000447. Manheimer E, White A, Berman B, Forys K, Ernst E. Meta-Analysis: Acupuncture for Low Back Pain. Ann Intern Med. 19, 2005 2005;142(8):651-663. Furlan A, van Tulder M, Cherkin D, et al. Acupuncture and Dry-Needling for Low Back Pain: An Updated Systematic Review Within the Framework of the Cochrane Collaboration. Spine. 2005;30(8):944-963. Furlan AD, van Tulder MW, Cherkin DC, et al. Acupuncture and dry-needling for low back pain Cochrane Database of Systematic Reviews. 2005(1):Art. No.: CD001351. Abdi S, Datta S, Trescot AM, et al. Epidural steroids in the management of chronic spinal pain: A systematic review. Pain Physician. 2007;10:185-212. Andersson GBJ, Mekhail NA, Block JE. Treatment of intractable discogenic low back pain. A systematic review of spinal fusion and intradiscal electrothermal therapy (IDET). Pain Physician. 2006;9:237-248. Appleby D, Andersson G, Totta M. Metaanalysis of the efficacy and safety of intradiscal electrothermal therapy (IDET). Pain Medicine. 2006;7(4):308-316. Armon C, Argoff CE, Samuels J, Backonja MM. Assessment: use of epidural steroid injections to treat radicular lumbosacral pain: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology. 2007;68(10):723-729. Boswell MV, Colson JD, Sehgal N, Dunbar EE, Epter R. A systematic review of therapeutic facet joint interventions in chronic
65.
53.
66.
54.
55.
67.
56.
68.
69.
57.
58.
70.
59.
71.
60.
72.
61.
73.
74.
62.
63.
75.

spinal pain. Pain Physician. 2007;10: 229-253. Dagenais S, Yelland MJ, Del Mar C, Schoene M. Prolotherapy injections for chronic lowback pain. Cochrane Database of Systematic Reviews. 2007(2):Art. No.: CD004059. DePalma MJ, Bhargava A, Slipman CW. A critical appraisal of the evidence for selective nerve root injection in the treatment of lumbosacral radiculopathy. Arch Phys Med Rehabil. 2005;86(7):1477-1483. Geurts J, van Wijk R, Stolker R, Groen G. Efficacy of radiofrequency procedures for the treatment of spinal pain: a systematic review of randomized clinical trials. Reg Anesth Pain Med. 2001;26(5):394-400. Gibson J, Waddell G. Surgery for degenerative lumbar spondylosis: updated Cochrane Review. Spine. 2005;30(20): 2312-2320. Gibson JNA, Waddell G, Grant IC. Surgery for degenerative lumbar spondylosis Cochrane Database of Systematic Reviews. 2005(2):Art. No.: CD001352. Gibson JN. Surgical interventions for lumbar disc prolapse: updated Cochrane Review. Spine. 2007;32:1735-1747. Gibson J, Waddell G. Surgical interventions for lumbar disc prolapse. Cochrane Database of Systematic Reviews. 2007(1):Art. No.: CD001350. Hansen H, McKenzie-Brown AM, Cohen SP, Swicegood JR, Colson JD, Manchikanti L. Sacroiliac joint Interventions: A systematic review. Pain Physician. 2007;10:165-184. Luijsterburg PAJ, Verhagen AP, Ostelo RWJG, van Os TAG, Peul WC, Koes BW. Effectiveness of conservative treatments for the lumbosacral radicular syndrome: a systematic review. Eur Spine J. 2007;16 (7):881-899. Mailis-Gagnon A, Furlan A, Sandoval J, Taylor R, Cochrane Pain Palliative Care and Supportive Care Group. Spinal cord stimulation for chronic pain Cochrane Database of Systematic Reviews. 2004(3):Art. No.: CD003783. Nelemans P, deBie R, deVet H, Sturmans F. Injection therapy for subacute and chronic benign low back pain. Spine. 2001;26(5): 501-515. NICE (National Institute for Clinical Excellence). Interventional procedures overview of percutaneous intradiscal radiofrequency thermocoagulation for lower back pain: 2004. NICE (National Institute for Clinical Excellence). Interventional procedures overview of percutaneous disc decompression using coblation for lower back pain. National Institute for Clinical Excellence; 2004. NICE (National Institute for Clinical Excellence). Interventional procedures overview of percutaneous intradiscal electrothermal therapy for lower back pain. National Institute for Clinical Excellence; 2004. Niemisto L, Kalso E, Malmivaara A, Seitsalo S, Hurri H. Radiofrequency denervation for neck and back pain. A systematic review of randomized controlled trials. Cochrane Database of Systematic Reviews. 2003(1):Art. No.: CD004058. Niemisto L, Kalso E, Malmivaara A, Seitsalo S, Hurri H, Cochrane Collaboration Back Review Group. Radiofrequency denervation for neck and back pain: a systematic review within the framework of the Cochrane Collaboration Back Review Group. Spine. 2003;28(16):1877-1888. Resnick DK, Choudhri TF, Dailey AT, et al. Guidelines for the performance of fusion procedures for degenerative disease of the lumbar spine. Part 13: injection therapies, low-back pain, and lumbar fusion. J Neurosurg Spine. 2005;2(6):707-715. Slipman CW, Bhat AL, Gilchrist RV, Issac Z, Chou L, Lenrow DA. A critical review of the evidence for the use of zygapophysial injections and radiofrequency denervation in the treatment of low back pain. Spine J. 2003;3(4):310-316. Staal J, de Bie R, De Vet H, Hildebrandt J, Nelemans P. Injection therapy for subacute and chronic low-back pain. Cochrane Database of Systematic Reviews. 2008(3):Art. No.: CD001824. Taylor RS. Spinal cord stimulation in complex regional pain syndrome and refractory neuropathic back and leg pain/failed back surgery syndrome: results of a systematic review and meta-analysis. J Pain Symptom Manage. 2006;31 (4 Suppl 1):S13-S19. Taylor RS, Van Buyten J, Buscher E. Spinal cord stimulation for chronic back pain and leg pain and failed back surgery syndrome: a systematic review and analysis of
76.
88.
77.
89.
78.
90.
79.
80.
91.
81.
82.
92.
83.
93.
84.
94.
85.
95.
86.
96.
87.

progressive factors. Spine. 2005;30(1): 152-160. Tonkovich-Quaranta L, Winkler S. Use of epidural corticosteroids in low back pain. Ann Pharmacother. 2000;34(10):1165-1172. Turner JA, Loeser JD, Deyo RA, Sanders SB. Spinal cord stimulation for patients with failed back surgery syndrome or complex regional pain syndrome: a systematic review of effectiveness and complications. Pain. 2004;108(1-2):137-147. Urrutia G, Kovacs F, Nishishinya MB, Olabe J. Percutaneous thermocoagulation intradiscal techniques for discogenic low back pain. Spine. 1 2007;32(10):1146-1154. Vroomen P, de Krom M, Slofstra P, Knottnerus J. Conservative treatment of sciatica: a systematic review. J Spinal Disord. 2000;13(6):463-469. Ackerman WE, 3rd, Ahmad M. The efficacy of lumbar epidural steroid injections in patients with lumbar disc herniations. Anesth Analg. 2007;104(5):1217-1222. Bonetti M, Fontana A, Cotticelli B, Volta GD, Guindani M, Leonardi M. Intraforaminal O2O3 versus periradicular steroidal infiltration in lower back pain: randomized controlled study. Am J Neuroradiol. 2005;26:996-1000. Burton AK, Tillotson KM, Cleary J. Singleblind randomised controlled trial of chemonucleolysis and manipulation in the treatment of symptomatic lumbar disc herniation. Eur Spine J. 2000;9:202-207. Buttermann GR. The effect of spinal steroid injections for degenerative disc disease. Spine J. 2004;4(5):495-505. Foster L, Clapp L, Erickson M, Jabbari B. Botulinum toxin A and chronic low back pain: a randomized, double-blind study. Neurology. 2001;56(10):1290-1293. Fukusaki M, Kobayashi I, Hara T, Sumikawa K. Symptoms of spinal stenosis do not improve after epidural steroid injection. Clin J Pain. 1998;14(2):148-151. Gallucci M, Limbucci N, Zugaro L, et al. Sciatica: treatment with intradiscal and intraforaminal injections of steroid and oxygen-ozone versus steroid only. Radiology. Mar 2007;242(3):907-913. Geurts J, van Wijk RM, Wynne HJ, et al. Radiofrequency lesioning of dorsal root ganglia for chronic lumbosacral radicular pain: a randomised, double-blind, controlled trial. Lancet. 2003;361(9351):21-26. 109. Graham C. Chemonucleolysis. A preliminary report on a double blind study comparing chemonucleolysis and intradiscal administration of hydrocortisone in the treatment of backache and sciatica. Orthop Clin North Am. 1975;6(1):259-263. Hoogland T, Schubert M, Miklitz B, Ramirez A. Transforaminal posterolateral endoscopic discectomy with or without the combination of a low-dose chymopapain: a prospective randomzied study in 280 consecutive cases. Spine. 2006;31:E890-E897. Jeong HS, Lee JW, Kim SH, Myung JS, Kim JH, Kang HS. Effectiveness of transforaminal epidural steroid injection by using a preganglionic approach: a prospective randomized controlled study. Radiology. 2007;245(2):584-590. Khot A, Bowditch M, Powell J, Sharp D. The use of intradiscal steroid therapy for lumbar spinal discogenic pain: a randomized controlled trial. Spine. 2004;29(8):833-836. Kumar K, Taylor RS, Jacques L, et al. Spinal cord stimulation versus conventional medical management for neuropathic pain: A multicentre randomised controlled trial in patients with failed back surgery syndrome. Pain. 2007;132(1-2):179-188. Luukkainen R, Wennerstrand P, Kautiainen H, Sanila M, Asikainen E. Efficacy of periarticular corticosteroid treatment of the sacroiliac join in non-spondyloarthropathic patients with chronic low back pain in the region of the sacroiliac joint. Clin Exper Rheumatol. 2002;20:52-54. Manchikanti L, Rivera JJ, Pampati V, et al. One day lumbar epidural adhesiolysis and hypertonic saline neurolysis in treatment of chronic low back pain: A randomized, doubleblind trial. Pain Physician. 2004;7(2):177-186. Nash TP. Facet joints - intra-articular steroids or nerve block? Pain Clinic. 1989;3(2):77-82. Nath S, Nath C, Pettersson K. Percutaneous lumbar zygapophysial (facet) joint neurotomy using radiofrequency current, in the management of chronic low back pain. Spine. 2008;33:1291-1297. Oh WS, Shim JC. A randomized controlled trial of radiofrequency denervation of the ramus communicans nerve for chronic discogenic low back pain. Clin J Pain. 2004;20(1):55-60. Simmons JW, McMillin JN, Emery SF, Kimmich SJ. Intradiscal steroids. A
97.
98.
110.
99.
111.
100.
112.
101.
113.
102.
103.
114.
104.
115.
105.
106.
116. 117.
107.
118.
108.
119.

prospective double-blind clinical trial. Spine. 1992;17(6 Suppl):S172-175. Tekin I, Mirzai H, Ok G, Erbuyun K, Vatansever D. A comparison of conventional and pulsed radiofrequency denervation in the treatment of chronic facet joint pain. Clin J Pain. 2007;23(6):524-529. van Wijk R, Geurts J, Wynne H, et al. Radiofrequency Denervation of Lumbar Facet Joints in the Treatment of Chronic Low Back Pain: A Randomized, Double-Blind, Sham Lesion-Controlled Trial. Clin J Pain 2005;21(4):335-344. Wittenberg RH, Oppel S, Rubenthaler FA, Steffen R. Five-year results from chemonucleolysis with chymopapain or collagenase. Spine. 2001;26:1835-1841. Arden NK, Price C, Reading I, et al. A multicentre randomized controlled trial of epidural corticosteroid injections for sciatica. Rheumatology. 2005;44:1399-1406. Barendse GA, van Den Berg SG, Kessels AH, Weber WE, van Kleef M. Randomized controlled trial of percutaneous intradiscal radiofrequency thermocoagulation for chronic discogenic back pain. Spine. 2001;26(3): 287-292. Beliveau P. A comparison between epidural anaesthesia with and without corticosteroid in the treatment of sciatica. Rheumatol Phys Med. 1971;11:40-43. Breivik H, Hesla PE, Molnar I, Lind B. Treatment of chronic low back pain and sciatica: comparison of caudal epidural injections of bupivicaine and methylprednisolone with bupivicaine followed by saline. Adv Pain Res Ther. 1976;1: 927-932. Bromley J, Varma A, Santoro A, Cohen P, Jacobs R, Berger L. Double-blind evaluation of collagenase injections for herniated lumbar discs. Spine. 1984;9:486-488. Bush K, Hillier S. A controlled study of caudal epidural injections of triamcinolone plus procaine for the management of intractable sciatica. Spine. 1991;16:572-575. Carette S, Marcoux S, Truchon R, et al. A controlled trial of corticosteroid injections into facet joints for chronic low back pain. N Engl J Med. 1991;325:1002-1007. Carette S, Leclaire R, Marcoux S, et al. Epidural corticosteroid injections for sciatica due to herniated nucleus pulposus. N Engl J Med. 1997;336:1634-1640. 131. Collee G, Dijkmans B, Vandenbroucke J, Cats A. Iliac crest pain syndrome in low back pain. A double blind, randomized study of local injection therapy. J Rheumatol. 1991;18:1060-1063. Cuckler J, Bernini P, Wiesel S, Booth R, Rothman R, Pickens G. The use of epidural steroids in the treatment of lumbar radicular pain. J Bone Joint Surg 1985;67A:63-66. Dabezies E, Langford K, Morris J, MShields C, Wilkinson H. Safety and efficacy of chymopapain (Discase) in the treatment of sciatica to a herniated nucleus pulposus. Results of a randomized, double-blind study. Spine. 1988;13(5):561-565. Dechow E, Davies R, Carr A, Thompson P. A randomized, double-blind, placebo-controlled trial of sclerosing injections in patients with chronic low back pain. Rheumatology. 1999;38:1255-1259. Dilke TFW, Burry HC, Grahame R. Extradural corticosteroid injection in management of lumbar nerve root compression. Br Med J. 1973;16(2):635-637. Fraser R. Chymopapain for the treatment of intervertebral disc herniation. A preliminary report of a double-blind study. Spine. 1982;7(6):608-612. Fraser R. Chymopapain for the treatment of intervertebral disc herniation. The final report of a double-blind study. Spine. 1984;9(8):815-818. Freeman BJ, Fraser RD, Cain CM, Hall DJ, Chapple DC. A randomized, double-blind, controlled trial: intradiscal electrothermal therapy versus placebo for the treatment of chronic discogenic low back pain. Spine. 2005;30(21):2369-2377. Gallagher J, Petriccione D, Wedley J, et al. Radiofrequency facet joint denervation in the treatment of low back pain: A prospective controlled double-blind study to assess its efficacy Pain Clinic. 1994;7(3):193-198. Garvey T, Marks M, Wiesel S. A prospective, randomized, double-blind evaluation of trigger-point injection therapy for low-back pain. Spine. 1989;14:962-964. Gogan W, Fraser R. Chymopapain. a 10year, double-blind study. Spine. 1992;17(4):388-394. Hameroff S, Crago B, Blitt C, Womble J, Kanel J. Comparison of bupivacaine, etidocaine, and saline for trigger-point therapy. Anaesth Analg. 1981;60:752-755.
120.
132.
121.
133.
122.
134.
123.
135.
124.
136.
125.
137.
126.
138.
127.
139.
128.
140.
129.
141.
130.
142.

143. Helliwell M. Outpatient treatment of low back pain and sciatica by a single extradural corticosteroid injection. Br J Clin Pract. 1985;39:228-231. Javid M, Nordby E, Ford L, et al. Safety and efficacy of chymopapain (chymodiactin) in herniated nucleus pulposus with sciatica. JAMA. 1983;249(18):2489-2494. Karppinen J, Malmivaara A, Kurunlahti M, et al. Periradicular infiltration for sciatica: a randomized controlled trial. Spine. 2001;26(9):1059-1067. Klein R, Eek B, DeLong W, Mooney V. A Randomized Double-Blind Trial of DextroseGlycerine-Phenol Injections for Chronic, Low Back Pain. J Spin Disord. 1993;6(1):23-33. Klenerman L, Greenwood R, Davenport H, White D, Peskett S. Lumbar epidural injections in the treatment of sciatica. Brit J Rheum. 1984;23(1):35-38. Kraemer J, Ludwig J, Bickert U, Owczarek V, Traupe M. Lumbar epidural perineural injection: a new technique. Eur Spine J. 1997;6(5):357-361. Leclaire R, Fortin L, Lambert R, Bergeron Y, Rossignol M. Radiofrequency facet joint denervation in the treatment of low back pain: a placebo-controlled clinical trial to assess efficacy. Spine. 2001;26(13):1411-1416. Lilius G, Lassonen AM, Myllynen P, Harilainen A, Salo L. The lumbar facet joint syndrome--significance of inappropriate signs. A randomized, placebo-controlled trial. French J Orthop Surg. 1989;3:479-486. Mathews J, Mills S, Jenkins V, et al. Back pain and sciatica: controlled trials of manipulation, traction, sclerosant and epidural injection. Br J Rheumatol. 1987;26:416-423. Ng L, Chaudhary N, Sell P. The efficacy of corticosteroids in periradicular infiltration for chronic radicular pain. A randomized, doubleblind, controlled trial. Spine. 2005;30: 857-862. North RB, Kidd DH, Farrokhi F, Piantadosi SA. Spinal cord stimulation versus repeated lumbosacral spine surgery for chronic pain: a randomized, controlled trial. Neurosurgery. 2005;56:98-107. North RB, Kidd DH, Lee MS. A prospective, randomized study of spinal cord stimulation versus reoperation for failed back surgery syndrome: inital results. Sterotact Funct Neurosurg.1994;62(1-4):267-272. 155. North RB, Kidd DH, Lee MS. Spinal cord stimulation versus reoperation for failed back surgery syndrome: a prospective study design. Acta Neurochir Suppl. 1995;64: 106-108. Ongley M, Klein R, Dorman T, Eek B, Hubert L. A new approach to the treatment of chronic low back pain. Lancet. 1987;2(8551): 143-146. Pauza KJ, Howell S, Dreyfuss P, Peloza JH, Dawson K, Bogduk N. A randomized, placebo-controlled trial of intradiscal electrothermal therapy for the treatment of discogenic low back pain. Spine J. 2004;4(1):27-35. Ridley M, Kingsley G, Gibson T, Grahame R. Outpatient lumbar epidural corticosteroids in the management of sciatica. Br J Rheumatol. 1988;27:295-299. Riew K, Yin Y, Gilula L, et al. The effect of nerve-root injections on the need for operative treatment of lumbar radicular pain. A prospective, randomized, controlled, double-blind study. J Bone Joint Surg 2000;82-A(11):1589-1593. Riew KD, Park J-B, Cho Y-S, et al. Nerve root blocks in the treatment of lumbar radicular pain. A minimum five-year follow-up. J Bone Joint Surg. 2006;88(8):1722-1725. Rogers P, Nash T, Schiller D, Norman J. Epidural steroids for sciatica. Pain Clin. 1992;5:67-72. Schwetschenau R, Ramirez A, Johnston J, Barnes E, Wiggs C, Martins AN. Double-blind evaluation of intradiscal chymopapain for herniated lumbar discs. Early results. J Neurosurg. 1976;45:622-627. Martins A, Ramirez A, Johnston J, Schwetschenau P. Double blind evaluation of chemonucleolysis for herniated lumbar discs. Late results. J Neurosurg. 1978;49:816-827. Snoek W, Weber H, Jorgensen B. Double blind evaluation of extradural methyl prednisolone for herniated lumbar discs. Acta Orthop Scand. 1977;48:635-641. Sonne M, Chirstensen K, Hansen S, Jensen E. Injection of steroids and local anaesthetics as therapy for low-back pain. Scand J Rheum. 1985;14:343-345. van Kleef M, Barendse G, Kessels A, Voets H, Weber W, de Lange S. Randomized trial of radiofrequency lumbar facet denervation for chronic low back pain. Spine. 1999;24(18):1937-1942.
144.
156.
145.
157.
146.
158.
147.
159.
148.
149.
160.
150.
161.
162.
151.
163.
152.
164.
153.
165.
154.
166.

167. Wilson-MacDonald J, Burt G, Griffin D, Glynn C. Epidural steroid injection for nerve root compression. A randomised, controlled trial. The J Bone Joint Surg. 2005;87(3):352-355. Yelland M, Glasziou P, Bogduk N, Schluter P, McKernon M. Prolotherapy injections, saline injections, and exercises for chronic low-back pain: a randomized trial. Spine. 2004;29(1): 9-16. Zahaar M. The value of caudal epidural steroids in the treatment of lumbar neural compression syndromes. J Neurol Orthop Med Surg. 1991;12:181-184. Abdi S, Datta S, Lucas L. Role of epidural steroids in the management of chronic spinal pain: A systematic review of effectiveness and complications: A systematic review. Pain Physician. 2005;8(1):127-143. Boswell M, Colson J, Spillane W. Therapeutic facet joint interventions in chronic spinal pain: A systematic review of their role in chronic spinal pain management and complications. Pain Physician. 2005;8(1):101-114. Boswell M, Hansen H, Trescot A, Hirsch J. Epidural steroids in the management of chronic spinal pain and radiculopathy. Pain Physician. 2003;6:319-334. Cannon D, Aprill C. Lumbosacral epidural steroid injections. Arch Phys Med Rehabil. 2000;81(3 supl 1):S87-S98. Chou LH, Lew HL, Coelho PC, Slipman CW. Intradiscal electothermal annuloplasty. Am J Phys Med Rehabil. 2005;84:538-549. Difazio M, Jabbari B. A focused review of the use of botulinum toxins for low back pain. Clin J Pain. 2002;18(6 suppl.):S155-S162. Gibson JNA, Grant IC, Waddell G. The Cochrane review of surgery for lumbar disc prolapse and degenerative lumbar spondylosis. Spine. 1999;24:1820-1832. Gibson JNA, Grant IC, Waddell G. Surgery for lumbar disc prolapse. Cochrane Database of Systematic Reviews. 2000(3):Art. No.: CD001350. Haselkorn JK. Epidural steroid injections and the management of sciatica: a meta-analysis. Arch Phys Med Rehabil. 1995;76:1037. Heary RF. Intradiscal electothermal annuloplasty: The IDET procedure. J Spin Disord. 2001;14:353-360. Hooten WM, Martin DP, Huntoon MA. Radiofrequency neurotomy for low back pain: Evidence-based procedural guidelines. Pain Med. 2005;6(2):129-138. 181. Kepes ER, Duncalf D. Treatment of backache with spinal injections of local anesthetics, spinal and systemic steroids. A review. Pain. 1985;22(1):33-47. Kim S, Stitik T, Foye P, Greenwald B, Campagnolo D. Critical review of prolotherapy for osteoarthritis, low back pain, and other musculoskeletal conditions: A physiatric perspective. Am J Phys Med Rehabil. 2004;83:379-389. Koes B. Epidural steroid injections for low back pain and sciatica: an updated systematic review of randomized clinical trials. Pain Digest. 1999;9:241. Koes BW, Scholten RJ, Mens JM, Bouter LM. Efficacy of epidural steroid injections for lowback pain and sciatica: a systematic review of randomized clinical trials. Pain. 1995;63(3):279-288. Manchikanti L, Singh V, Vilims B, Hansen H, Schultz D, Kloth D. Medial Branch Neurotomy in Management of Chronic Spinal Pain: Systematic Review of the Evidence Pain Physician. 2002;5(4):405-418. McKenzie-Brown AM, Shah RV, Sehgal N, Everett Cr. A systematic review of sacroiliac joint interventions. Pain Physician. 2005;8:115-125. Nelemans PJ, de Bie RA, de Vet HCW, Sturmans F. Injection therapy for subacute and chronic benign low-back pain Cochrane Database of Systematic Reviews. 1999(4):Art. No.: CD001824. Rozenberg S. Glucocorticoid therapy in common lumbar spinal disorders. Rev Rhum - Engl Ed. 1998;65(11):649-655. Rozenberg S, Dubourg G, Khalifa P, Paolozzi L, Maheu E, Ravaud P. Efficacy of epidural steroids in low back pain and sciatica. A critical appraisal by a French Task Force of randomized trials. Critical Analysis Group of the French Society for Rheumatology. Rev Rhum - Engl Ed. 1999;66(2):79-85. Scheer S, Radack K, O'Brien D. Randomized controlled trials in industrial low back pain relating to return to work. Part 2. Discogenic low back pain. Arch Phys Med Rehabil. 1996;77(11):1189-1197. Stevens C, Dubois R, Larequi-Lauber T, Vader J. Efficacy of lumbar discectomy and percutaneous treatments for lumbar disc herniation. Soz Praventivmed. 1997;42(6):367-369.
168.
182.
169.
183.
170.
184.
171.
185.
172.
186.
173.
187.
174.
175.
188.
176.
189.
177.
190.
178.
179.
191.
180.

192. Turner JA, Loeser JD, Bell KG. Spinal cord stimulation for chronic low back pain: a systematic literature synthesis. Neurosurgery. 1995;37(6):1088-1095. van Tulder MW, Koes BW, Bouter LM. Conservative treatment of acute and chronic nonspecific low back pain: a systematic review of randomized controlled trials of the most common interventions. Spine. 1997;22(18):2128-2156. Watts RW, Silagy CA. A meta-analysis on the efficacy of epidural corticosteroids in the treatment of sciatica. Anaesth Intensive Care. 1995;23(5; Oct):564-569. Wetzel FT, Hassenbusch S, Oakley JC, Willis KD, Simpson RK, Ross EL. Treatment of chronic pain in failed back surgery patients with spinal cord stimulation: a review of current literature and proposal for future investigation. Neuromodulation. 2000;3: 59-74. Wetzel FT, McNally TA, Phillips FM. Intradiscal electrothermal therapy used to manage chronic discogenic low back pain: New directions and interventions. Spine. 2002;27:2621-2626. Yelland MJ, Del Mar C, Pirozzo S, Schoene ML, Vercoe P. Prolotherapy injections for chronic low-back pain Cochrane Database of Systematic Reviews. 2004(2):Art. No.: CD004059. Bogduk N, Karasek M. Two-year follow-up of a controlled trial of intradiscal electrothermal anuloplasty for chronic low back pain resulting from internal disc disruption. Spine J. 2002;2(5):343-350. Dallas T, Lin R, Wu W, Wolskee P. Epidural morphine and methylprednisolone for lowback pain. Anesthesiology. 1987;67(3): 408-411. Frost F. A controlled, double-blind comparison of mepivacaine injection versus saline injection for myofascial pain. Lancet. 1980;1(8):499-501. Glynn C, Dawson D, Sanders R. A doubleblind comparison between epidural morphine and epidural clonidine in patients with chronic noncancer pain. Pain. 1988;34:123-128. Gobel H, Heinze A, Reichel G, Hefter H, Benecke R, Dysport myofascial pain study g. Efficacy and safety of a single botulinum type A toxin complex treatment (Dysport) for the relief of upper back myofascial pain syndrome: results from a randomized doubleblind placebo-controlled multicentre study. Pain. 2006;125(1-2):82-88. Kapural L, Hayek S, Malak O, Arrigain S, Mehail N. Intradiscal thermal annuloplasty versus intradiscal radiofrequency ablation for the treatment of discogenic pain: a prospective matched control trial. Pain medicine. 2005;6(6):425-431. Lew HL, Lee EH, Castaneda A, Klima R, Date E. Therapeutic use of botulinum toxin type A in treating neck and upper-back pain of myofascial origin: a pilot study. Arch Phys Med Rehabil. 2008;89(1):75-80. Maugars Y, Mathis C, Berthelot JM, Charlier C, Prost A. Assessment of the efficacy of scroiliac corticosteroid injections in spondylarthropathies: a double-blind study. Rheumatology. 1996;35(8):767-770. Pneumaticos SG, Chatziioannou SN, Hipp JA, Moore WH, Esses SI. Low back pain: prediction of short-term outcome of facet joint injection with bone scintigraphy. Radiology. 2006;238(2):693-698. Subin B, Saleemi S, Morgan G, Zavisca F, Randall C. Treatment of Chronic Low Back Pain by Local Injection of Botulinum Toxin-A. Internet J Anesthesiol. 2003;6(2):1-8. Vad VB, Bhat AL, Lutz GE, Cammisa F. Transforaminal epidural steroid injections in lumbosacral radiculopathy: a prospective randomized study. Spine. 2002;27(1):11-16. Yates D. A comparison of the types of epidural injection commonly used in the treatment of low back pain and sciatica. Rheum Rehabil. 1978;17:181-186. Bono CM, Lee CK. Critical analysis of trends in fusion for degenerative disc disease over the past 20 years: influence of technique on fusion rate and clinical outcome. Spine. 2004;29(4):455-463. Boult M, Fraser R, Jones N, et al. Percutaneous endoscopic laser discectomy. Aust N Z J Surg. 2000;70(7):475-479. de Kleuver M, Oner F, Jacobs W. Total disc replacement for chronic low back pain: background and a systematic review of the literature. Eur Spine J. 2003;12(2):108-116. Freeman B, Davenport J. Total disc replacement in the lumbar spine: a systematic review of the literature. Eur Spine J. 2006;15 Suppl 3:S439-447. Geisler FH, Blumenthal SL, Guyer RD, et al. Neurological complications of lumbar artificial disc replacement and comparison of clinical
203.
193.
204.
194.
205.
195.
206.
196.
207.
197.
208.
198.
209.
210.
199.
200.
211.
212.
201.
213.
202.
214.

results with those related to lumbar arthrodesis in the literature: results of a multicenter, prospective, randomized investigational device exemption study of Charite intervertebral disc. Invited submission from the Joint Section Meeting on Disorders of the Spine and Peripheral Nerves, March 2004. J Neurosurg Spine. 2004;1(2):143-154. Ibrahim T, Tleyjeh M, Gabbar O. Surgical versus non-surgical treatment of chronic low back pain: a meta-analysis of randomised trials. Int Orthop. 2008;32:107-113. Kwon B, Hilibrand A, Malloy K, et al. A critical analysis of the literature regarding surgical approach and outcome for adult low-grade isthmic spondylolisthesis. J Spin Disord Tech. 2005;18(Supplement 1):S30-S40. Martin C, Gruszczynski A, Braunsfurth H, Fallatah S, O'Neil J, Wai E. The surgical management of degenerative lumbar spondylolisthesis: a systematic review. Spine. 2007;32(16):1791-1798. Mirza S, Deyo R. Systematic review of randomized trials comparing lumbar fusion surgery to nonoperative care for treatment of chronic back pain. Spine. 2007;32(7):816823. NICE (National Institute for Clinical Excellence). Interventional procedure overview of laser lumbar discectomy. IP Guidance Number IPG027. National Institute for Clinical Excellence; 2003. NICE (National Institute for Clinical Excellence). Interventional procedure overview of endoscopic laser foraminoplasty. IP Guidance Number IPG031. National Institute for Clinical Excellence; 2003. NICE (National Institute for Clinical Excellence). Prosthetic intervertebral disc replacement. IP Guidance Number IPG100.National Institute for Clinical Excellence; 2004. NICE Interventional Procedures Advisory Committee (IPAC). Interventional procedures overview of automated percutaneous mechanical lumbar discectomy: National Institute for Clinical Excellence; 2005. NICE (National Institute for Clinical Excellence). Interventional procedure overview of interspinous distraction procedures for spinal stenosis causing neurogenic claudication in the lumbar spine: National Institute for Clinical Excellence; 2005. 224. Resnick DK, Choudhri TF, Dailey AT, et al. Guidelines for the performance of fusion procedures for degenerative disease of the lumbar spine. Part 7: intractable low-back pain without stenosis or spondylolisthesis. J Neurosurg Spine. 2005;2(6):670-672. Resnick DK, Choudhri TF, Dailey AT, et al. Guidelines for the performance of fusion procedures for degenerative disease of the lumbar spine. Part 11: interbody techniques for lumbar fusion. J Neurosurg Spine. 2005;2:692-699. Resnick DK, Choudhri TF, Dailey AT, et al. Guidelines for the performance of fusion procedures for degenerative disease of the lumbar spine. Part 12: pedicle screw fixation as an adjunct to posterolateral fusion for lowback pain. J Neurosurg Spine. 2005;2(6): 700-706. Resnick DK, Choudhri TF, Dailey AT, et al. Guidelines for the performance of fusion procedures for degenerative disease of the lumbar spine. Part 8: lumbar fusion for disc herniation and radiculopathy. J Neurosurg Spine. 2005;2:673-678. Resnick DK, Choudhri TF, Dailey AT, et al. Guidelines for the performance of fusion procedures for degenerative disease of the lumbar spine. Part 9: fusion in patients with stenosis and spondylolisthesis. J Neurosurg Spine. 2005;2(679-685). Resnick DK, Choudhri TF, Dailey AT, et al. Guidelines for the performance of fusion procedures for degenerative disease of the lumbar spine. Part 10: fusion following decompression in patients with stenosis without spondylolisthesis. J Neurosurg Spine. 2005;2:686-691. Fenton JJ, Mirza SK, Lahad A, Stern BD, Deyo RA. Variation in reported safety of lumbar interbody fusion. Spine. 2007;32(4):471-480. Anderson PA, Tribus CB, Kitchel SH. Treatment of neurogenic claudication by interspinous decompression: application of the X STOP device in patients with lumbar degenerative spondylolisthesis. J Neurosurg Spine. 2006;4(6):463-471. Fernandez-Fairen M, Sala P, Ramirez H, Gil J. A prospective randomized study of unilateral versus bilateral instrumented posterolateral lumbar fusion in degenerative spondylolisthesis. Spine. 2007;32(4): 395-401.
225.
215.
216.
226.
217.
227.
218.
228.
219.
229.
220.
221.
230.
222.
231.
223.
232.

233. Inamdar DN, Alagappan M, Shyam L, Devadoss S, Devadoss A. Posterior lumbar interbody fusion versus intertransverse fusion in the treatment of lumbar spondylolisthesis. J Orthop Surg. 2006;14(1):21-26. Katayama Y, Matsuyama Y, Yoshihara H, et al. Comparison of surgical outcomes between macro discectomy and micro discectomy for lumbar disc herniation: a prospective randomized study with surgery performed by the same spine surgeon. J Spinal Disord Tech. 2006;19(5):344-347. Kim K-T, Lee S-H, Lee Y-H, Bae S-C, Suk KS. Clinical outcomes of 3 fusion methods through the posterior approach in the lumbar spine. Spine. 20 2006;31(12):1351-1357; discussion 1358. Malmivaara A, Slatis P, Heliovaara M, et al. Surgical or nonoperative treatment for lumbar spinal stenosis?: A randomized controlled trial. Spine. 2007;32(1):1-8. Osterman H, Seitsalo S, Karppinen J, Malmivaara A. Effectiveness of microdiscectomy for lumbar disc herniation: a randomized controlled trial with 2 years of follow-up. Spine. 2006 31(21):2409-2414. Peul WC, van Houwelingen HC, van den Hout WB, et al. Surgery versus prolonged conservative treatment for sciatica. N Engl J Med. 2007;356(22):2245-2256. Peul W, van den Hout W, Brand R, Thomeer R, Koes B. Prolonged conservative care versus early surgery in patients with sciatica caused by lumbar disc herniation: two year results of a randomised controlled trial. BMJ. 2008;336:1355-1358. Ruetten S, Komp M, Merk H, Godolias G. Full-endoscopic interlaminar and transforaminal lumbar discectomy versus conventional microsurgical technique. Spine. 2008;33:931-939. Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical versus nonsurgical treatment for lumbar degenerative spondylolisthesis. N Engl J Med. 2007;356(22):2257-2270. Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical versus nonoperative treatment for lumbar disc herniation. Four-year results for the Spine Patient Outcomes Research Trial. Spine. 2008;33:2789-2800. Amundsen T, Weber H, Nordal H, Magnaes B, Abdelnoor M, Lilleas F. Lumbar Spinal Stenosis: Conservative or Surgical Management? : A Prospective 10-Year Study. Spine. 2000;25(11):1424-1436. Brox J, Reikeras O, Nygaard O, et al. Lumbar instrumented fusion compared with cognitive intervention and exercises in patients with chronic back pain after previous surgery for disc herniation: A prospective randomized controlled study. Pain. 2006;122(1-2): 145-155. Brox JI, Sorensen R, Friis A, et al. Randomized clinical trial of lumbar instrumented fusion and cognitive intervention and exercises in patients with chronic low back pain and disc degeneration. Spine. 2003;28(17):1913-1921. Fairbank J, Frost H, Wilson-MacDonald J, et al. Randomised controlled trial to compare surgical stabilisation of the lumbar spine with an intensive rehabilitation programme for patients with chronic low back pain: the MRC spine stabilisation trial. BMJ. 2005;330(7502):1233. Fritzell P, Hagg O, Wessberg P, Nordwall A, Swedish Lumbar Spine Study G. 2001 Volvo Award Winner in Clinical Studies: Lumbar Fusion Versus Nonsurgical Treatment for Chronic Low Back Pain: A Multicenter Randomized Controlled Trial From the Swedish Lumbar Spine Study Group. Spine. 2001;26(23):2521-2532. Weber H. Lumbar disc herniation. A controlled, prospective study with ten years of observation. Spine. 1983;8(2):131-140. Weinstein JN, Tosteson TD, Lurie JD, et al. Surgical vs Nonoperative Treatment for Lumbar Disk Herniation: The Spine Patient Outcomes Research Trial (SPORT): A Randomized Trial. JAMA. 2006;296(20):2441-2450. Weinstein J, Tosteson T, Lurie J, et al. Surgical versus nonsurgical therapy for lumbar spinal stenosis. N Engl J Med. 2008;358(8):794-810. Zucherman J, Hsu K, Hartjen C, et al. A prospective randomized multi-center study for the treatment of lumbar spinal stenosis with the X STOP interspinous implant: 1-year results. Eur Spine J. 2004;13(1):22-31. Blumenthal S, McAfee PC, Guyer RD, et al. A prospective, randomized, multicenter Food and Drug Administration investigational device exemptions study of lumbar total disc replacement with the CHARITE artificial disc
244.
234.
245.
235.
246.
236.
237.
247.
238.
239.
248.
249.
240.
250.
241.
251.
242.
252.
243.

versus lumbar fusion: Part I: Evaluation of clinical outcomes. Spine. 2005;30:1565-1575. Zigler J, Delamarter R, Spivak JM, et al. Results of the prospective, randomized, multicenter Food and Drug Administration investigational device exemption study of the ProDisc-L total disc replacement versus circumferential fusion for the treatment of 1level degenerative disc disease. Spine. 2007;32(11):1155-1162; discussion 1163. Aalto TJ, Malmivaara A, Kovacs F, et al. Preoperative predictors for postoperative clinical outcome in lumbar spinal stenosis. Spine. 2006;31:E648-E663. Gibson JNA, Waddell G, Grant IC. Surgery for degenerative lumbar spondylosis. Cochrane Database of Systematic Reviews. 2000(3):Art. No.: CD001352. Hoffman RM, Wheeler KJ, Deyo RA. Surgery for herniated lumbar discs: a literature synthesis. J Gen Intern Med. 1993;8: 487-496. Mardjetko SM, Connolly PJ, Shott S. Degenerative lumbar spondylolisthesis. A meta-analysis of literature 1970-1993. Spine. 1994;19(20S):2256S-2265S. Niggemeyer O, Strauss J, Schulitz K. Comparison of surgical procedures for degenerative lumbar stenosis: a metaanalysis of the literature from 1975 to 1995. Eur Spine J. 1997;6(6):423-429. 1997;6(6):423-429. Turner JA, Ersek M, Herron L, Deyo R. Surgery for lumbar spinal stenosis. Attempted meta-analysis of the literature. Spine. 1992;17(1):1-8. Turner JA, Ersek M, Herron L, et al. Patient outcomes after lumbar spinal fusions. JAMA. 1992;268(7):907-911. Turner JA, Herron L, Deyo RA. Meta-analysis of the results of lumbar spine fusion. Acta Orthop Scand Suppl. 1993;251:120-122. Greenfield K, Nelson RJ, Findlay GD, Egger M, Sanford E. Microdiscectomy and conservative treatment for lumbar disc herniation with back pain and sciatica: a randomised clnical trial. Abstract from the Proceedings of the NASS Annual Meeting, Oct 31-Nov 3, 2001. The Spine Journal. 2002;2(2 (Suppl 1)):47. Hallett A, Huntley JS, Gibson JNA. Foraminal stenosis and single-level degenerative disc disease. A randomized controlled trial comparing decompression with decompression and instrumented fusion. Spine. 2007;32:1375-1380. Livesey JPS. Laser discectomy versus lumbar epidural steroid injection: a randomised comparative study of two treatments for sciatica J Bone Joint Surg. 2000;82(Suppl 1):74. Sasso RC, Foulk DM, Hahn M. Prospective, randomized trial of metal-on-metal artificial lumbar disc replacement: initial results for treatment of discogenic pain. Spine. 2008;33(2):123-131. Deyo R, Diehl A. Cancer as a cause of back pain: frequency, clinical presentation, and diagnostic strategies. J Gen Intern Med. 1988;3(3):230-238. Underwood MR, Dawes P. Inflammatory back pain in primary care. Br J Rheumatol. 1995;34:1074-1077. Jarvik JG, Deyo RA. Diagnostic evaluation of low back pain with emphasis on imaging. Ann Intern Med. 2002;137:586-597. de Graaf I. Diagnosis of lumbar spinal stenosis. A systematic reivew of the accuracy of diagnostic tests. Spine. 2006;31(10): 1168-1176. Devill W, van der Windt D, Dzaferagic A, Bezemer P, Bouter L. The test of Lasegue: systematic review of the accuracy of diagnosing herniated discs. Spine. 2000;25(9):1140-1147. van den Hoogen HM, Koes BW, van Eijk JT, Bouter LM. On the accuracy of history, physical examination, and erythrocyte sedimentation rate in diagnosing low back pain in general practice. A criteria-based review of the literature. Spine. 1995;20(3):318-327. Vroomen PC, de Krom MC, Knottnerus JA. Diagnostic value of history and physical examination in patients suspected of sciatica due to disc herniation: a systematic review. J Neurol. 1999;246:899-906. Deyo RA, Rainville J, Kent DL. What can the history and physical examination tell us about low back pain? JAMA. 1992;268(6):760-765. Rebain R, Baxter D, McDonough S. A systematic review of the passive straight leg raising test as a diagnostic aid for low back pain. Spine. 2002;27(17):E388-E395. Jarvik JG. Imaging of adults with low back pain in the primary care setting. Neuroimag Clin N Am. 2003;13:293-305.
253.
264.
265.
254.
266.
255.
267.
256.
268.
257.
269.
258.
270.
271.
259.
260.
272.
261.
262.
273.
274.
263.
275.

276. Rudwaleit M, Metter A, Listing J, Sieper J, Braun J. Inflammatory back pain in ankylosing spondylitis. A reassessment of the clincial histroy for application as classification and diagnostic criteria. Arth Rheum. 2006;54:569-578. Rudwaleit M, Khan M. The challenge of diagnosis and classification in early ankylosing spondylitis: Do we need new criteria? Arth Rheum. 2005;52(4):1000-1008. Hudgins WR, Hudgins WR. Computer-aided diagnosis of lumbar disc herniation. Spine. 1983;8(6):604-615. Hirsch C, Nachemson A. The reliability of lumbar disc surgery. Clin Orthop. 1963;29:189-195. Jensen OH, Schmidt-Olsen S. A new functional test in the diagnostic evaluation of neurogenic intermittent claudication. Clin Rheumatol. 1989;8:363-367. Roach KE, Brown MD, Albin RD, Delaney KG, Lipprandi HM, Rangelli D. The sensitivity and specificity of pain response to activity and position in categorizing patients with low back pain. Physical Therapy. 1997;77(7): 730-738. Katz JN, Dalgas M, Stucki G, et al. Degenerative lumbar spinal stenosis: diagnostic value of the history and physical examination. Arth Rheum. 1995;38(9):12361241. Fritz JM, Erhard RE, Delitto A, Welch WC, Nowakowski PE. Preliminary results of the use of a two-stage treadmill test as a clinical diagnostic tool in the differential diagnosis of lumbar spinal stenosis. J Spinal Disord. 1997;10:410-416. Crook J, Milner R, Schultz IZ, Stringer B. Determinants of Occupational Disability Following a Low Back Injury: A Critical Review of the Literature. J Occup Rehabil. 2002;12(4):277-295. Dionne CE, Von Korff M, Koepsell TD, Deyo RA, Barlow WE, Checkoway H. Formal education and back pain: a review. J Epidemiol Comm Health. 2001;55(7):455468. Fayad F, Lefevre-Colau MM, Poiraudeau S, et al. Chronicity, recuurence, and return to work in low back pain: common prognostic factors. Annales de readaption et de medecine physique. 2004;47:179-189. Hartvigsen J. Psychococial factors at work in relation to low back pain and consequences of low back pain; a systematic, critical review of prospective cohort studies. Occup Environ Med. 2004;61(2):1-10. Kuijer W, Groothoff JW, Brouwer S, Geertzen JHB, Dijkstra PU. Prediction of sickness absence in patients with chronic low back pain: a systematic review. J Occup Rehabil. 2006;16(3):430-458. Linton SJ. A review of psychological risk factors in back and neck pain. Spine. 2000;25:1148-1156. McIntosh G, Frank JW, Hogg-Johnson S, Hall H, Bombardier C. Low back pain prognosis: structured review of the literature. J Occup Rehab. 2000;10(2):101-115. Pincus T, Burton AK, Vogel S, Field AP. A systematic review of psychological factors as predictors of chronicity/disability in prospective cohorts of low back pain. Spine. 2002;27(5):E109-E120. Pincus T, Vogel S, Burton AK, Santos R, Field AP. Fear avoidance and prognosis in back pain. A systematic review and synthesis of current evidence. Arth Rheum. 2006;54(12):3999-4010. Shaw WS, Pransky G, Fitzgerald TE. Early prognosis for low back disability: intervention strategies for health care providers. Disability & Rehab. 2001;23(18):815-828. Steenstra I, Verbeek J, Heymans M, Bongers P. Prognostic factors for duration of sick leave in patients sick listed with acute low back pain: a systematic review of the literature. Occup Environ Med. 2005;62: 851-860. Truchon M, Fillion L. Biopsychosocial determinants of chronic disability and lowback pain: a review. J Occup Rehab. 2000;10(2):117-142. Borge JA, Leboef-Yde C, Lothe J. Prognostic vlaues of physical examination findings in patients with chronic low back pain treated conservatively: a systematic literature review. J Manipulative Physiol Ther. 2001;24(4): 292-295. Hazard RG, Haugh LD, Reid S, Preble JB, MacDonald L. Early prediction of chronic disability after occupational low back injury. Spine. 1996;21:945-951. Fritz JM, Wainner RS, Hicks GE. The use of nonorganic signs and symptoms as a screening tool for return-to-work in patients with acute low back pain. Spine. 2000;25:1925-1931.
288.
277.
289.
278.
290.
279.
291.
280.
281.
292.
293.
282.
294.
283.
295.
284.
296.
285.
297.
286.
298.
287.

299. Karjalainen K, Malmivaara A, van Tulder M, et al. Multidisciplinary biopsychosocial rehabilitation for subacute low back pain among working age adults.[update of Cochrane Database Syst Rev. 2000;(3):CD002193]. Cochrane Database of Systematic Reviews. 2003(2):Art. No.: CD002194. Karjalainen K, Malmivaara M, van Tulder M, et al. Multidisciplinary biopsychosocial rehabilitation for subacute low-back pain among working age adults. Spine. 2001;26(3):262-269. Ostelo RWJG, van Tulder MW, Vlaeyen JWS, Linton SJ, Morley SJ, Assendelft WJJ. Behavioural treatment for chronic low-back pain Cochrane Database of Systematic Reviews. 2005(1):Art. No.: CD002014. Schonstein E, Kenny D, Keating J, Koes B, Herbert RD. Physical conditioning programs for workers with back and neck pain: a Cochrane systematic review. Spine. 2003;28(19):E391-395. Schonstein E, Kenny DT, Keating J, Koes BW. Work conditioning, work hardening and functional restoration for workers with back and neck pain. Cochrane Database of Systematic Reviews. 2003(3):Art. No.: CD001822. Hay EM, Lewis M, Vohora K, et al. Comparison of physical treatments versus a brief pain-management programme for back pain in primary care: a randomised clinical trial in physiotherapy practice. Lancet. 2005;365:2024-2030. Jellema P, van der Windt D, van der Horst H, Twisk J, Stalman W, Bouter L. Should treatment of (sub)acute low back pain be aimed at psychosocial prognostic factors? Cluster randomised clinical trial in general practice. BMJ. 2005;331(7508):84. Gatchel RJ, Polatin PB, Noe C, Gardea M, Pulliam C, Thompson J. Treatment- and costeffectiveness of early intervention for acute low-back pain patients: a one-year prospective study. J Occup Rehabil. 2003;13(1):1-9. Von Korff M, Balderson BHK, Saunders K, et al. A trial of an activating intervention for chronic back pain in primary care and physical therapy settings. Pain. 2005;113(3):323-330. George SZ, Fritz JM, Bialosky JE, Donald DA. The effect of a fear-avoidance-based physical therapy intervention for patients with acute low back pain: results of a randomized clinical trial. Spine. 2003;28(23):2551-2560. Linton SJ, Andersson T. Can chronic disability be prevented? A randomized trial of a cognitive-behavior intervention and two forms of information for patients with spinal pain. Spine. 2000;25:2825-2831. Linton SJ, Boersma K, Jansson M, Svard L, Botvalde M. The effects of cognitivebehavioral and physical therapy preventive interventions on pain-related sick leave: a randomized controlled trial. Clin J Pain. 2005;21(2):109-119. Jellema P, van der Windet DA, van der Horst HE, Blankenstein AH, Bouter LM, Stalman WA. Why is a treatment aimed at psychosocial factors not effective in patients with (sub)acute low back pain? Pain. 2005;118:350-359. Jellema P, van der Horst HE, Vlaeyeyn JWS, Stalman WAB, Bouter LM, van der Windt DAWM. Predictors of outcome in patients with (sub)acute low back pain differ across treatment groups. Spine. 2006;31:1699-1705. van Tulder MW, Assendelft WJ, Koes BW, Bouter LM. Spinal radiographic findings and nonspecific low back pain: a systematic review of observational studies. Spine. 1997;22(4):427-434. Lurie JD, Birkmeyer NJ, Weinstein JN. Rates of advanced spinal imaging and spine surgery. Spine. 2003;28(6):616-620. Sox HJ, Margulies I, Sox C. Psychologically mediated effects of diagnostic tests. Ann Intern Med. 1981;95:680-685. Boos N, Lander PH. Clinical efficacy of imaging modalities in the diagnosis of lowback pain disorders. Eur Spine J. 1996;5(1):2-22. Littenberg B, Siegel A, Tosteson AN, Mead T. Clinical efficacy of SPECT bone imaging for low back pain. J Nucl Med. 1995;36(9): 1707-1713. Joines JD, McNutt RA, Carey TS, Deyo RA, Rouhani R. Finding cancer in primary care outpatients with low back pain: a comparison of diagnostic strategies. J Gen Intern Med. 2001;16(1):14-23. Hollingworth W, Gray DT, Martin BI, Sullivan SD, Deyo RA, Jarvik JG. Rapid magnetic resonance imaging for diagnosing cancerrelated low back pain: A cost-effectiveness
309.
300.
310.
301.
311.
302.
312.
303.
313.
304.
314.
315.
305.
316.
317.
306.
318.
307.
319.
308.

analysis. J Gen Intern Med. 2003;18: 303-312. Hoffman RM, Kent DL, Deyo RA. Diagnostic accuracy and clinical utility of thermography for lumbar radiculopathy. A meta-analysis. Spine. 1991;16(6):623-628. Pullman SL, Goodin DS, Marquinez A, Tabbal S, Rubin M. Clinical utility of surface EMG: Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology. 2000;55(2):171-177. Geisser ME, Ranavaya M, Haig AJ, et al. A meta-analytic review of surface electromyography among persons with low back pain and normal, healthy controls. J Pain. 2005;6(11):711-726. Mohseni-Bandpei M. Application of surface electromyography in the assessment of low back pain: a literature review. Phys Ther Rev. 2000;5(2):93-105. Kardaun JW, Schipper J, Braakman R. CT, myelography, and phlebography in the detection of lumbar disk herniation: an analysis of the literature. Am J Neuroradiol. 1989;10(5):1111-1122. Kent DL, Haynor DR, Larson EB, Deyo RA. Diagnosis of lumbar spinal stenosis in adults: a metaanalysis of the accuracy of CT, MR, and myelography. Am J Roentgenol. 1992;158(5):1135-1144. Hanly JG, Barnes DC, Mitchell MJ, MacMillan L, Docherty P. Single photon emission computed tomography in the diagnosis of inflammatory spondyloarthropathies. J Rheumatol. 1993;20:2062-2068. Marc V, Dromer C, Le Guennec P, Manelfe C, Fournie B. Magnetic resonance imaging and axial involvement in spondylarthropathies. Delineation of the spinal entheses. Rev Rhum - Engl Ed. 1997;64:465-473. Khan MA, Khan MK. Diagnostic value of HLA-B27 testing in ankylosing spondylitis and Reiters syndrome. Ann Intern Med. 1982;96:70-76. Fisher M. Electrophysiology of radiculopathies. Clin Neurophysiol. 2002;113(3):317-335. Deyo RA, Diehl AK, Rosenthal M. Reducing roentgenography use. Can patient expectations be altered? Arch Intern Med. 1987;147(1):141-145. 331. Djais N, Kalim H. The role of lumbar spien radiography in the outcomes of patients with simple acute low back pain. J Rheumatol. 2005;8:45-50. Kendrick D, Fielding K, Bentley E, Kerslake R, Miller P, Pringle M. Radiography of the lumbar spine in primary care patients with low back pain: randomised controlled trial. BMJ. 2001;322:400-405. Kerry S, Hilton S, Dundas D, Rink E, Oakeshott P. Radiography for low back pain: a randomised controlled trial and observational study in primary care. Br J Gen Pract. 2002;52(479):469-474. Gilbert FJ, Grant AM, Gillan MGC, et al. Low Back Pain: Influence of Early MR Imaging or CT on Treatment and Outcome - Multicenter Randomized Trial. Radiology. 2004;231(2):343-351. Modic MT, Obuchowski NA, Ross JS, et al. Acute Low Back Pain and Radiculopathy: MR Imaging Findings and Their Prognostic Role and Effect on Outcome. Radiology. 2005;237(2):597-604. Ash LM, MOdic MT, Obuchowski NA, Ross JS, Brant-Zawadzki MN, Grooff PN. Effects of diagnostic information, per se, on patient outcomes in acute radiculopathy and low back pain. Am J Neuroradiol. 2008;29: 1098-1103. Jarvik JG, Hollingworth W, Martin B, et al. Rapid magnetic resonance imaging vs. radiographs for patients with low back pain: a randomized controlled trial. JAMA. 2003;289(21):2810-2818. Jarvik JG, Maravilla KR, Haynor DR, Levitz M, Deyo RA. Rapid MR imaging versus plain radiography in patients with low back pain: initial results of a randomized study. Radiology. 1997;204(2):447-454. Miller P, Kendrick D, Bentley E, Fielding K. Cost-effectiveness of lumbar spine radiography in primary care patients with low back pain. Spine. 2002;27(20):2291-2297. Liang M, Komaroff AL. Roentgenograms in primary care patients with acute low back pain. A cost-effectiveness analysis. Arch Intern Med. 1982;142:1108-1112. Hagen KB, Hilde G, Jamtvedt G. The Cochrane review of bed rest for acute low back pain and sciatica. Spine. 2000;25(22):2932-2939. Allen C, Glasziou P, Del Mar C. Bed rest: a potentially harmful treatment needing more
320.
332.
321.
333.
322.
334.
323.
335.
324.
336.
325.
337.
326.
338.
327.
339.
328.
340.
329.
341.
330.
342.

careful evaluation. Lancet. 1999;354(9186):1229-1233. Koes BW. Efficacy of bed rest and orthoses for low back pain: A review of randomized clinical trials. Eur J Phys Med Rehabil. 1994;4:86-93. Maher C, Latimer J, Refshauge K. Prescription of activity for low back pain: What works? Aust J Physiother. 1999;45(2):121-132. Scheer SJ, Radack K, O'Brien D, Jr. Randomized controlled trials in industrial low back pain relating to return to work. Part 1. Acute interventions. Arch Phys Med Rehabil. 1995;76(10):966-973. van der Weide WE, Verbeek JH, van Tulder MW. Vocational outcome of intervention for low-back pain. Scand J Work Environ Health. 1997;23(3):165-178. Waddell G, Feder G, Lewis M. Systematic reviews of bed rest and advice to stay active for acute low back pain. Br J Gen Pract. 1997;47(423; Oct):647-652. Coomes NE. A comparison between epidural anaesthesia and bed rest in sciatica. BMJ. 1961;1:20-24. Malmivaara A, Hakkinen U, Aro T, et al. The treatment of acute low back pain--bed rest, exercises, or ordinary activity? N Engl J Med. 1995;332:351-355. Rozenberg S, Delval C, Rezvani Y, al e. Bed rest or normal activity for patients with acute low back pain: a randomized controlled trial. Spine. 2002;27:1487-1493. Hofstee DJ, Gijtenbeek JMM, Hoogland PH, et al. Westeinde Sciatica Trial: randomized controlled study of bed rest and physiotherapy for acute sciatica. J Neurosurg: Spine. 2002;96:45-49. Vroomen PCAJ, Marc CTFM, Wilmink JT, Kester ADM, Knottnerus JA. Lack of effectiveness of bed rest for sciatica. N Engl J Med. 1999;340:418-423. Wiesel S, Cuckler J, Deluca F, Jones F, Zeide M, Rothman R. Acute low back pain: an objective analysis of conservative therapy. Spine. 1980;5(4):324-330. Wilkinson MJ. Does 48 hours' bed rest influence the outcome of acute low back pain? Br J Gen Pract. 1995;45:481-484. Gilbert JR, Taylor DW, Hildebrand A, Evans C. Clinical trial of common treatments for low back pain in family practice. Br Med J Clin Res Ed. 1985;291:791-794. 356. Postacchini F, Facchini M, Palieri P. Efficacy of various forms of conservative treatment in low back pain: A comparative study. NeuroOrthopedics. 1988;6:28-35. Szpalski M, Hayez JP. How many days of bed rest for acute low back pain? Objective assessment of trunk function. Eur Spine J. 1992;1:29-31. Deyo RA, Diehl AK, Rosenthal M. How many days of bed rest for acute low back pain? A randomized clinical trial. N Engl J Med. 1986;315:1064-1070. Hagen K, Hilde G, Jamtvedt G, Winnem M. The Cochrane review of advice to stay active as a single treatment for low back pain and sciatica. Spine. 2002;27(16):1736-1741. Hilde G, Hagen KB, Jamtvedt G, Winnem M. Advice to stay active as a single treatment for low-back pain and sciatica. Cochrane Database of Systematic Reviews. 2002;2:Art. No.: CD003632. Frost H, Lamb SE, Doll HA, Carver PT, Stewart-Brown S. Randomised controlled trial of physiotherapy compared with advice for low back pain. Br Med J. 2004;329(7468):708-711. Pengel LHM, Refshauge KM, Maher CG, Nicholas MK, Herbert RD, McNair P. Physiotherapist-directed exercise, advice, or both for subacute low back pain. A randomized trial. Ann Intern Med. 2007;146:787-796. Little P, Roberts L, Blowers H, et al. Should we give detailed advice and information booklets to patients with back pain? A randomized controlled factorial trial of a selfmanagement booklet and doctor advice to take exercise for back pain. Spine. 2001;26(19):2065-2072. Stankovic R, Johnell O. Conservative treatment of acute low back pain. A 5-year follow-up study of two methods of treatment. Spine. 1995;20:469-472. Stankovic R, Johnell O. Conservative treatment of acute low-back pain. A prospective randomized trial: McKenzie method of treatment versus patient education in "mini back school". Spine. 1990;15(2): 120-123. Burton, Waddell GT. Information and advice to patients with back pain does have a positive effect: an RCT of an educational booklet in primary care. J Bone Joint Surg Br. 1999;82(Suppl 1):3.
343.
357.
344.
358.
345.
359.
346.
360.
347.
361.
348.
349.
362.
350.
363.
351.
352.
364.
353.
365.
354.
366.
355.

367. Cherkin DC, Deyo RA, Battie M, Street J, Barlow W. A comparison of physical therapy, chiropractic manipulation, and provision of an educational booklet for the treatment of patients with low back pain. New Engl J Med. 1998;339(15):1021-1029. Cherkin DC, Deyo RA, Street JH, Hunt M, Barlow W. Pitfalls of patient education. Limited success of a program for back pain in primary care. Spine. 1996;21(3):345-355. Cherkin DC, Eisenberg D, Sherman KJ, et al. Randomized trial comparing traditional Chinese medical acupuncture, therapeutic massage, and self-care education for chronic low back pain. Arch Intern Med. 2001;161(8):1081-1088. Hazard RG, Reid S, Haugh LD, McFarlane G. A controlled trial of an educational pamphlet to prevent disability after occupational low back injury. Spine. 2000;25(11):1419-1423. Sherman KJ, Cherkin DC, Erro J, Miglioretti DL, Deyo RA. Comparing Yoga, Exercise, and a Self-Care Book for Chronic Low Back Pain: A Randomized, Controlled Trial. Ann Intern Med. December 20, 2005 2005;143(12):849-856. Roberts L, Little P, Chapman J, Cantrell T, Pickering R, Langridge J. The back home trial: general practitioner-supported leaflets may change back pain behavior. Spine. 2002;27(17):1821-1828. Roland M, Dixon M. Randomized controlled trial of an educational booklet for patients presenting with back pain in general practice. Brit J Gen Pract. 1989;39(323):244-246. Burton AK, Waddell G, Tillotson KM, Summerton N. Information and advice to patients with back pain can have a positive effect. A randomized controlled trial of a novel educational booklet in primary care. Spine. 1999;24(23):2484-2491. Lorig K, Laurent D, Deyo R. Can a back pain e-mail discussion group improve health status and lower health care costs? Arch Intern Med. 2002;162(7):792-796. Miller JS, Stanley I, Moore K. Videotaped exercise instruction: A randomised controlled trial in musculoskeletal physiotherapy. Physiother Theory Pract. 2004;20(3): 145-154. Snook SH, Webster BS, McGorry RW. The reduction of chronic, nonspecific low back pain through the control of early morning lumbar flexion: 3-year follow-up. J Occup Rehabil. 2002;12(1):13-19. Snook SH, Webster BS, McGorry RW, Fogleman MT, McCann KB. The reduction of chronic nonspecific low back pain through the control of early morning lumbar flexion. Spine. 1998;23(23):2601-2607. Haas M, Groupp E, Muench J, et al. Chronic disease self-management progrm for low back pain in the elderly. J Manipulative Physiol Ther. 2005;28(4):228-237. Von Korff M, Moore JE, Lorig K, et al. A randomized trial of a lay person-led selfmanagement group intervention for back pain patients in primary care. Spine. 1998;23(23):2608-2615. Strong LLMPH, Von Korff MS, Saunders KJD, Moore JEP. Cost-Effectiveness of Two Self-Care Interventions to Reduce Disability Associated With Back Pain. Spine. 2006;31(15):1639-1645. Deyo RA, Cherkin DC, Weinstein J, Howe J, Ciol M, Mulley AG. Involving patients in clinical decisions: impact of an interactive video program on use of back surgery. Medical care. 2000;38(9):959-969. Phelan EA, Deyo RA, Cherkin DC, et al. Helping patients decide about back surgery: a randomized trial of an interactive video program. Spine. 2001;26(2):206-211. van Tulder MW, Jellema P, van Poppel MNM, Nachemson AL, Bouter LM. Lumbar supports for prevention and treatment of low-back pain. Cochrane Database of Systematic Reviews. 2000(3):Art. No.: CD001823. Jellema P, van Tulder M, van Poppel MNM, Nachemson AL, Bouter LM. Lumbar supports for prevention and treatment of low back pain. Spine. 2001;26(4):377-386. Scheer S, Watanabe T, Radack K. Randomized controlled trials in industrial low back pain. Part 3. Subacute/chronic pain interventions. Arch Phys Med Rehabil. 1997;78(4):414-423. van Poppel MN, de Looze MP, Koes BW, Smid T, Bouter LM. Mechanisms of action of lumbar supports: a systematic review. Spine. 2000;25(16):2103-2113. Penrose KW, Chook K, Stump JL. Acute and chronic effects of pneumatic lumbar support on muscular strength, flexibility, and functional impairment index. Sports Training Med Rehab. 1991;2:121-129.
378.
368.
379.
369.
380.
370.
381.
371.
382.
372.
383.
384.
373.
374.
385.
386.
375.
387.
376.
388.
377.

389. Coxhead CE, Inskip H, Meade TW, North WR, Troup JD. Multi-centre trial of physiotherapy in the management of sciatic symptoms. Lancet. 1981;1(8229):1065-1068. Hsieh CY, Phillips RB, Adams AH, Pope MH. Functional outcomes of low back pain: Comparison of four treatment groups in a randomized controlled trial. J Manipulative Physiol Ther. 1992;15(1):4-9. Pope MH, Phillips RB, Haugh LD, Hsieh CY, MacDonald L, Haldeman S. A prospective randomized three-week trial of spinal manipulation, transcutaneous muscle stimulation, massage and corset in the treatment of subacute low back pain. Spine. 1994;19:2571-2577. Valle-Jones J, Walsh H, O'Hara J, O'Hara H, Davey N, Hopkin-Richards H. Controlled trial of a back support ('Lumbotrain') in patients with non-specific low back pain. Curr Med Res Opin. 1992;12(9):604-13. 1992;12(9):604-613. Doran DM, Newell DJ, Doran DM, Newell DJ. Manipulation in treatment of low back pain: a multicentre study. Br Med J. 1975;2(5964):161-164. Million R, Nilsen K, Jayson M, Baker R. Evaluation of low back pain and assessment of lumbar corsets with and without back supports. Ann Rheum Dis. 1981;40(5): 449-454. Kovacs FM, Abraira V, Pena A, et al. Effect of firmness of mattress on chronic non-specific low-back pain: Randomised, double-blind, controlled, multicentre trial. Lancet. 2003;362(9396):1599-1604. Atherton J, Clarke AK, Harrison RA, Maddison MC. Low back pain - the use of the isometric mattress. Br J Occup Ther Vol. 1983;46:133-134. Garfin SR, Pye SA. Bed design and its effect on chronic low back pain--a limited controlled trial. Pain. 1981;10(1):87-91. French S, Cameron M, Walker B, Reggars J, Esterman A. Superficial heat or cold for low back pain. Cochrane Database of Systematic Reviews. 2006(1):Art. No.: CD004750. Philadelphia Panel. Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for low back pain. Physical Therapy. 2001;81(10):1641-1674. Nadler SF, Steiner DJ, Erasala GN, Hengehold DA, Abeln SB, Weingand KW. Continuous low-level heatwrap therapy for treating acute nonspecific low back pain. Arch Phys Med Rehabil. 2003;84(3):329-334. Nadler SF, Steiner DJ, Erasala GN, et al. Continuous low-level heat wrap therapy provides more efficacy than Ibuprofen and acetaminophen for acute low back pain. Spine. May 15 2002;27(10):1012-1017. Nadler SF, Steiner DJ, Petty SR, Erasala GN, Hengehold DA, Weingand KW. Overnight use of continuous low-level heatwrap therapy for relief of low back pain. Arch Phys Med Rehabil. 2003;84(3):335-342. Nuhr M, Hoerauf K, Bertalanffy A, et al. Active warming during emergency transport relieves acute low back pain. Spine. 2004;29(14):1499-1503. Mayer JM, Ralph L, Look M, et al. Treating acute low back pain with continuous low-level heat wrap therapy and/or exercise: A randomized controlled trial. Spine Journal. 2005;5(4):395-403. Melzack R, Jeans M, Stratford J, Monks R. Ice massage and transcutaneous electrical stimulation: comparison of treatment for lowback pain. Pain 1980;9(2):209-217. Landen BR. Heat or cold for the relief of low back pain? Phys Ther. 1967;47(12): 1126-1128. Roberts D, Walls C, Carlile J, Wheaton C. Chapter 13 - Relief of chronic low back pain: heat versus cold. 2nd edition ed: Baltimore: Urban & Schwarzenberg, 1992. Lloyd A, Scott DA, Akehurst RL, Lurie-Luke E, Jessen G. Cost-effectiveness of low-level heat wrap therapy for low back pain. Value in Health. 2004;7(4):413-422. Crofford L. Rational use of analgesic and antiinflammatory drugs. N Engl J Med. 2001;345:1844-1846. Graham G, Graham R, Day R. Comparative analgesia, cardiovascular and renal effects of celecoxib, rofecoxib and acetaminophen (paracetamol). Curr Pharm Des. 2002;8(12):1063-1075. Schnitzer TJ, Ferraro A, Hunsche E, Kong SX. A comprehensive review of clinical trials on the efficacy and safety of drugs for the treatment of low back pain. J Pain & Sympt Mgmt. 2004;28(1):72-95. van Tulder MW, Schotten RJPM, Koes BW, Deyo RA. Non-steroidal anti-inflammatory drugs for low-back pain Cochrane Database
401.
390.
402.
391.
403.
392.
404.
393.
405.
394.
406.
407.
395.
408.
396.
409.
397.
410.
398.
411.
399.
412.
400.

of Systematic Reviews. 2000(2):Art. No.: CD000396. van Tulder MW, Schotten RJPM, Koes BW, Deyo RA. Nonsteroidal anti-inflammatory drugs for low back pain: a systematic review within the framework of the Cochrane Collaboration back review group. Spine. 2000;25(19):2501-2513. Hickey RF, Hickey RF. Chronic low back pain: a comparison of diflunisal with paracetamol. NZ Med J. 1982;95(707): 312-314. Deyo RA. Drug therapy for back pain: Which drugs help which patients? Spine. 1996;21(24):2840-2849. Evans DP, Burke MS, Newcombe RG. Medicines of choice in low back pain. Curr Med Res Opin. 1980;6(8):540-547. Milgrom C, Finestone A, Lev B, et al. Overexertional lumbar and thoracic back pain among recruits: a prospective study of risk factors and treatment regimens. J Spinal Disord. 1993;6(3):187-193. Muckle DS. Flurbiprofen for the treatment of soft tissue trauma. Am J Med. 1986;80(3A):76-80. Stein D, Peri T, Edelstein E, et al. The efficacy of amitriptyline and acetaminophen in the management of acute low back pain. Psychosomatics. 1996;37(1):63-70. Hackett GI, Seddon D, Kaminski D, Hackett GI, Seddon D, Kaminski D. Electroacupuncture compared with paracetamol for acute low back pain. Practitioner. 1988;232(1443):163-164. McGuinness BW, Lloyd-Jones M, Fowler PD, McGuinness BW, Lloyd-Jones M, Fowler PD. A double-blind comparative trial of 'parazolidin' and paracetamol. Br J Clin Pract. 1969;23(11):452-455. Moore N, Van Ganse E, Le Parc JM, et al. The PAIN study: Paracetamol, aspirin and ibuprofen new tolerability study. A largescale, randomised clinical trial comparing the tolerability of aspirin, ibuprofen and paracetamol for short-term analgesia. Clin Drug Invest. 1999;18(2):89-98. Brown FL, Jr., Bodison S, Dixon J, et al. Comparison of diflunisal and acetaminophen with codeine in the treatment of initial or recurrent acute low back strain. Clin Ther. 1986;9 Suppl C:52-58. De Almeida Coimbra F, De Araujo Filho NC. A double-blind study comparing the associations chlormezanone-acetaminophen and thiocolchicoside-glafenin in patients with low back pain. Folha Medica. 1980;81(1):113-116. Gammaitoni AR, Galer BS, Lacouture P, Domingos J, Schlagheck T. Effectiveness and safety of new oxycodone/acetaminophen formulations with reduced acetaminophen for the treatment of low back pain. Pain Medicine. 2003;4(1):21-30. Hale M, Speight K, Harsanyi Z, et al. Efficacy of 12 hourly controlled-release codeine compared with as required dosing of acetaminophen plus codeine in patients with chronic low back pain. Pain Res Manage. 1997;2(1):33-38. Hingorani K. Orphenadine/Paracetamol in backache: a double-blind controlled trial. Br J Clin Pract. 1971;25:227-231. Innes GD, Croskerry P, Worthington J, Beveridge R, Jones D. Ketorolac versus acetaminophen-codeine in the emergency department treatment of acute low back pain. J Emerg Med. 1998;16(4):549-556. Lovell SJ, Taira T, Rodriguez E, Wackett A, Gulla J, Singer AJ. Comparison of valdecoxib and an oxycodone-acetaminophen combination for acute musculoskeletal pain in the emergency department: a randomized controlled trial. Academ Emerg Med. 2004;11(12):1278-1282. Palangio M, Damask MJ, Morris E, et al. Combination hydrocodone and ibuprofen versus combination codeine and acetaminophen for the treatment of chronic pain. Clin Ther. 2000;22(7):879-892. Palangio M, Morris E, Doyle RT, Dornseif BE, Valente TJ. Combination hydrocodone and ibuprofen versus combination oxycodone and acetaminophen in the treatment of moderate or severe acute low back pain. Clin Ther. 2002;24(1):87-99. Ruoff GE, Rosenthal N, Jordan D, Karim R, Kamin M, Protocol C-SG. Tramadol/acetaminophen combination tablets for the treatment of chronic lower back pain: a multicenter, randomized, double-blind, placebo-controlled outpatient study. Clin Ther. 2003;25(4):1123-1141. Thurel C, Bardin T, Boccard E. Analgesic efficacy of an association of 500-mg paracetamol plus 30-mg codeine versus 400mg paracetamol plus 30-mg dextropropoxyphene in repeated doses for
413.
425.
414.
426.
415.
416.
427.
417.
428.
418.
429.
419.
420.
430.
421.
431.
422.
432.
423.
433.
424.

chronic lower back pain. Curr Ther Res Clin Exp. 1991;50(4):463-473. McGuinness BW, McGuinness BW. A double-blind comparison in general practice of a combination tablet containing orphenadrine citrate and paracetamol ('Norgesic') with paracetamol alone. J Intern Med Res. 1983;11(1):42-45. Lee C, Straus W, Balshaw R, Barlas S, Vogel S, Schnitzer T. A comparison of the efficacy and safety of nonsteroidal antiinflammatory agents versus acetaminophen in the treatment of osteoarthritis: a meta-analysis. Arth Rheum. 2004;51(5):746-754. Towheed T, Judd M, Hochberg M, Wells G. Acetaminophen for osteoarthritis. Cochrane Database of Systematic Reviews. 2006(1):Art. No.: CD004257. Wegman A, van der Windt D, van Tulder M, Stalman W, de Vries T. Nonsteroidal antiinflammatory drugs or acetaminophen for osteoarthritis of the hip or knee? A systematic review of evidence and guidelines. J Rheumatol. 2004;31(2):344-354. Zhang W, Jones A, Doherty M. Does paracetamol (acetaminophen) reduce the pain of osteoarthritis? A meta-analysis of randomised controlled trials. Ann Rheum Dis. 2004;63(8):901-907. Garcia Rodriguez L, Hernandez-Diaz S. Risk of uncomplicated peptic ulcer among users of aspirin and nonaspirin nonsteroidal antiinflammatory drugs. Am J Epidemiol. 2004;159(1):23-31. Rahme E, Pettitt D, LeLorier J. Determinants and sequelae associated with utilization of acetaminophen versus traditional nonsteroidal antiinflammatory drugs in an elderly population. Arth Rheum. 2002;46(11):3046-3054. Curhan G, Willett W, Rosner B, Stampfer M. Frequency of analgesic use and risk of hypertension in younger women. Arch Intern Med. 2002;162:2204-2208. Dedier J, Stampfer M, Hankinson S, Willett W, Speizer F, Curhan G. Nonnarcotic analgesic use and the risk of hypertension in US women. Hypertension. 2002;40:604-608. Forman J, Stampfer M, Curhan G. Nonnarcotic analgesic dose and risk of incident hypertension in US women. Hypertension. 2005;46:500-507. McLaughlin J, Lipworth L, Chow W-H, Blot WJ. Analgesic use and chronic renal failure: a critical review of the epidemiologic literature. Kidney International. 1998;54: 679-686. Chan AT, Manson JE, Albert CM, et al. Nonsteroidal antiinflammatory drugs, acetaminophen, and the risk of cardiovascular events. Circulation. 2006;113:1578-1587. Watkins PB, Kaplowitz N, Slattery JT, et al. Aminotransferase elevations in healthy adults receiving 4 grams of acetaminophen daily. JAMA. 2006;296:87-93. FitzGerald GA, Patrono C. The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med. 2001;245(6):433-442. Koes BW, Scholten RJPM, Mens JMA, Bouter LM. Efficacy of non-steroidal antiinflammatory drugs for low back pain: a systematic review of randomized clinical trials. Ann Rheum Diseases. 1997;56(4): 514-223. Berry H, Hutchinson D. Tizanidine and ibuprofen in acute low-back pain: results of a double-blind multicentre study in general practice. J Intern Med Res. 1988;16:83-91. Listrat V, Dougados M, Chevalier X, Kramer F, Amor B. Comparison of the analgesic effect of tenoxicam after oral or intramuscular administration. Drug Invest. 1990;2 (Suppl 3):51-52. Hernandez-Diaz S, Garcia Rodriguez L. Association between nonsteroidal antiinflammatory drugs and upper gastrointestinal tract bleeding/perforation. An overview of epidemiologi studies published in the 1990s. Arch Intern Med. 2000;160: 2093-2099. Garcia Rodriguez, Hernandez-Diaz. Relative risk of upper gastrointestinal complications among users of acetaminophen and nonsteroidal anti-inflammatory drugs. Epidemiology. 2001;12:570-576. Kearney PM, Baigent C, Godwin J, Halls H, Emberson JR, Patrono C. Do selective cyclooxygenase-2 inhibitors and traditional nonsteroidal anti-infalmmatory drugs increase the risk of atherothrombosis? Meta-analysis of randomized trials. BMJ. 2006;332:1302-1308. Food and Drug Administration. FDA Public Advisory: FDA announces important changes and additional warnings for COX-2 selective and non-selective non-steroidal antiinflammatroy drugs (NSAIDS). 2006.
434.
445.
435.
446.
447.
436.
448.
437.
449.
438.
450.
439.
451.
440.
452.
441.
453.
442.
443.
454.
444.

455. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med. 2000;343(21):1520-1528. Furberg CD, Psaty BM, FitzGerald GA. Parecoxib, valdecoxib, and cardiovascular risk. Circulation. 2005;111:249. Birbara CA, Puopolo AD, Munoz DR, et al. Treatment of chronic low back pain with etoricoxib, a new cyclo-oxygenase-2 selective inhibitor: improvement in pain and disability--a randomized, placebo-controlled, 3-month trial. J Pain. 2003;4(6):307-315. Chrubasik S, Kunzel O, model A, conradt C, Black A. Treatment of low back pain with a herbal or synthetic anti-rheumatic: a randomized controlled study. Willow bark extract for low back pain. Rheumatology. 2001;40:1388-1393. Chrubasik S, Model A, Black A, Pollak S. A randomized double-blind pilot study comparing Doloteffin and Vioxx in the treatment of low back pain. Rheumatology. 2003;42(1):141-148. Coats TL, Borenstein DG, Nangia NK, Brown MT. Effects of valdecoxib in the treatment of chronic low back pain: results of a randomized, placebo-controlled trial. Clin Ther. 2004;26(8):1249-1260. Curtis SP, Ng J, Yu Q, et al. Renal Effects of Etoricoxib and Comparator Nonsteroidal AntiInflammatory Drugs in Controlled Clinical Trials. Clin Ther. 2004;26(1):70-83. Ju W, Krupa D, Walters D, Newman T, Borenstein D, Katz N. A placebo-controlled trial of rofecoxib in the treatment of chronic low back pain. Pain Medicine. 2001;2(3): 242-243. Katz N, Ju WD, Krupa DA, et al. Efficacy and safety of rofecoxib in patients with chronic low back pain: results from two 4-week, randomized, placebo-controlled, parallelgroup, double-blind trials. Spine. 2003;28(9):851-858. Katz N, Rodgers DB, Krupa D, Reicin A. Onset of pain relief with rofecoxib in chronic low back pain: results of two four-week, randomized, placebo-controlled trials. Curr Med Res Opin. 2004;20(5):651-658. Pallay RM, Seger W, Adler JL, et al. Etoricoxib reduced pain and disability and improved quality of life in patients with chronic low back pain: a 3 month, randomized, controlled trial. Scand J Rheumatol. 2004;33(4):257-266. Pohjolainen T, Jekunen A, Autio L, Vuorela H. Treatment of acute low back pain with the COX-2-selective anti-inflammatory drug nimesulide: results of a randomized, doubleblind comparative trial versus ibuprofen. Spine. 2000;25(12):1579-1585. Bekker A, Cooper P, Frempong-Boadu A, Babu R, Errico T, Lebovits A. Evaluation of preoperative administration of the cyclooxygenase-2 inhibitor rofecoxib for the treatment of postoperative pain after lumbar disc surgery. Neurosurgery. 2002;50(5):10531057; Karst M, Kegel T, Lukas A, Ludemann W, Hussein S, Piepenbrock S. Effect of celecoxib and dexamethasone on postoperative pain after lumbar disc surgery. Neurosurgery. 2003;53(2):331-336. Reuben S, Connelly N, Reuben S, Connelly N. Postoperative analgesic effects of celecoxib or rofecoxib after spinal fusion surgery. Anesth Analg. 2000;91(5): 1221-1225. Deeks J, Smith L, Bradley M. Efficacy, tolerability, and upper gastrointestinal safety of celecoxib for treatment of osteoarthritis and rheumatoid arthritis: systematic review of randomised controlled trials. BMJ. 2002;325(7365):619. Moore R, Derry S, Makinson G, McQuay H. Tolerability and adverse events in clinical trials of celecoxib in osteoarthritis and rheumatoid arthritis: systematic review and meta-analysis of information from company clinical trial reports. Arth Res Ther. 2005;7:R644-R655. Silverstein F, Faich G, Goldstein J, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: A randomized controlled trial. Celecoxib Long-term Arthritis Safety Study. JAMA. 2000;284(10):1247-1255. Juni P, Rutjes A, Dieppe P. Are selective COX 2 inhibitors superior to traditional non steroidal anti-inflammatory drugs? BMJ. 2002;324:1287-1288. Oyemade GA, Onadeko BO, Oyemade GA, Onadeko BO. A controlled clinical study comparing sulindac with ibuprofen and aspirin in the treatment of musculo-skeletal diseases. J Int Med Res. 1979;7(6):556-559.
466.
456.
457.
467.
458.
468.
459.
469.
460.
470.
461.
471.
462.
472.
463.
464.
473.
474.
465.

475. Dominguez GC. Double blind comparative randomized studio, between a combination of meloxicam and methocarbamol vs acetylsalicylic acid and methocarbamol in patients suffering lumbago. Invest Med Int. 2001;28(1):8-13. Derry S, Loke Y. Risk of gastointestinal haemorrhage with long term use of aspirin: meta-analysis. BMJ. 2000;321:1183-1187. McQuaid KR, Laine L. Systematic review and meta-analysis of adverse events of low-dose aspirin and clopidogrel in randomized controlled trials. Am J Med. 2006;119: 624-638. Saarto T, Wiffen PJ. Antidepressants for neuropathic pain. Cochrane Database of Systematic Reviews. 2005(3):Art. No.: CD005454. Salerno SM, Browning R, Jackson JL. The effect of antidepressant treatment on chronic back pain: a meta-analysis. Arch Intern Med. 2002;162(1):19-24. Staiger TO, Gaster B, Sullivan MD, Deyo RA. Systematic review of antidepressants in the treatment of chronic low back pain. Spine. 2003;28(22):2540-2545. Schreiber S, Vinokur S, Shavelzon V, Pick CG, Zahavi E, Shir Y. A randomized trial of fluoxetine versus amitriptyline in musculoskeletal pain. Isr J Psychiatry Relat Sci. 2001;38(2):88-94. Goodkin K, Gullion CM. Antidepressants for the relief of chronic pain. Do they work? Ann Behav Med. 1989;11(3):83-101. Onghena P, Van Houdenhove B. Antidepressant-induced analgesia in chronic non-malignant pain: a meta-analysis of 39 placebo-controlled studies. Pain. 1992;49:205-219. Turner JA, Denny MC. Do antidepressant medications relieve chronic low back pain? J Family Pract. 1993;37(6):545-553. Fishbain D. Evidence based data on pain relief with antidepressants. Ann Med. List 2000;32(5):305-316. Atkinson JH, Slater MA, Williams RA, et al. A placebo-controlled randomized clinical trial of nortriptyline for chronic low back pain. Pain. 1998;76(3):287-296. Atkinson JH, Slater MA, Wahlgren DR, et al. Effects of noradrenergic and serotonergic antidepressants on chronic low back pain intensity. Pain. 1999;83(2):137-145. 488. van Tulder MW, Touray T, Furlan AD, Solway S, Bouter LM. Muscle relaxants for nonspecific low-back pain Cochrane Database of Systematic Reviews. 2003(4):Art. No.: CD004252. van Tulder MW, Touray T, Furlan AD, Solway S, Bouter LM, Cochrane Back Review Group. Muscle relaxants for nonspecific low back pain: a systematic review within the framework of the Cochrane Collaboration. Spine. 2003;28(17):1978-1992. Hingorani K. Diazepam in backache: a double-blind controlled trial. Ann Phys Med. 1966;12:125-131. Moll W. Zur therapie akuter lumbovertebraler syndrome durch optimale medikamentose muskelrelaxation mittels diazepam. Med Welt. 1973;24:1747-1751. Arbus L, Fajadet B, Aubert D, Morre M, Goldfinger E. Activity of tetrazepam in low back pain. Clin Trials J. 1990;27(4):258-267. Salzmann E, Pforringer W, Paal G, Gierend M. Treatment of chronic low-back syndrome with tetrazepam in a placebo controlled double-blind trial. J Drug Dev. 1992;4: 219-228. Basmajian JV. Cyclobenzaprine hydrochloride effect on skeletal muscle spasm in the lumbar region and neck: two double-blind controlled clinical and laboratory studies. Arch Phys Med Rehabil. 1978;59(2):58-63. Hennies O. A new skeletal muscle relaxant (DS 103-282) compared to diazepam in the treatment of muscle spasm of local origin. J Int Med Res. 1981;9(1):62-68. Boyles W, Glassman J, Soyka J. Management of acute musculoskeletal conditions: Thoracolumbar strain or sprain. A double-blind evaluationg comparing the efficacy and safety of carisoprodol with diazepam. Today's Ther Trends. 1983;1: 1-16. Nibbelink D, Strickland S, McLean L, Gould A. Cyclobenzaprine, diazepam, and placebo in the treatment of skeletal muscle spasm of local origin. Clin Therapeutics. 1978;1: 409-424. Collins SL, Moore RA, McQuay HJ, Wiffen P. Antidepressants and anticonvulsants for diabetic neuropathy and posteherpetic neuralgia: a quantitative systematic review. J Pain Symptom Manage. 2000;20(6):449-458.
489.
476.
477.
490.
478.
491.
479.
492.
493.
480.
481.
494.
482.
495.
483.
496.
484.
485.
497.
486.
498.
487.

499. Finnerup NB, Otto M, McQuay HJ, Jensen TS, Sindrup SH. Algorithm for neuropathic pain treatment: an evidence based proposal. Pain. 2005;118:289-305. Mellegers M, Furlan A, Mailis A. Gabapentin for Neuropathic Pain: Systematic Review of Controlled and Uncontrolled Literature. Clin J Pain. 2001;17(4):284-295. McCleane GJ. Does gabapentin have an analgesic effect on background, movement and referred pain? A randomised, doubleblind, placebo controlled study. Pain Clinic. 2001;13(2):103-107. Yildirim K, Sisecioglu M, Karatay S, et al. The effectiveness of gabapentin in patients with chronic radiculopathy. The Pain Clinic. 2003;15(3):213-218. Khoromi S, Patsalides A, Parada S, Salehi V, Meegan JM, Max MB. Topiramate in chronic lumbar radicular pain. J Pain. 2005;6: 829-836. Muehlbacher M, Nickel MK, Kettler C, et al. Topiramate in treatment of patients with chronic low back pain. A randomized, doubleblind, placebo-controlled study. Clin J Pain. 2006;22:526-531. Chou R, Peterson K, Helfand M. Comparative efficacy and safety of skeletal muscle relaxants for spasticity and musculoskeletal conditions: a systematic review. J Pain Symptom Manage. 2004;28(2):140-175. Browning R, Jackson JL, O'Malley PG. Cyclobenzaprine and back pain: a metaanalysis. Arch Intern Med. 2001;161(13):1613-1620. Berry H, Hutchinson DR. A multicentre placebo-controlled study in general practice to evaluate the efficacy and safety of tizanidine in acute low-back pain. J Int Med Res. 1988;16:75-82. Bartleson JD. Evidence for and against the use of opioid analgesics for chronic nonmalignant low back pain: a review. Pain Medicine. 2002;3(3):260-271. Brown RL, Fleming MF, Patterson JJ. Chronic opioid analgesic therapy for chronic low back pain. J Amer Board Family Prac. 1996;9(3):191-204. Furlan AD, Sandoval JA, Mailis-Gagnon A, Tunks E. Opioids for chronic noncancer pain: A meta-analysis of effectiveness and side effects. CMAJ. 2006;174(11):1589-1594. Kalso E, Edwards JE, Moore RA, McQuay HJ. Opioids in chronic non-cancer pain:systematic review of efficacy and safety. Pain. 2004;112:372-380. Hale ME, Dvergsten C, Gimbel J. Efficacy and safety of oxymorphone extended release in chronic low back pain: results of a randomized, double-blind, placebo- and active-controlled phase III study. J Pain. 2005;6(1):21-28. Allan L, Richarz U, Simpson K, Slappendel R. Transdermal fentanyl versus sustained release oral morphine in strong-opoiods naive patients with chronic low back pain. Spine. 2005;30:2484-2490. Baratta RR. A double-blind comparative study of carisoprodol, propoxyphene, and placebo in the management of low back syndrome. Curr Ther Res Clin Exp. 1976;20(3):233-240. Gostick N, Allen J, Cranfield R, et al. A comparison of the efficacy and adverse effects of controlled release dihydrocodeine and immediate release dihydrocodeine in the treatment of pain in osteoarthritis and chronic back pain. In: Twycross RG, ed. The Edinburgh Symposium on Pain Control and Medical Education. Edinburgh: Royals Society of medicine Press: London, England.; 1989:137-143. Hale ME, Fleischmann R, Salzman R, et al. Efficacy and safety of controlled-release versus immediate-release oxycodone: randomized, double-blind evaluation in patients with chronic back pain. Clin J Pain. 1999;15(3):179-183. Jamison RN, Raymond SA, Slawsby EA, Nedeljkovic SS, Katz NP. Opioid therapy for chronic noncancer back pain. A randomized prospective study. Spine. 1998;23(23): 2591-2600. Salzman RT, Roberts MS, Wild J, Fabian C, Reder RF, Goldenheim PD. Can a controlledrelease oral dose form of oxycodone be used as readily as an immediate-release form for the purpose of titrating to stable pain control? J Pain Symptom Manage. 1999;18(4): 271-279. Muller FO, Odendaal CL, Muller FR, Raubenheimer J, Middle MV, Kummer M. Comparison of the efficacy and tolerability of a paracetamol/codeine fixed-dose combination with tramadol in patients with refractory chronic back pain. Arzneimittel Forschung. 1998;48(6):675-679.
512.
500.
501.
513.
502.
514.
503.
515.
504.
505.
516.
506.
517.
507.
518.
508.
509.
519.
510.
511.

520. Mullican WS, Lacy JR, Group T-A-S. Tramadol/acetaminophen combination tablets and codeine/acetaminophen combination capsules for the management of chronic pain: a comparative trial. Clinical therapeutics. 2001;23(9):1429-1445. Murphy JE, Donald JF, Layes Molla A. Analgesic efficacy and acceptability of fenoprofen combined with paracetamol and compared with dihydrocodeine tartrate in general practice. J Int Med Res. 1978;6(5):375-380. Veenema K, Leahey N, Schneider S. Ketorolac versus meperidine: ED treatment of severe musculoskeletal low back pain. Am J Emerg Med. 2000;18(4):404-407. Moulin D, Iezzi A, Amireh R, Sharpe W, Boyd D, Merskey H. Randomised trial of oral morphine for chronic non-cancer pain. Lancet. 1996;347:143-147. Muncie H, Jr., King D, DeForge B. Treatment of mild to moderate pain of acute soft tissue injury: diflunisal vs acetaminophen with codeine. J Fam Pract. 1986;23(2):125-127. Ingpen ML. A controlled clinical trial of sustained-action dextropropoxyphene hydrochloride. Br J Clin Pract. 1969;23(3):113-115. Rashiq S, Koller M, Haykowsky M, Jamieson K. The effect of opioid analgesia on exercise test performance in chronic low back pain. Pain. 2003;106(1-2):119-125. Chou R, Clark E, Helfand M. Comparative efficacy and safety of long-acting oral opioids for chronic non-cancer pain: a systematic review. J Pain Symptom Manage. 2003;26(5):1026-1048. Moore RA, McQuay HJ. Prevalence of opioid adverse events in chronic non-malignant pain: systematic review of randomised trials of oral opioids. Arth Res Ther. 2005;7: R1046-R1051. Metscher B, Kubler U, Jahnel-Kracht H. [Dexketoprofen-trometamol and tramadol in acute lumbago]. Fortschritte der Medizin Originalien. 2001;118(4):147-151. Schnitzer TJ, Gray WL, Paster RZ, Kamin M. Efficacy of tramadol in treatment of chronic low back pain. J Rheum. 2000;27(3): 772-778. Raber M, Hofmann S, Junge K, Momberger H, Kuhn D. Analgesic efficacy and tolerability of tramadol 100 mg sustained-release capsules in patients with moderate to severe chronic low back pain. Clin Drug Invest. 1999;17(6):415-423. Sorge J, Stadler T. Comparison of the analgesic efficacy and tolerability of tramadol 100mg sustained-release tablets and tramadol 50mg capsules for the treatment of chronic low back pain. Clin Drug Invest. 1997;14(3):157-164. Peloso PM, Fortin L, Beaulieu A, Kamin M, Rosenthal N, Protocol TRPCANSG. Analgesic efficacy and safety of tramadol/ acetaminophen combination tablets (Ultracet) in treatment of chronic low back pain: a multicenter, outpatient, randomized, double blind, placebo controlled trial. J Rheumatol. 2004;31(12):2454-2463. Bamigbade TA, McCartney C, Paes M, Langford RM, Gallagher WJ. A randomised, double blind, crossover study comparing oral tramadol with oral co-proxamol for the treatment of chronic back pain. Paper presented at: 9th World congress on Pain; August 22-27, 1999; Vienna, Austria. Relja M. The role of tramadol in the treatment of acute low back pain. J Neurol Sci. 1990;98:334. Finckh A, Zufferey P, Schurch M, Balague F, Waldburger M, So A. Short-Term Efficacy of Intravenous Pulse Glucocorticoids in Acute Discogenic Sciatica. A Randomized Controlled Trial. Spine. 2006;31(4):377-381. Haimovic IC, Beresford HR. Dexamethasone is not superior to placebo for treating lumbosacral radicular pain. Neurology. 1986;36:1593-1594. Porsman O, Friis H. Prolapsed lumbar disc treated with intramuscularly administered dexamethasonephosphate. A prospectively planned, double-blind, controlled clinical trial in 52 patients. Scand J Rheumatol. 1979;8(3):142-144. Friedman BW, Holden L, Esses D, et al. Parenteral corticosteroids for emergency department patients with non-radicular low back pain. J Emerg Med. 2006;31(4): 365-370. King J. Dexamethasone--a helpful adjunct in management after lumbar discectomy. Neurosurgery. 1984;14(6):697-700. Lundin A, Magnuson A, Axelsson K, Kogler H, Samuelsson L. The effect of perioperative corticosteroids on the outcome of microscopic lumbar disc surgery. Eur Spine J. 2003;12(6):625-630.
532.
521.
533.
522.
523.
534.
524.
535.
525.
536.
526.
537.
527.
538.
528.
539.
529.
530.
540.
541.
531.

542. Watters W, Temple A, Granberry M. The use of dexamethasone in primary lumbar disc surgery. A prospective, randomized, doubleblind study. Spine. 1989;14(4):440-442. Hofferberth B, Gottschaldt M, Grass H, Buttner K. [The usefulness of dexamethasonephosphate in the conservative treatment of lumbar pain--a double-blind study (author's transl)]. Arch Psychiatr Nervenkr. 1982;231:359-367. Nicholson B, Galer BS, Oleka N, Gammaitoni AR. A randomized, open-label study comparing the efficacy and safety of lidocaine patch 5% with celecoxib 200 mg in patients with chronic axial low-back pain. Arth Rheum. 2005;52(9 [suppl]):S523-S524. Gagnier JJ, van Tulder MW, Berman B, Bombardier C. Herbal medicine for low back pain. Cochrane Database of Systematic Reviews. 2006(2):Art. No.: CD004504. Gagnier JJ, van Tulder MW, Berman B, Bombardier C. Herbal medicine for low back pain. A Cochrane review. Spine. 2007;32: 82-92. Gagnier JJ, Chrubasik S, Manheimer E. Harpgophytum procumbens for osteoarthritis and low back pain: A systematic review. BMC Complement Altern Med. 2004;4:13. Chrubasik S, Junck H, Breitschwerdt H, Conradt C, Zappe H. Effectiveness of Harpagophytum extract WS 1531 in the treatment of exacerbation of low back pain: a randomized, placebo-controlled, double-blind study. Eur J Anaesth. 1999;16:118-129. Chrubasik S, Zimpfer CH, Schutt U, Ziegler R. Effectiveness of harpagophytum procumbens in the treatment of acute low back pain. Phytomedicine. 1996;3:1-10. Chrubasik S, Eisenberg E, Balan E, Weinberger T, Luzzati R, Conradt C. Treatment of low back pain exacerbation with willow bark extract: A randomzied doubleblind study. Am J Med. 2000;109:9-14. Ginsberg F, Famaey JP. A double-blind study of topical massage with Rado-Salil Ointment in mechanical low back pain. J Int Med Res. 1987;15:148-153. Keitel W, Frerick H, Kun U, Schmidt U, Kuhlmann M, Bredehoorst A. Capsicum pain plaster in c hronic non-specific low back pain. Arzneimtel Forschung. 2001;51. Frerick H, Keitel W, Kuhn U, Schmidt S, Bredehorst A, Kuhlmann M. Topical treatment of chornic low back pain with a capsicum plaster. Pain. 2003;106:59-64. Stam C, Bonnet MS, van Haselen RA. The efficacy and safety of a homeopathic gel in the treatment of acute low back pain: a multicentre, randomzied, double-blind comparative clinical trial. Br Homeo J. 2001;90:21-28. Cherkin DC, Sherman KJ, Deyo RA, Shekelle PG. A review of the evidence for the effectiveness, safety, and cost of acupuncture, massage therapy, and spinal manipulation for back pain. Ann Intern Med. 2003;138(11):898-906. Ernst E, White AR. Prospective studies of the safety of acupuncture: a systematic review. Am J Med. 2001;110:481-485. van Tulder MW, Cherkin DC, Berman B, Lao L, Koes BW. The effectiveness of acupuncture in the management of acute and chronic low back pain. A systematic review within the framework of the Cochrane Collaboration Back Review Group. Spine. 1999;24(11):1113-1123. Ernst E, White A. Life-threatening adverse reactions after acupuncture? A systematic review. Pain. 1997;71:123-126. Ernst E, White AR. Acupuncture for back pain: a meta-analysis of randomized controlled trials. Arch Intern Med. 1998;158(20):2235-2241. Strauss A. Acupuncture and the treatment of chronic low-back pain: a review of the literature. Chirop J Aust. 1999;29:213. ter Riet G, Kleijnen J, Knipschild P. Acupuncture and chronic pain: a criteriabased meta-analysis. J Clin Epidemiol. 1990;43(11):1191-1199. Patel M. A meta-analysis of acupuncture for chronic pain. Int J Epidemiol. 1989;18(4; Dec):900-906. Ezzo JM, Berman B, Hadhazy VA, Jadad AR, Lao L, Singh BB. Is acupuncture effective for the treatment of chronic pain? A systematic review. Pain. 2000;86:217-225. Smith LA, Oldman AD, McQuay HJ, Moore RA. Teasing apart quality and validity in systematic reviews: an example from acupuncture trials in chronic neck and back pain. Pain. 2000;86(1-2):119-132. Kalauokalani D, Cherkin DC, Sherman KJ, Koepsell TD, Deyo RA. Lessons from a trial of acupuncture and massage for low back pain. Spine. 2001;26:1418-1424.
554.
543.
555.
544.
556.
545.
557.
546.
547.
558.
559.
548.
560.
549.
561.
550.
562.
563.
551.
564.
552.
565.
553.

566. Brinkhaus B, Witt CM, Jena S, et al. Acupuncture in patients with chronic low back pain. Arch Intern Med. 2006;166:450-457. Thomas KJ, MacPherson H, Thorpe L, et al. Randomised controlled trial of a short course of traditional acupuncture compared with usual care for persistent non-specific low back pain. BMJ. 2006;333(7569):623. Witt CM, Jena S, Selim D, et al. Pragmatic randomized trial evaluating the clinical and economic effectiveness of acupuncture for chronic low back pain. Am J Epid. 2006;164(5):487-496. Duplan B, Cabanel G, Piton JL, Grauer JL, Phelip X. [Acupuncture and sciatica in the acute phase. Double-blind study of 30 cases]. Sem Hop. 1983;59:3109-3114. Araki S, Kawamura O, Mataga T, Fujioka H, Huruhigashi S, Nabeta T. Randomized controlled trial comparing manual acupuncture with sham acupuncture for acute low back pain. J Japan Soc Acupunct Moxibustion. 2001;50:356. Shekelle P. Acupuncture for low back pain. Ann Intern Med. 2005;143:691-692. Manheimer E, White A, Ernst E, Langenberg P. Acupuncture for low back pain. Ann Intern Med. 2005;143:692-693. Ceccherelli F, Rigoni MT, Gagliardi G, Ruzzane L. Comparison of superficial and deep acupuncture in the treatment of lumbar myofascial pain: a double-blind randomized controlled study. Clin J Pain. 2002;18(3): 149-153. Carlsson CP, Sjolund BH. Acupuncture for chronic low back pain: a randomized placebo-controlled study with long-term follow-up. Clin J Pain. 2001;17(4):296-305. MacDonald AJ, Macrae KD, Master BR, Rubin AP. Superficial acupuncture in the relief of chronic low back pain. Ann Royal Coll Surg Engl. 1983;65:44-46. Gunn CC, Milbrandt WE, Little AS, Mason KE. Dry needling of muscle motor points for chronic low-back pain: a randomized clinical trial with long-term follow-up. Spine. 1980;5:279-291. Ratcliffe J, Thomas KJ, MacPherson H, Brazier J. A randomised controlled trial of acupuncture care for persistent low back pain: cost effectiveness analysis. BMJ. 2006;333(7569):626. Hsieh LL-C, Kuo C-H, Yen M-F, Chen TH-H. A randomized controlled clinical trial for low back pain treated by acupressure and physical therapy. Prev Med. 2004;39(1): 168-176. Hsieh LL, Kuo CH, Lee LH, Yen AM, Chien KL, Chen TH. Treatment of low back pain by acupressure and physical therapy: randomised controlled trial. BMJ. 2006(332):696-700. Kovacs FM, Abraira V, Lopez-Abente G, F. P. [Neuro-reflexotherapy intervention in the treatment of non specified low back pain: a randomized, controlled, double-blind clinical trial]. Medicina Clinica. 1993;101(15): 570-575. Kovacs FM, Abraira V, Pozo F, et al. Local and remote sustained trigger point therapy for exacerbations of chronic low back pain. A randomized, double-blind, controlled, multicenter trial. Spine. 1997;22(7):786-797. Urrutia G, Burton AK, Morral A, Bonfill X, Zanoli G. Neuroreflexotherapy for nonspecific low-back pain Cochrane Database of Systematic Reviews. 2004(2):Art. No.: CD003009. Kovacs FM, Llobera J, Abraira V, Lazaro P, Pozo F, Kleinbaum D. Effectiveness and cost-effectiveness analysis of neuroreflexotherapy for subacute and chronic low back pain in routine general practice. Spine. 2002;27:1149-1159. Forssell M. The back school. Spine. 1981;6:104-106. Forssell M. The Swedish back school. Physiotherapy. 1980;66:112-114. Heymans MW, van Tulder MW, Esmail R, Bombardier C, Koes BW. Back schools for nonspecific low back pain. A systematic review within the framework of the Cochrane Collaboration Back Review Group. Spine. 2005;2005:2153-2163. Heymans MW, van Tulder MW, Esmail R, Bombardier C, Koes BW. Back schools for non-specific low-back pain Cochrane Database of Systematic Reviews. 2004(4):Art. No.: CD000261. Elders LAM, van der Beek AJ, Burdorf A,. Return to work after sickness absence due to back disorders - a systematic review on intervention strategies. Int Arch Occup Environ Health. 2000;73(5):339-348. Maier-Riehle B, Harter M. The effects of back schools--a meta-analysis. Int J Rehab Res. 2001;24(3):199-206.
567.
579.
568.
580.
569.
581.
570.
582.
571. 572.
583.
573.
584. 585. 586.
574.
575.
587.
576.
588.
577.
589.
578.

590. van der Hulst M, Vollenbroek-Hutten MMR, IJzerman MJ. A systematic review of sociodemographic, physical, and psychological predictors of multidisciplinary rehabilitation - or, back school treatment outcome in patients with chronic low back pain. Spine. 2005;30(7):813-825. van Tulder MW, Esmail R, Bombardier C. Back schools for non-specific low-back pain. Cochrane Database of Systematic Reviews. 1999(3):Art. No.: CD000261. Cohen JE, Goel V, Frank JW, Bombardier C, Peloso P, Guillemin F. Group education interventions for people with low back pain: an overview of the literature. Spine. 1994;19(11):1214-1222. Di Fabio RP. Efficacy of comprehensive rehabilitation programs and back school for patients with low back pain: a meta-analysis. Physical Therapy. 1995;75(10):865-878. Keijsers JFE, Bouter LM, Meertens RM. Validity and comparability of studies on the effects of back schools. Physiother Theory Pract. List 4/20/05 1991;7(3):177-184. Koes BW, van Tulder MW, van der Windt WM, Bouter LM. The efficacy of back schools: a review of randomized clinical trials. J Clin Epidemiol. 1994;47(8):851-862. Nentwig CG. [Effectiveness of the back school. A review of the results of evidencebased evaluation]. Der Orthopde. 1999 1999;28(11):958-965. Turner JA. Educational and behavioral interventions for back pain in primary care. Spine. 1996;21(24):2851-2859. Bergquist-Ullman M, Larsson U. Acute lowback pain in industry. Acta Orthop Scand. 1977;170:1-117. Linton SJ, Bradley LA, Jensen IB, Spangfort E, Sundell L. The secondary prevention of low back pain: A controlled study with followup. Pain. 1989;36:197-207. Lonn JH, Glomsrod B, Soukup MG, Bo K, Larsen S. Active back school: Prophylactic management for low back pain: A randomized controlled 1-year follow-up study. Spine. 1999;24:865-871. Donchin M, Woolf O, Kaplan L, Floman Y. Secondary prevention of low-back pain. A clinical trial. Spine. 1990;15(12):1317-1320. Lankhorst GJ, van der Stadt RJ, Vogelaar TW, van der Korst JK, Prevo AJH. The effect of the Swedish back school in chronic idiopathic low-back pain. Scand J Rehab Med. 1983;15:141-145. Indahl A, Haldorsen EH, Holm S, Reikeras O, Ursin H. Five-year follow-up study of a controlled clinical trial using light mobilization and an informative approach to low back pain. Spine. 1998;23:2625-2630. Indahl A, Velund L, Reikeraas O. Good prognosis for low back pain when left untampered. A randomized clinical trial. Spine. 1995;20:473-477. Leclaire R, Esdaile JM, Suissa S, Rossignol M, Proulx R, Dupuis M. Back school in a first episode of compensated acute low back pain: a clinictal trial to assess efficacy and prevent relapse. Arch Phys Med Rehabil. 1996;77:673-679. Lindequist SL, Lundberg B, Wikmark R, Bergstad B, Loof G, Ottermark AC. Information and regime at low-back pain. Scand J Rehab Med. 1984;16:113-116. Molde Hagen E, Eriksen HR, Ursin H. Does early intervention with a light mobilization program reduce long-term sick leave for low back pain? Spine. 2000;25:1973-1976. Molde Hagen E, Grasdal A, Eriksen HR. Does early intervention with a light mobilization program reduce long-term sick leave for low back pain: A 3-year follow-up study. Spine. 2003;28(20):2309-2315. Karjalainen K, Malmivaara A, Pohjolainen T, et al. Mini-intervention for subacute low back pain: A randomized controlled trial. Spine. 2003;28(6):533-540. Karjalainen K, Malmivaara A, Mutanen P, Roine R, Hurri H, Pohjolainen T. Miniintervention for subacute low back pain: Twoyear follow-up and modifiers of effectiveness. Spine. 2004;29(10):1069-1076. Niemisto L, Lahtinen-Suopanki T, Rissanen P, Lindgren KA, Sarna S, Hurri H. A randomized trial of combined manipulation, stabilizing exercises, and physician consultation compared to physician consultation alone for chronic low back pain. Spine. 2003;28(19):2185-2191. Niemisto L, Rissanen P, Sarna S, LahtinenSuopanki T, Lindgren KA, Hurri H. Costeffectiveness of combined manipulation, stabilizing exercises, and physician consultation compared to physician consultation alone for chornic low back pain: a prospective randomized trial with 2-year follow-up. Spine. 2005;30:1109-1115.
603.
591.
604.
592.
605.
593.
606.
594.
607.
595.
608.
596.
609.
597.
610.
598.
599.
611.
600.
612.
601.
602.

613. Hayden JA, van Tulder MW, Malmivaara AV, Koes BW. Meta-analysis: Exercise therapy for nonspecific low back pain. Ann Intern Med. 2005;142:765-775. Hayden J, van Tulder M, Malmivaara A, Koes B. Exercise therapy for low-back pain Cochrane Database of Systematic Reviews. 2005(3):Art. No.: CD000335. Hayden JA, van Tulder MW, Tomlinson G. Systematic review: strategies for using exercise therapy to improve outcomes in chronic low back pain. Ann Intern Med. 2005;142:776-785. Clare HA, Adams R, Maher CG. A systematic review of efficacy of McKenzie therapy for spinal pain. Aust J Physiother. 2004;50: 209-216. Kool J, de Bie R, Oesch P, Knusel O, van den Brandt P, Bachmann S. Exercise reduces sick leave in patients with non-acute non-specific low back pain: a meta-analysis. J Rehab Med. 2004;36(2):49-62. Liddle S, Baxter G, Gracey J. Exercise and chronic low back pain: what works? Pain. 2004;107(1-2):176-190. Machado LA, de Souza MS, Ferreira PH, et al. The McKenzie method for low back pain: a systematic review of the literature with a meta-analysis approach. Spine. Apr 20 2006;31(9):E254-262. McNeely ML, Torrance G, Magee D. A systematic review of physiotherapy for spondylolysis and spondylolisthesis. Man Ther. 2003;8(2):80-89. van Tulder M, Malmivaara A, Esmail R, Koes B. Exercise therapy for low back pain: a systematic review within the framework of the Cochrane Collaboration back review group. Spine. 2000;25(21):2784-2796. Colle F, Rannou F, Revel M, Fermanian J, Poiraudeau s. Impact of quality scales on levels of evidence inferred from a systematic review of exercise therapy and low back pain. Arch Phys Med Rehabil. 2002;83:1745-1752. Beckerman H, Bouter LM, van der Heijden GJ, de Bie RA, Koes BW. Efficacy of physiotherapy for musculoskeletal disorders: what can we learn from research? Br J Gen Pract. 1993;43(367):73-77. Faas A. Exercises: which ones are worth trying, for which patients, and when? Spine. 1996;21(24):2874-2879. Hilde G, Bo K. Effect of exercise in the treatment of chronic low back pain: a systematic review, emphasising type and dose of exercise. Phys Ther Rev. 1998;3(2):107-117. Koes BW, Bouter LM, Beckerman H, van der heijden GJMG, Knipschild PG. Physiotherapy exercises and back pain: a blinded review. BMJ. 1991;302:1572-1576. Cleland J, Schulte C, Durall C. The role of therapeutic exercise in treating instabilityrelated lumbar spine pain: A systematic review. J Back Musculoskeletal Rehabil. 2002;16:105-115. Ostelo RWJG, de Vet HCW, Waddell G, Kerckhoffs MR, Leffers P, van Tulder MW. Rehabilitation after lumbar disc surgery Cochrane Database of Systematic Reviews. 2002(2):Art. No.: CD003007. UK BEAM Trial Team. United Kingdom back pain exercise and manipulation (UK BEAM) randomised trial: effectiveness of physical treatments for back pain in primary care. BMJ. 2004;329(7479):1377 (Epub). Whitman JM, Flynn TW, Childs JD, et al. A comparison between two physical therapy treatment programs for patients with lumbar spinal stenosis. A randomized clinical trial. Spine. 2006;31:2541-2549. Staal JB, Hlobil H, Twisk JWR, Smid T, Koke AJA, Van Mechelen W. Graded Activity for Low Back Pain in Occupational Health Care: A Randomized, Controlled Trial. Annals of Internal Medicine. 2004;140(2):77-84+I24. Lindstrom I. The effect of graded activity on patients with subacute low back pain: a randomized prospective clnical study with an operant-conditioning behavioural approach. Physical Therapy. 1992;72:279-293. Moffett JK, Torgerson D, Bell-Syer S, et al. Randomised controlled trial of exercise for low back pain: clinical outcomes, costs, and preferences. BMJ. 1999;319:279-283. O'Sullivan PB, Twomey LT, Allison GT. Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis of spondylolisthesis. Spine. 1997;22(24): 2959-2967. Spratt KF, Weinstein JN, Lehmann TR, Woody J, Sayre H. Efficacy of flexion and extension treatments incorporating braces for low-back pain patients with retrodisplacement, spondylolisthesis, or normal sagittal translation. Spine. 1993;18(1839-1849).
626.
614.
627.
615.
628.
616.
617.
629.
618.
630.
619.
631.
620.
632.
621.
633.
622.
634.
623.
635.
624.
625.

636. Rivero-Arias O, Gray A, Frost H, Lamb SE, Stewart-Brown S. Cost-utility analysis of physiotherapy treatment compared with physiotherapy advice in low back pain. Spine. 2006;31:1381-1387. Seferlis T, Lindholm L, Nemeth G. Costminimisation analysis of three conservative treatment programmes in 180 patients sicklisted for acute low-back pain. Scand J Prim Health Care. 2000;18:53-57. McIlveen B, Robertson VJ. A randomised controlled study of the outcome of hydrotherapy for subjects with low back or back and leg pain. Physiotherapy. 1998;84(1):17-26. Sjogren T, Long N, Storay I, Smith J. Group hydrotherapy versus group land-based treatment for chronic low back pain. Physiotherapy research international : the journal for researchers and clinicians in physical therapy. 1997;2(4):212-222. Yozbatiran N, Yildirim Y, Parlak B. Effects of fitness and aquafitness exercises on physical fitness in patients with chronic low back pain. The Pain Clinic. 2004;16(1):35-42. Williams KA, Petronis J, Smith D, et al. Effect of Iyengar yoga therapy for chronic low back pain. Pain. 2005;115(1-2):107-117. Galantino ML, Bzdewka TM, Eissler-Russo JL, et al. The impact of modified Hatha yoga on chronic low back pain: a pilot study. Alternative Therapies in Health & Medicine. 2004;10(2):56-59. Guzman J, Esmail R, Karjalainen K, Malmivaara A, Irvin E, Bombardier C. Multidisciplinary rehabilitation for chronic low back pain: systematic review. BMJ. 2001;322(7301):1511-1516. Guzman J, Esmail R, Karjalainen K, Malmivaara A, Irvin E, Bombardier C. Multidisciplinary bio-psycho-social rehabilitation for chronic low-back pain Cochrane Database of Systematic Reviews. 2002(1):Art. No.: CD000963. Tveito T, Hysing M, Eriksen H. Low back pain interventions at the workplace: a systematic literature review. Occup Med 2004;54(1): 3-13. Loisel P, Abenhaim L, Durand P, et al. A population-based, randomized clinical trial on back pain management. Spine. 1997;22:2911-2918. Bendix AF, Bendix T, Vaegtar KV, Lund C, Frolund L, Holm L. Multidisciplinary intensive treatment for chronic low back pain: a randomized, prospective study. Cleve Clinc J Med. 1996;63:62-69. Bendix AF, Bendix T, Ostenfeld S, Bush E, Andersen. Active treatment programs for patients with chronic low back pain: a prospective, randomized, observer-blinded study. Eur Spine J. 1995;4:148-152. Alaranta H, Rytokoski U, Rissanen A, et al. Intensive physical and psychosocial training program for patients with ch ronic low back pain. A controlled trial. Spine. 1994;19: 1340-1349. Mitchell RI, Carmen GM. The functional restoration approach to the treatment of chronic pain in patients with soft titssue and back injuries. Spine. 1994;19:633-642. Loisel P, Lemaire J, Poitras S, et al. Costbenefit and cost-effectiveness analysis of a disability prevention model for back pain management: a six year follow up study. Occupational & Environmental Medicine. 2002;59(12):807-815. Skoen JS, Grasdal AL, Haldorsen EMH, Ursin H. Relative cost-effectiveness of extensive and light multidisciplinary treatment programs versus treatment as usual for patients with chronic low back pain on longterm sick leave. Spine. 2002;27:901-910. Lechner DE. Work hardening and work conditioning interventions: do they affect disability? Physical Therapy. 1994;74(5): 471-493. Faas A, van Eijk JTM, Chavannes AW, Gubbels JW. A randomized trial of exercise therapy in patients with acute low back pain: Efficacy on sickness absence. Spine. 1995;20(8):941-947. Seferlis T, Nemeth G, Carlsson AM, Gillstrom P. Conservative treatment in patients sicklisted for acute low-back pain: a prospective randomised study with 12 months' follow-up. European Spine Journal. 1998;7(6):461-470. Dettori JR, Bullock SH, Sutlive TG, Franklin RJ, Patience T. The effects of spinal flexion and extension exercises and their associated postures in patients with acute low back pain. Spine. 1995;20(21):2303-2312. Bendix AF, Bendix T, Vaegter K, Busch E, Kirkbak S, Ostenfeld S. [Intensive work rehabilitation]. Videnskab og parksis. 1994;156:2388-2395. Bendix AF, Bendix T, Lund C, Kirkbak S, Ostenfeld S. Comparison of three intensive
648.
637.
649.
638.
650.
639.
651.
640.
652.
641.
642.
653.
643.
654.
644.
655.
656.
645.
646.
657.
647.
658.

programs fro chronic low back pain patients: a prospective, randomzied, observer-blinded study with one year follow-up. Scand J Rehab Med. 1997;29:81-89. Hurley DA, McDonough SM, Dempster M, Moore AP, Baxter GD. A randomized clinical trial of manipulative therapy and interferential therapy for acute low back pain. Spine. 2004;29(20):2207-2216. Werners R, Pynsent P, Bulstrode C. Randomized trial comparing interferential therapy with motorized lumbar traction and massage in the management of low back pain in a primary care setting. Spine. 1999;24(15):1579-1584. Hurley DA, Minder PM, McDonough SM, Walsh DM, Moore AP, Baxter DG. Interferential therapy electrode placement technique in acute low back pain: a preliminary investigation. Archives of physical medicine and rehabilitation. 2001;82(4): 485-493. Beckerman H, de Bie R, Bouter L, Oostendorp R. The efficacy of laser therapy for musculoskeletal and skin disorders: a criteria-based meta-analysis of randomized clinical trials. Phys Ther. 1992;72(7):483-491. Bjordal JM, Couppe C, Chow RT, Tuner J, Ljunggren EA. A systematic review of low level laser therapy with location-specific doses for pain from chronic joint disorders. Australian Journal of Physiotherapy. 9/15/05 2003;49(2):107-116. de Bie R. Efficacy of 904 nm laser therapy in musculoskeletal disorders. Physical Therapy Reviews. 1998;3:1-14. Gam AN, Thorsen H, Lonnberg F. The effect of low-level laser therapy on musculoskeletal pain: A meta-analysis. Pain. 1993;52(1): 63-66. Basford JR, Sheffield CG, Harmsen WS. Laser therapy: a randomized, controlled trial of the effects of low-intensity Nd:YAG laser irradiation on musculoskeletal back pain. Archives of physical medicine and rehabilitation. 1999;80(6):647-652. Klein RG, Eek BC. Low-energy laser treatment and exercise for chronic low back pain: double-blind controlled trial. Arch Phys Med Rehabil. 1990;71(1):34-37. Soriano F, Rios R. Gallium arsenide laser treatment of chronic low back pain: a prospective, randomized and double blind study. Laser Therapy. 1998;10:175-180. 669. Toya S, Motegi M, Inomata K, Ohshiro T, Maed T. Report on a computer-randomized double blind clinical trial to determine the effectiveness of the GaAlAs(830nm) diode laser for attenuation in selected pain groups. Laser Therapy. 1994;6:143-148. Gur A, Karakoc M, Cevik R, Nas K, Sarac AJ. Efficacy of low power laser therapy and exercise on pain and functions in chronic low back pain. Lasers in Surgery & Medicine. 2003;32(3):233-238. Longo L, Tamburini A, Monti A, Cattaneo L, Sesti A. Treatment with 904 nm and 10 600 nm laser of acute lumbago: double blind control. Laser Clinical Research. 1988;3:1620. 1988;3:16-20. Monticone M, Barbarino A, Testi C, Arzano S, Moschi A, Negrini S. Symptomatic efficacy of stabilizing treatment versus laser therapy for sub-acute low back pain with positive tests for sacroiliac dysfunction: a randomised clinical controlled trial with 1 year follow-up. Europa Medicophysica. 2004;40(4):263-268. Gibson T, Harkness J, Blagrave P, Grahame R, Woo P, Hills R. Controlled comparison of short-wave diathermy treatment with osteopathic treatment in non-specific low back pain. Lancet. 1985;1(8440):1258-1261. Rasmussen G. Manipulation in treatment of low back pain: a randomized clinical trial. Manuelle Medizin. 1979;1:8-10. Sweetman BJ, Heinrich I, Anderson JAD. A randomzied controlled trial of exercises, short wave diathermy, and traction for low back pain, with evidence of diagnosis-related response to treatment. Journal of Orthopaedic Rheumatology. 1993;6:159-166. Clarke J, van Tulder M, Blomberg S, de Vet H, van der Heijden G, Bronfort G. Traction for low back pain with or without sciatica: an updated systematic review within the framework of the Cochrane Collaboration. Spine. 2006;31:1591-1599. Clarke J, van Tulder M, Blomberg S, de Vet H, van der Heijden G, Bronfort G. Traction for low-back pain with or without sciatica. The Cochrane Database of Systematic Reviews 2005(4):Art. No.: CD003010. Harte A, Baxter G, Gracey J. The efficacy of traction for back pain. Arch Phys Med Rehabil 2003;84(10):1542-1553. van der Heijden GJ, Beurskens AJ, Koes BW, Assendelft WJ, de Vet HC, Bouter LM. The efficacy of traction for back and neck
659.
670.
660.
671.
661.
672.
662.
673.
663.
674.
675.
664.
665.
676.
666.
677.
667.
678.
668.
679.

pain: a systematic, blinded review of randomized clinical trial methods. Physical Therapy. 1995;75(2):93-104. Beurskens AJ, de Vet HC, van der Heijden GJ, et al. Efficacy of traction for non-specific low back pain: a randomised clinical trial. The Lancet. 1995;346(8990):1596-1600. Beurskens AJP, de Vet HCP, Koke AJPT, et al. Efficacy of Traction for Nonspecific Low Back Pain: 12-Week and 6-Month Results of a Randomized Clinical Trial. Spine. 1997;22(23):2756-2762. van der Heijden G, Beurskens A, Dirx M, Bouter L, Lindeman E. Efficacy of Lumbar Traction: A Randomised Clinical Trial. Physiotherapy. 1995;81(1):29-35. Larsson U, Choler U, Lidstrom A, et al. Autotraction for treatment of lumbago-sciatica. A multicentre controlled investigation. Acta Orthop Scand. 1980;81(5):791-798. Lind G. Auto-traction, treatment of lowback pain and sciatica. An electromyographic, radiographic and clinical study (thesis). Linkping, Sweden, University of Linkping; 1974. Walker L, Svenkerud T, Weber H. [Traksjonbehandling ved lumbago-ischias: en kontrollert undersolske med Spina-trac]. Fysioterapeuten. 1982;49:161-163, 177. Bihaug O. Autotraksjon for ishialgpasienter: en kontollert sammenlikning mellom effekten av Auto-traksjon-B og isometriske ovelser ad modum Hume endall og enkins. Fysioterapeuten. 1978;45:377-379. Lidstrom A, Zachrisson M. Physical therapy on low back pain and sciatica. An attempt at evaluation. Scandinavian Journal of Rehabilitation Medicine. 1970;2(1):37-42. Mathews W, Morkel M, Mathews J. Manipulation and traction for lumbago and sciatica: Physiotherapeutic techniques used in two controlled trials. Physiother Pract. 1988;4(4):201-206. Weber H, Ljunggren A, Walker L. Traction therapy in patients with herniated lumbar intervertebral discs. J Oslo City Hosp. 1984;34:61-70. Ljunggren A, Walker L, Weber H, Amundsen T. Manual traction versus isometric exercises in patients with herniated intervertebral. Physiotherapy Theory and Practice. 1992;8:207-213. Sherry E, Kitchener P, Smart R. A prospective randomized controlled study of VAX-D and TENS for the treatment of chronic low back pain. Neurol Res. 2001;23(7): 780-784. Tesio L, Merlo A. Autotraction versus passive traction: an open controlled study in lumbar disc herniation. Arch Phys Med Rehabil. 1993;74(8):871-876. Letchuman R, Deusinger R. Comparison of sacrospinalis myoelectric activity and pain levels in patients undergoing static and intermittent lumbar traction. Spine. 1993;18(10):1361-1365. Ljunggren A, Weber H, Larsen S. Autotraction versus manual traction in patients with prolapsed lumbar intervertebral discs. Scand J Rehabil Med. 1984;16(3): 117-124. Reust P, Chantraine A, Vischer T. [Traitment par tractions mcaniques des lombosciatalgies avec ou sans d cit neurologique]. Schweiz med Wschr 1988;118(8):271-274. Konrad K, Tatrai T, Hunka A, Vereckei E, Korondi I. Controlled trial of balneotherapy in treatment of low back pain. Annals of the Rheumatic Diseases. 1992;51(6):820-822. Deyo RA, Walsh NE, Martin DC, Schoenfeld LS, Ramamurthy S. A controlled trial of transcutaneous electrical nerve stimulation (TENS) and exercise for chronic low back pain. New England Journal of Medicine. 1990;322:1627-1634. Khadilkar A, Milne A, Brosseau L, et al. Transcutaneous electrical nerve stimulation (TENS) for chronic low-back pain Cochrane Database of Systematic Reviews. 2005(3):Art. No.: CD003008. Khadilkar A, Milne S, Brosseau L, et al. Transcutaneous electrical nerve stimulation for the treatment of chronic low back pain: A systematic review. Spine. 2005;30: 2657-2666. Furlan AD, Brosseau L, Imamura M, Irvin E. Massage for low-back pain Cochrane Database of Systematic Reviews. 2002(2):Art. No.: CD001929. Furlan AD, Brosseau L, Imamura M, Irvin E. Massage for low-back pain: a systematic review within the framework of the Cochrane Collaboration Back Review Group. Spine. 2002;27:1896-1910. Brosseau L, Milne S, Robinson V, et al. Efficacy of the transcutaneous electrical nerve stimulation for the treatment of chronic
680.
692.
681.
693.
682.
694.
683.
695.
684.
696.
685.
697.
686.
698.
687.
699.
688.
700.
689.
701.
690.
702.
691.

low back pain: a meta-analysis. Spine. 2002;27(6):596-603. Milne S, Welch V, Brosseau L, et al. Transcutaneous electrical nerve stimulation (TENS) for chronic low-back pain Cochrane Database of Systematic Reviews. 2001(2):Art. No.: CD003008. Gadsby JG, Flowerdew MW. Transcutaneous electrical nerve stimulation and acupuncturelike transcutaneous electrical nerve stimulation for chronic low back pain. Cochrane Database of Systematic Reviews. 2000(2):Art. No.: CD000210. Flowerdew M, Gadsby G. A review of the treatment of chronic low back pain with acupuncture-like transcutaneous electrical nerve stimulation and transcutaneous electrical nerve stimulation. Comp Ther Med. 1997;5(4):193-201. Pengel HM, Maher CG, Refshauge KM. Systematic review of conservative interventions for subacute low back pain. Clin Rehab. 2002;16:811-820. Cheing GL, Hui-Chan CW. Transcutaneous electrical nerve stimulation: nonparallel antinociceptive effects on chronic clinical pain and acute experimental pain. Archives of Physical Medicine & Rehabilitation. 1999;80(3):305-312. Melzack R, Vetere P, Finch M. Transcutaneous electrical nerve stimulation for low back pain. A comparison of TENS and massage for pain and range of motion. Phys Ther. 1983;63:489-493. Tsukayama H, Yamashita H, Amagai H, Tanno Y. Randomised controlled trial comparing the effectiveness of electroacupuncture and TENS for low back pain: a preliminary study for a pragmatic trial. Acupuncture in Medicine. 2002;20(4): 175-180. Cummings M. Percutaneous electrical nerve stimulation--electroacupuncture by another name? A comparative review. Acupuncture in Medicine. 2001:32-35. Yokoyama M, Sun X, Oku S, et al. Comparison of percutaneous electrical nerve stimulation with transcutaneous electrical nerve stimulation for long-term pain relief in patients with chronic low back pain. Anesthesia and Analgesia. 2004;98(6): 1552-1556. Ghoname EA, Craig WF, White PF, et al. Percutaneous electrical nerve stimulation for low back pain: a randomized crossover study. JAMA. 1999;281(9):818-823. Weiner DK, Rudy TE, Glick RM, et al. Efficacy of percutaneous electrical nerve stimulation for the treatment of chronic low back pain in older adults. Journal of the American Geriatrics Society. 2003;51(5): 599-608. Ghoname EA, White PF, Ahmed HE, Hamza MA, Craig WF, Noe CE. Percutaneous electrical nerve stimulation: an alternative to TENS in the management of sciatica. Pain. 1999;83(2):193-199. van der Windt DAWM, van der Heijden GJMG, van den Berg SGM, ter Riet G, de Winter AF, Bouter LM. Ultrasound therapy for musculoskeletal disorders: a systematic review. Pain. 1999;81:257-271. Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Physical Therapy. 2001;81:1339-1350. Gam AN. Ultrasound therapy in musculoskeletal disorders: a meta-analysis. Pain. 1995;63:85-91. Ansari N. A randomized, single blind placebo controlled trial on the effect of continuous ultrasound on low back pain. Electromyography and Clinical Neurophysiology. 2006;46(6):329-336. Nwuga VC. Ultrasound in treatment of back pain resulting from prolapsed intervertebral disc. Archives of Physical Medicine and Rehabilitation. 1983;64(2):88-89. Roman MP. A clinical evaluation of ultrasound by use of a placebo technic. Phys Ther Rev. 1960;40:649-652. Gnatz SM. Increased radicular pain due to therapeutic ultrasound applied to the back. Archives of Physical Medicine and Rehabilitation. 1989;70(6):493-494. Hoffman BM, Papas RK, Chatkoff DK, Kerns RD. Meta-Analysis of Psychological Interventions for Chronic Low Back Pain. Health Psychology. 2007;26(1):1-9. van Tulder MW, Osteleo R, Vlaeyen JWS, Linton SJ, Morely SJ, Assendelft WJJ. Behavioral treatment for chronic low back pain: a systematic review within the framework of the Cochrane back review group. Spine. 2000;26(3):270-281. Morley S, Eccleston C, Williams A. Systematic review and meta-analysis of randomized controlled trials of cognitive behaviour therapy and behaviour therapy for
703.
713.
704.
714.
705.
715.
716.
706.
717.
707.
718.
708.
719.
720.
709.
721.
722.
710.
723.
711.
724.
712.

chronic pain in adults, excluding headache. Pain. 1999;80:1-13. Turner JA, Clancy S. Comparison of operant behavioral and cognitive-behavioral group treatment for chronic low back pain. J Consult Clin Psychol. 1988;56:261-266. Nouwen A. EMG biofeedback used to reduce standing levels of paraspinal muscle tension in chronic low back pain. Pain. 1983;17: 353-360. Turner JA, Clancy S, McQuade KJ, Cardenas DD. Effectiveness of behavioral therapy for chronic low back pain: a component analysis. J Consult Clin Psychol. 1990;58:573-579. van den Hout JHC, Vlaeyen JWS, Heuts PHTG, Zijlema JHL, Wijnen JAG. Secondary prevention of work-related disability in nonspecific low back pain: Does problemsolving therapy help? A randomized clinical trial. Clinical Journal of Pain. 2003;19(2): 87-96. Kole-Snijders AMJ, Vlaeyen JWS, Goossens MEJB. Chronic low back pain: what does cognitive coping skills training add to operational behavioral treatment? Results of a randomized clinical trial. J Consult Clin Psychol. 1999;67:931-944. Turner JA, Jensen MP. Efficacy of cognitive therapy for chronic low back pain. Pain. 1993;52:169-177. Turner JA. Comparison of group progressiverelaxation training and cognitive-behavioral group therapy for chronic low back pain. J Consult Clin Psychol. 1982;50:757-765. Bru E, Mykletun R, Berge W, Svebak S. Effects of different psychological interventions on neck, shoulder and low back pain in female hospital staff. Psychology and Health. 1994;9:371-382. Goossens MEJB, Rutten-Van Molken MPMH, Kole-Snijders AMJ, Vlaeyen JWS, van Breukelen GV, Leidl R. Health economic assessment of behavioural rehabilitation in chronic low back pain: a randomised clinical trial. Health Economics. 1998;7:39-51. Ernst E. Massage therapy for low back pain: a systematic review. Journal of Pain and Symptom Management. 1999;17:65-69. Ernst E. The safety of massage therapy. Rheumatology. 2003;42:1101-1106. Preyde M. Effectiveness of massage therapy for subacute low-back pain: a randomized controlled trial. CMAJ. 2000;162:1815-1820. 737. Godfrey CM, Morgan PP, Schatzker J. A randomized trial of manipulation for low-back pain in a medical setting. Spine. 1984;9: 301-304. Hoehler FK, Tobis JS, Buerger AA. Spinal manipulation for low back pain. JAMA. 1981;245:1835-1838. Hernandez-Reif M, field T, Krasnegor J, Theadston H. Lower back pain is reduced and range of motion increased after massage therapy. Int J Neurosci. 2001;106:131-145. Franke A, Gebauer S, Franke K, Brockow T. Acupuncture massage vs Swedish massage and individual exercise vs group exercise in low back pain sufferes--a randomised controlled clinical trial in a 2 x 2 factorial trial design. Forschende Komplementarmedizin und Klassische Naturheilkunde. 2000;7: 286-293. Hlobil H, Staal JB, Spoelstra M, Ariens GAM, Smid T, van Mechelen W. Effectiveness of a return-to-work intervention for subacute lowback pain. Scand J Work Environ Health. 2005;31:249-257. Krause N, Dasinger L, Neuhauser F. Modified work and return to work: a review of the literature. J Occup Rehabil. 1998;8(2): 113-139. Scheel IB, Hagen KB, Herrin J, Carling C, Oxman AD. Blind faith? the effects of promoting active sick leave for back pain patients. A cluster-randomized controlled trial. Spine. 2002;27:2734-2740. Scheel IB, Hagen KB, Herrin J, Oxman AD. A call for action. A randomized controlled trial of two strategies to implement active sick leave for patients with low back pain. Spine. 2002;27:561-566. Pittler MH, Karagulle MZ, Karagulle M, et al. Spa therapy and balneotherapy for treating low back pain: meta-analysis of randomized trials. Rheumatology. Jul 2006;45(7): 880-884. Guillemin F, Constant F, Collin JF, Boulange M. Short and long-term effect of spa therapy in chronic low back pain. British journal of rheumatology. 1994;33(2):148-151. Constant F, Guillemin F, Collin JF, Boulange M. Use of spa therapy to improve the quality of life of chronic low back pain patients. Medical Care. 1998;36(9):1309-1314. Constant F, Collin JF, Guillemin F, Boulange M. Effectiveness of spa therapy in chronic low back pain: a randomized clinical trial. The
725.
738.
726.
739.
727.
740.
728.
741.
729.
742.
730.
743.
731.
744.
732.
745.
733.
746.
734.
747.
735. 736.
748.

Journal of Rheumatology. 1995;22(7): 1315-1320. Yurtkuran M, Kahraman Z, Sivrioglu K, Afsin Y, Dogan M. Balneotherapy in Low Back Pain. European Journal of Physical Medicine and Rehabilitation. 1997;7(4):120-123. Brown A, Angus A, Chen S, et al. Costs and outcomes of chiropractic treatment for low back pain. Technology Report No. 56. Ottawa, Canada: Canadian Coordinating Office for Health Technology; July 2005. Ferreira ML, Ferreira PH, Latimer J, Herbert R, Maher CG. Efficacy of spinal manipulative therapy for low back pain of less than three months' duration. Journal of Manipulative & Physiological Therapeutics. 2003;26(9): 593-601. Ferreira ML, Ferreira PH, Latimer J, Herbert R, Maher CG. Does spinal manipulative therapy help people with chronic low back pain? Aust J Physiother. 2002;48:277-284. Avery S, O'Driscoll ML. Randomised controlled trials on the efficacy of spinal manipulation therapy in the treatment of low back pain. Physical Therapy Reviews. 2004;9:146-152. Bronfort G, Haas M, Evans R, Bouter L. Efficacy of spinal manipulation and mobilization for low back pain and neck pain: a systematic review and best evidence synthesis. The Spine Journal. 2004;4: 335-356. Ernst E, Canter P. Chiropractic spinal manipulation treatment for back pain? A systematic review of randomised clinical trials. Phys Ther Rev. 2003;8(2):85-91. Gay RE, Bronfort G, Evans RL. Distraction manipulation of the lumbar spine: A review of the literature. Journal of Manipulative & Physiological Therapeutics. 2005;28: 266-273. Licciardone JC, Brimhall AK, King LN. Osteopathic manipulative treatment for low back pain: a systematic review and metaanalysis of randomized controlled trials. BMC Musculoskeletal Disorders. 2005;6:43. Woodhead T, Clough A. A systematic review of the evidence for manipulation in the treatment of low back pain. Journal of Orthopaedic Medicine. 2005;27:99-120. Ernst E. Prospective investigations into the safety of spinal manipulation. Journal of Pain and Symptom Management. 2001;21(3): 238-242. 760. Meeker WC, Haldeman S. Chiropractic: A profession at the crossroads of mainstream and alternative medicine. Ann Intern med. 2002;136:216-227. Oliphant D. Safety of spinal manipulation in the treatment of lumbar disk herniations: a systematic review and risk assessment. Journal of Manipulative and Physiological Therapeutics 2004;27(3):197-210. Stevinson C, Ernst E. Risks associated with spinal manipulation. The American Journal of Medicine. 2002;112(7):566-571. Kent P. Clinical rule predicts patients likely to benefit from spinal manipulation. Australian Journal of Physiotherapy. 2005; 51(1):56-7. Abenhaim L, Bergeron AM. Twenty years of randomized clinical trials of manipulative therapy for back pain: a review. Clinical & Investigative Medicine. 1992;15(6):527-535. Anderson R, Meeker WC, Wirick BE, Mootz RD, Kirk DH, Adams A. A meta-analysis of clinical trials of spinal manipulation. J Manipulative Physiol Ther. 1992;15(3): 181-194. Assendelft WJJ, Koes BW, van der Heijden GJMG, Bouter LM. The efficacy of chiropractic manipulation for back pain: blinded review of relevant randomized clinical trials. J Manipulative Physiol Ther. 1992;15(8):487-494. Assendelft WJJ, Koes BW, Knipschild PG, Bouter LM. The relationship between methodological quality and conclusions in reviews of spinal manipulation. JAMA. 1995;274(24):1942-1948. Assendelft WJJ, Koes BW, van der Heijden GJMG, Bouter LM. The effectiveness of chiropractic for treatment of low back pain: an update and attempt at statistical pooling. Journal of Manipulative & Physiological Therapeutics. 1996;19(8):499-507. Assendelft WJJ, Bouter LM, Knipschild PG. Complications of spinal manipulation: a comprehensive review of the literature. J Family Pract. 1996;42(5):475-480. Brox JI, Hagen KB, Juel NG, Storheim K. [Is exercise therapy and manipulation effective in low back pain?]. Tidsskrift for Den Norske Laegeforening. 1999;119(14):2042-2050. Di Fabio RP. Efficacy of manual therapy. Phys Ther. 1992;72(12):853-864. Koes BW, Assendelft WJJ, van der Heijden GJMG, Bouter LM. Spinal manipulation for low back pain. An updated systematic review
749.
761.
750.
762.
751.
763.
764.
752.
765.
753.
766.
754.
767.
755.
768.
756.
757.
769.
770.
758.
771. 772.
759.

of randomized clinical trials. Spine. 1996;21(24):2860-2873. Koes BW, Assendelft WJJ, van der Heijden GJMG, Bouter LM, Knipschild PG. Spinal manipulation and mobilisation for back and neck pain: a blinded review. BMJ. 1991;303(6813):1298-1303. Mohseni-Bandpei M. Spinal manipulation in the treatment of low back pain: a review of the literature with particular emphasis on randomized controlled clinical trials. Phys Ther Rev. 1998;3:185-194. Ottenbacher K, DiFabio RP. Efficacy of spinal manipulation/mobilization therapy. A metaanalysis. Spine. 1985;10(9):833-837. Shekelle PG, Adams AH, Chassin MR, Hurwitz EL, Brook RH. Spinal manipulation for low-back pain. Ann Intern Med. 1992;117:590-598. Ernst E. Does spinal manipulation have specific treatment effects? Family Prac. 2000;17(6):554-556. Ernst E, Harkness E. Spinal manipulation: a systematic review of sham-controlled, double-blind, randomized clinical trials. J Pain & Sympt Mgmt. 2001;22(4):879-889. Ernst E. Cerebrovascular complications associated with spinal manipulation. Physical Therapy Reviews. 2004;9(1):5-15. Hurwitz EL, Morgenstern H, Harber P, et al. A randomized trial of medical care with and without physical therapy and chiropractic care with and without physical modalities for patients with low back pain: 6-Month followup outcomes from the UCLA low back pain study. Spine. 2002;27(20):2193-2204. Hurwitz EL, Morgenstern HP, Kominski GFP, Yu FP, Chiang L-MMS. A Randomized Trial of Chiropractic and Medical Care for Patients With Low Back Pain: Eighteen-Month Followup Outcomes From the UCLA Low Back Pain Study. Spine. 2006;31(6):611-621. Santilli V, Behghi E, Finucci S. Chiropractic manipulation in the treatment of acute back pain and sciatica with disc protrusion: a randomized double-blind clinical trial of active and simulated spinal manipulations. Spine. 2006;6:131-137. Wreje U, Nordgren B, Aberg H. Treatment of pelvic joint dysfunction in primary care--a controlled study. Scan J Prim Health Care. 1992;10:310-315. Hadler NM, Curtis P, Gillings DB, Stinnett S. A benefit of spinal manipulation as adjunctive therapy for acute low-back pain: a stratified controlled trial. Spine. 1987;12:702-706. Hadler NM, Curtis P, Gillings DB, Stinnett S. Der Nutzen von Manipulation als zusatzliche Therapie bei akuten Lumbalgien: eine gruppenkontrolierte Studie. Manuelle Medizin. 1990;28:2-6. Kohlbeck F, Haldeman S. Medicationassisted spinal manipulation. The Spine Journal. 2002;2:288-302. Kominski GF, Heslin KC, Morgenstern H, Hurwitz EL, Harber PI. Economic evaluation of four treatments for low-back pain: results from a randomized controlled trial. Medical Care. 2005;43(5):428-435. Skargren EI, Carlsson PG, Oberg BE. Oneyear follow-up comparison of the cost and effectiveness of chiropractic and physiotherapy as primary management for back pain: subgroup analysis, recurrence, and additional health care utilization. Spine. 1998;23(17):1875-1883. Skargren EI, Oberg B, Carlsson PG, Gade M. Cost and effectiveness analysis of chiropractic and physiotherapy treatment for low back and neck pain: six-month follow-up. Spine. 1997;22(18):2167-2177. Ostelo RW, de Vet HC, Berfelo MW, et al. Effectiveness of behavioral graded activity after first-time lumbar disc surgery: short term results of a randomized controlled trial. European Spine Journal. 2003;12(6): 637-644. Najm WI, Seffinger MA, Mishra SI, et al. Content validity of manual spinal palpatory exams - a systematic review. BMC Complement Altern Med. 2003;3(1):1. Seffinger MA, Najm WI, Mishra SI, et al. Reliability of spinal palpation for diagnosis of back and neck pain. Spine. 2004;29(19):E413-E425. van der Wurff P, Hagmeijer RHM, Meyne W. Clinical tests of the sacroiliac joint: A systematic methodological review. Part 1: Reliability. Manual Therapy. 2000;5(1):30-36. van der Wurff P, Meyne W, Hagmeijer RHM. Clinical tests of the sacroiliac joint. A systematic methodological review. Part 2: Validity. Manual Therapy. 2000;5(2):89-96. Hestboek L, Leboef-Yde C. Are Chiropractic Tests for the Lumbo-Pelvic Spine Reliable and Valid? A Systematic Critical Literary Review. Journal of Manipulative & Physiological Therapeutics. 2000;23(4):258.
773.
785.
774.
786.
787.
775.
776.
788.
777.
778.
789.
779.
790.
780.
791.
781.
792.
793.
782.
794.
783.
795.
784.

796. Childs JD, Fritz JM, Flynn TW, et al. A clinical prediction rule to identify patients with low back pain most likely to benefit from spinal manipulation: a validation study [summary for patients in Ann Intern Med. 2004 Dec 21;141(12):I39; PMID: 15611484]. Annals of Internal Medicine. 2004;141(12):920-928. Flynn T, Fritz J, Whitman J, et al. A clinical prediction rule for classifying patients with low back pain who demonstrate short-term improvement with spinal manipulation. Spine. 2002;27(24):2835-2843. Fritz JM, Whitman JM, Childs JD. Lumbar spine segmental mobility assessment: An examination of validity for determining intervention strategies in patients with low back pain. Arch Phys Med Rehabil. 2005;86:1745-1752. Fritz JM, Delitto A, Erhard RE. Comparison of classification-based physical therapy with therapy based on clinical practice guidelines for patients with acute low back pain: a randomized clinical trial. Spine. 2003;28(13):1363-1371. Brennan GP, Fritz JM, Hunter SJ, Thackeray A, Delitto A, Erhard RE. Identifying Subgroups of Patients With Acute/Subacute "Nonspecific" Low Back Pain: Results of a Randomized Clinical Trial. Spine. 2006;31(6):623-631. Fritz JM, Childs JD, Flynn TW. Pragmatic application of a clinical prediction rule in primary care to identify patients with low back pain with a good prognosis following a brief spinal manipulation intervention. BMC Family Practice. 2005;6:29. McGinn TG, Guyatt GH, Wyer PC, Naylor CD, Stiell IG, Richardson WS. Users' guides to the medical literature: XXII: How to use articles about clinical decision rules. JAMA. 2000;284:279-284. Hicks GE, Fritz JM, Delitto A, Mishock J. Interrater reliability of clinical examination measures for identification of lumbar segmental instability. Archives of Physical Medicine & Rehabilitation. 2003;84(12): 1858-1864. Bogduk N, Modic M. Lumbar discography. Spine. 1996;21(3):402-404. Rhyne A, Smith S, Darden B, Wood K. Outcome of unoperated discogram-positive low back pain. Spine. 1995;20(18): 1997-2000. 806. Cohen SP, Larkin TM, Barna SA, Palmer WE, Hecht AC, Stojanovic MP. Lumbar discography: A comprehensive review of outcome studies, diagnostic accuracy, and principles. Regional Anesthesia and Pain Medicine. 2005;30(2):163-183. Shah R, Everett C, McKenzie-Brown A, Sehgal N. Discography as a diagnostic test for spinal pain: A systematic and narrative review Pain Physician. 2005;8(2):187-209. Sackett DL, Haynes RB. Evidence base of clinical diagnosis: The architecture of diagnostic research. BMJ. 2002 2002;324(7336):539-541. Buenaventura R, Shah RV, Patel V, Benyamin R, Singh V. Systematic review of discography as a diagnostic test for spinal pain: An update. Pain Physician. 2007;10:147-164. Willems P, Jacobs W, Duinkerke E, De Kleuver M. Lumbar discography: should we use prophylactic antibiotics? A study of 435 consecutive discograms and a systematic review of the literature. J Spinal Disord Tech. 2004;17(3):243-247. Walsh TR, Weinstein JN, Spratt KF, Lehmann TR, Aprill C, Sayre H. Lumbar discography in normal subjects. A controlled, prospective study. Journal of Bone & Joint Surgery - American Volume. 1990;72(7):1081-1088. Carragee EJ, Alamin TF, Miller J, Grafe M. Provocative discography in volunteer subjects with mild persistent low back pain. The Spine Journal. 2002;2(1):25-34. Carragee EJ, Chen Y, Tanner CM, Truong T, Lau E, Brito JL. Provocative discography in patients after limited lumbar discectomy: A controlled, randomized study of pain response in symptomatic and asymptomatic subjects. Spine. 2000;25(23):3065-3071. Carragee EJ, Tanner CM, Khurana S, et al. The rates of false-positive lumbar discography in select patients without low back symptoms. Spine. Jun 1 2000;25(11):1373-1380; discussion 1381. Carragee EJ, Tanner CM, Yang B, Brito JL, Truong T. False-positive findings on lumbar discography. Reliability of subjective concordance assessment during provocative disc injection. Spine. Dec 1 1999;24(23):2542-2547. Carragee EJ, Alamin TF, Carragee JM. Lowpressure positive discography in subjects
807.
797.
808.
798.
809.
799.
810.
800.
811.
801.
812.
813.
802.
803.
814.
815.
804. 805.
816.

asymptomatic of significant low back pain illness. Spine. 2006;31(5):505-509. Derby R, Kim B-J, Lee S-H, Chen Y, Seo KS, Aprill C. Comparison of discographic findings in asymptomatic subject discs and the negative discs of chronic LBP patients: Can discography distinguish asymptomatic discs among morphologically abnormal discs? The Spine Journal. 2005;5(4): 389-394. Block AR, Vanharanta H, Ohnmeiss DD, et al. Discographic pain report. Influence of psychological factors. Spine. 1996;21(3): 334-338. Heggeness M, Watters W, Gray P. Discography of Lumbar Discs After Surgical Treatment for Disc Herniation. Spine. 1997;22(14):1606-1609. Manchikanti L, Singh V, Pampati V, et al. Provocative Discography in Low Back Pain Patients With or Without Somatization Disorder. Pain Physician. 2001;4(32): 227-239. Ohnmeiss D, Vanharanta H, Guyer R. The association between pain drawings and computed tomographic/discographic pain responses. Spine. 1995;20(6):729-733. Madan S, Gundanna M, Harley JM, Boeree NR, Sampson M. Does provocative discography screening of discogenic back pain improve surgical outcome? Journal of Spinal Disorders & Techniques. 2002;15(3):245-251. Carragee EJ, Lincoln T, Parmar VS, Alamin T. A gold standard evaluation of the "discogenic pain" diagnosis as determined by provocative discography. Spine. 2006;31(18):2115-2123. Willems PC, Elmans L, Anderson PG, van der Schaaf DB, de Kleuver M. Provocative discography and lumbar fusion: is preoperative assessment of adjacent discs useful? Spine. 2007;32(10):1094-1099. Holt EJ. The question of lumbar discography. J Bone Joint Surg Am. 1968;50(4):720-726. Massie W, Stevens D. A critical evaluation of discography. J Bone Joint Surgery. 1967;49A:1243. O'Neill C, Kurgansky M. Subgroups of Positive Discs on Discography. Spine. 2004;29(19):2134-2139. Derby R, Lee S-H, Kim B-J, Chen Y, Aprill C, Bogduk N. Pressure-Controlled Lumbar Discography in Volunteers Without Low Back Symptoms. Pain Medicine. 2005;6(3): 213-221. Gill K, Blumenthal S. Functional results after anterior lumbar fusion at L5-S1 in patients with normal and abnormal MRI scans. Spine. 1992;17(8):940-942. Colhoun E, McCall IW, Williams L, Cassar Pullicino VN. Provocation discography as a guide to planning operations on the spine. Journal of Bone & Joint Surgery - British Volume. 1988;70(2):267-271. Osti OL, Fraser RD, Vernon-Roberts B. Discitis after discography. The role of prophylactic antibiotics. J Bone Joint Surg (Br). 1990;72:271-274. Flanagan MN, Chung BU. Roentgenographic changes in 188 patients 10-20 years after discography and chemonucleolysis. Spine. 1986;11:444-448. Everett C, Shah R, Sehgal N, McKenzieBrown A. A systematic review of diagnostic utility of selective nerve root blocks Pain Physician. 2005;8(2):225-233. Datta S, Everett CR, Trescot AM, et al. An updated systematic review of the diagnostic utility of selective nerve root blocks. Pain Physician. 2007;20:113-128. Sehgal N, V. R, Shah R, McKenzie-Brown A, Everett C. Diagnostic utility of facet (zygapophysial) joint injections in chronic spinal pain: A systematic review of evidence Pain Physician. 2005;8(2):211-224. Boswell M, Singh V, Staats P, Hirsch J. Accuracy of precision diagnostic blocks in the diagnosis of chronic spinal pain of facet or zygapophysial joint origin. Pain Physician. 2003;6(4):449-456. Sehgal N, Dunbar EE, Shah RV, Colson J. Systematic review of diagnostic utility of facet (zygapophysial) joint injections in chronic spinal pain: An update. Pain Physician. 2007;10:213-228. Birkenmaier C, Veihelmann A, Trouillier HH, Hausdorf J, von Schulze Pellengahr C. Medial branch blocks versus pericapsular blocks in selecting patients for percutaneous cryodenervation of lumbar facet joints. Regional Anesthesia and Pain Medicine. 2007;32(1):27-33. Li M, Goldberger BA, Hopkins C. Fatal case of BOTOX-related anaphylaxis? Journal of Forensic Sciences. 2005;50(1):169-172. Aldrete JA. Epidural injections of indomethacin for postlaminectomy syndrome:
817.
829.
830.
818.
831.
819.
832.
820.
833.
821.
834.
822.
835.
836.
823.
837.
824.
838.
825. 826.
827.
839.
828.
840.

a preliminary report. Anesth Analg. 2003;96:463-468. Buchner N, Zeifang F, Brocai D, Schiltenwolf M. Epidural corticosteroid injection in the conservative management of sciatica. Clinical Orthopaedics & Related Research. 2000;375:149-156. Buttermann GR, Buttermann GR. Treatment of lumbar disc herniation: epidural steroid injection compared with discectomy. A prospective, randomized study. Journal of Bone & Joint Surgery - American Volume. 2004;86-A(4):670-679. Dashfield AK, Taylor MB, Cleaver JS, Farrow D. Comparison of caudal steroid epidural with targeted steroid placement during spinal endoscopy for chronic sciatica: a prospective, randomized, double-blind trial. Br. J. Anaesth. 2005 2005;94(4):514-519. Devulder J, Deene P, De Laat M, van Bastelaere M, Brusselmans G, Rolly G. Nerve root sleeve injections in patients with failed back surgery syndrome: a comparison of three solutions. Clin J Pain. 1999;15: 132-135. Helsa PE, Breivik H. Epidural analgesia and epidural steroid injection for treatment of chronic low back pain and sciatica. Tidsskrift for Den Norske Laegeforening. 1979;21: 511-513. Kolsi I, Delecrin J, Berthelot JM, Thomas L, Prost A, Maugars Y. Efficacy of nerve root versus interspinous injections of glucocorticoids in the treatment of diskrelated sciatica. A pilot, prospective, randomized, double-blind study. Joint Bone Spine. 2000;67:113-118. McGregor AH, Anjarwalla NK, Stambach T. Does the method of injection alter the outcome of epidural injections? J Spinal Disord. 2001;14:507-510. Meadeb J, Rozenberg S, Duquesnoy B. Forceful sacroccoccygeal injections in the treatment of postdiscectomy sciatica: A controlled study versus gluccocorticoid injections. Joint Bone Spine. 2001;68:43-49. Pirbudak L, Karakurum G, Oner U, Gulec A, Karadasli H. Epidural corticosteroid injection and amitriptyline for the treatment of chronic low back pain associated with radiculopathy. Pain Clinic. 2003;15(3):247-253. Revel M, Auleley G, Alaoui S. Forceful epidural injections for the treatment of lumbosciatic pain with post-operative lumbar spinal fibrosis. Rev Rheum Eng Ed. 1996;63:270-277. Rocco A, Frank E, Kaul A, Lipson S, Gallo J. Epidural steroids, epidural morphine and epidural steroids combined with morphine in the treatment of post-laminectomy syndrome. Pain. 1989;36(3):297-303. Serrao JM, Marks RL, Morley SJ, Goodchild CS. Intrathecal midazolam for the treatment of chronic mechanical low back pain: a controlled comparison with epidural steroid in a pilot study. Pain. 1992;48(1):5-12. Thomas E, Cyteval C, Abiad L, Picot MC, Taourel P, Blotman F. Efficacy of transforaminal versus interspinous corticosteroid injections in discal radiculalgia-a prospective, randomized, double-blind study. Clin Rheumatol. 2003;22:299-304. Cuckler JM, Bernini PA, Wiesel SW, Booth RE, Rothman RH, Pickens GT. The use of epidural steroids in the treatment of lumbar radicular pain. A prospective, randomized, double-blind study. J Bone Joint Surg Am. 1985;67(1):63-66. Glaser SE, Falco F. Paraplegia following a thoracolumbar transforaminal epidural steroid injection. Pain Physician. 2005;8:309-314. Hooten WM, Mizerak A, Carns PE, Huntoon MA. Discitis after lumbar epidural corticosteroid injection: a case report and analysis of the case report literature. Pain Medicine. 2006;7(1):46-51. Huntoon MA, Martin DP. Paralysis after transoforaminal epidural injection and previous spinal surgery. Reg Anesth Pain Med. 2004;29(5):494-495. Price C, Arden N, Coglan L, Rogers P. Cost effectiveness and safety of epidural steroids in the management of sciatica. Health Technology Assessment: NHS R&D HTA Programme; 2005. Fuchs S, Erbe T, Fischer HL, Tibesku CO. Intraarticular hyaluronic acid versus glucocorticoid injections for nonradicular pain in the lumbar spine. J Vac Interv Radiol. 2005;16:1493-1498. Lilius G, Harilainen A, Laasonen M, Myllynen P. Chronic unilateral lowback pain. Predictors of outcome of facet joint injections. Spine. 1990;15:780-782. Manchikanti L, Manchikanti KN, Manchukonda R, et al. Evaluation of lumbar facet joint nerve blocks in the management of chronic low back pain: preliminary report of a
841.
851.
842.
852.
853.
843.
854.
844.
855.
845.
856.
846.
857.
858.
847.
848.
859.
849.
860.
861.
850.

randomized, double-blind controlled trial: clinical trial NCT00355914. Pain Physician. 2007;10(3):425-440. Manchikanti L. Effectiveness of lumbar facet joint nerve blocks in chronic low back pain: a randomized clinical trial. Pain Physician. 2001;4:101-117. Marks RC, Houston T, Thulbourne T. Facet joint injection and facet nerve block: a randomised comparison in 86 patients with chronic low back pain. Pain. 1992;49(3): 325-328. Mayer TG, Gatchel RJ, Keeley J, McGeary D, Dersh J, Anagnostis C. A randomized clinical trial of treatment for lumbar segmental rigidity. Spine. 2004;29(20):2199-2205. Bontoux D, Alcalay M, Debiais F, et al. Treatment of lumbar disk hernia by intra-disk injection of chymopapain or triamcinolone hexacetonide. Comparative study of 80 cases. Rev Rhum Mal Osteoartic. 1990;57:327-331. Bourgeois P, Benoist M, Palazzo E, et al. [Multicenter randomized double-blind study of triamcinolone hexacetonide versus chymopapain in the treatment of disc lumbosciatica. Initial results at 6 months]. Rev Rhum Mal Osteoartic. 1988;55:767-769. Graham C. Chemonucleolysis: a double blind study comparing chemonucleolysis with intra discal hydrocortisone: in the treatment of backache and sciatica. Clinical Orthopaedics & Related Research. 1976(117):179-192. Benoist M, Bonneville JF, Lassale B, et al. A randomized, double-blind study to compare low-dose with standard-dose chymopapain in the treatment of herniated lumbar intervertebral discs. Spine. 1993;18:28-34. Crawshaw C, Frazer AM, Merriam WF, Mulholland RC, Webb JK. A comparison of surgery and chemonucleolysis in the treatment of sciatica. A prospective randomized trial. Spine. 1984;9(2):195-198. Ejeskar A, Nachemson A, Herberts P, et al. Surgery versus chemonucleolysis for herniated lumbar discs. A prospective study with random assignment. Clin Orthop. 1983;174:236-242. Feldman J, Menkes CJ, Pallardy G, et al. [Double-blind study of the treatment of disc lumbosciatica by chemonucleolysis]. Rev Rhum Mal Osteoartic. 1986;53(3):147-152. Hedtmann A, Fett H, Steffen R, Kramer J. [Chemonucleolysis using chymopapain and collagenase. 3-year results of a prospective randomized study]. Z Orthop Ihren Grenzgeb. 1992;130:36-44. Krugluger J, Knahr K. Chemonucelolysis and automated percutaneous discectomy--a prospective randomized comparison. International Orthopaedics. 2000;24:167-169. Lavignolle B, Vital J, Baulny D, Grenier F, Castagnera L. [Comparative study of surgery and chemonucleolysis in the treatment of sciatica caused by a herniated disc]. Acta Orthop Belg. 1987;53(2):244-249. Muralikuttan KP, Hamilton A, Kernohan WG, Mollan RA, Adair IV. A prospective randomized trial of chemonucleolysis and conventional disc surgery in single level lumbar disc herniation. Spine. 1992;17(4):381-387. Revel M, Payan C, Vallee C, et al. Automated percutaneous lumbar discectomy versus chemonucleolysis in the treatment of sciatica. A randomized multicenter trial. Spine. 1993;18:1-7. Steffen R, Luetke A, Wittenbert R, Kraemer J. A prospective comparative study of chemonucleolysis and laser discectomy. Orthop Trans. 1996;20:388. van Alphen HA, Braakman R, Bezemer PD, Broere G, Berfelo MW. Chemonucleolysis or discectomy: a randomized multicenter trial. J Neurosurg. 1989;70(6):869-875. Day PL. Early, interim, and long-term observations on chemonucleolysis in 876 patients with special comments on the lateral approach. Clin Orthop. 1974;99:64. Wiltse LL, Widell EH, Yuan HA. Chymopapain chemonucleolysis in lumbar disk disease. JAMA. 1975;231:474-479. Agre K, Wilson RR, Brim M, McDermott DJ. Chymodiactin postmarketing surveillance: demographic and adverse experience data. Spine. 1984;9:479-486. Brown MD, Tompkins JS. Chemonucleolysis (discolysis) with collagenase. Spine. 1989;11:123-130. Launois R, Henry B, Marty JR, et al. Chemonucleolysis versus surgical discectomy for sciatica secondary to lumbar disc herniation: A cost and quality-of-life evaluation. PharmacoEconomics. 1994;6(5):453-463. Ramirez LF, Javid MJ. Cost effectiveness of chemonucleolysis versus laminectomy in the
862.
873.
863.
874.
864.
875.
865.
876.
866.
877.
878.
867.
879.
868.
880.
869.
881.
882.
870.
883.
871.
884.
872.

treatment of herniated nucleus pulposus. Spine. 1985;10(4):363-367. Sanders M, Zuurmond WW. Percutaneous intraarticular lumbar facet joint denervation in the treatment of low back pain: A comparison with percutaneous extra-articular lumbar facet denervation. Pain Clin. 1999;11: 329-335. Bogduk N. Lumbar radiofrequency neurotomy. Clinical Journal of Pain. 2006 2006;22(4):409. Gofeld M. Radiofrequency Facet Denervation: A Randomized Control Placebo Versus Sham Procedure. Clinical Journal of Pain. 2006;22(4):410-411. Freeman B, Fraser RD, Cain CJ, Hall D, Chapple D. A randomized, double-blind, controlled efficacy study: Intradiscal electrothermal therapy (IDET) versus placebo for the treatment of chronic discogenic low back pain. European Spine Journal. 2003;12:S23. Gerszten PC, Welch WC, McGrath PM, Willis SL. A prospective outcomes study of patients undergoing intradiscal electrothermy (IDET) for chronic low back pain. Pain Physician. 2002;5(4):360-364. Ercelen O, Bulutcu E, Oktenoglu T, et al. Radiofrequency lesioning using two different time modalities for the treatment of lumbar discogenic pain: a randomized trial. Spine. 2003;28(17):1922-1927. Taylor RJ, Taylor RS. Spinal cord stimulation for failed back surgery syndrome: A decisionanalytic model and cost-effectiveness analysis. International Journal of Technology Assessment in Health Care. 2005;21(3): 351-358. Boden SD. Use of Recombinant Human Bone Morphogenetic Protein-2 to Achieve Posterolateral Lumbar Spine Fusion in Humans: A Prospective, Randomized Clinical Pilot Trial 2002 Volvo Award in Clinical Studies. Spine. 2002;27:2662-2673. Burkus JK, Transfeldt EE, Kitchel SH, Watkins RG, Balderston RA. Clinical and radiographic outcomes of anterior lumbar interbody fusion using recombinant human bone morphogenetic protein-2. Spine. 2002;27(21):2396-2408. Burkus JK, Gornet MF, Dickman CA, Zdeblick TA. Anterior lumbar interbody fusion using rhBMP-2 with tapered interbody cages. Journal of Spinal Disorders & Techniques. 2002;15(5):337-349. France JC, Yaszemski MJ, Lauerman WC, Cain JE, Glover JM, Lawson KJ. A Randomized Prospective Study of Posterolateral Lumbar Fusion: Outcomes With and Without Pedicle Screw Instrumentation. Spine. 1999;24:553-560. Gibson S, McLeod I, Wardlaw D, Urbaniak S. Allograft Versus Autograft in Instrumented Posterolateral Lumbar Spinal Fusion: A Randomized Control Trial. Spine. 2002;27:1599-1603. Goodwin CB, Brighton CT, Guyer RD, Johnson JR, Light KI, Yuan HA. A doubleblind study of capacitively coupled electrical stimulation as an adjunct to lumbar spinal fusions. Spine. 1999;24(13):1349-1356. Jenis LG, An HS, Stein R, Young B. Prospective comparison of the effect of direct current electrical stimulation and pulsed electromagnetic field on instrumented posterolateral lumbar arthrodesis. Journal of Spine Disorders. 2000;13:290-296. Linovitz R, Pathria M, Bernhardt M, et al. Combined magnetic fields accelerate and increase spine fusion. A double-blind, randomized, placebo controlled study. Spine. 2002;27:1383-1389. Lorenz M, Zindrick M, Schwaegler P, et al. A comparison of single level fusions with and without hardware. Spine. 1991;16(Suppl 8):S455-S458. Madan S, Boeree N. Outcome of the Graf ligamentoplasty procedure compared with anterior lumbar interbody fusion with the Hartshill horseshoe cage. European Spine Journal. 2003;12(4):361-368. Mooney V. A randomized double-blind prospective study of the efficacy of pulsed electromagnetic fields for interbody lumbar fusions. Spine. 1990;15:708-712. Sasso RC, Kitchel SH, Dawson EG. A prospective, randomized controlled clinical trial of anterior lumbar interbody fusion using a titanium cylindrical threaded fusion device. Spine. 2004;29(2):113-121. Schofferman J, Slosar P, Reynolds J, Goldthwaite N, Koestler M. A prospective randomized comparison of 270 degrees fusions to 360 degrees fusions (circumferential fusions). Spine. May 15 2001;26(10):E207-212.
885.
895.
886.
896.
887.
897.
888.
898.
889.
899.
890.
900.
891.
901.
892.
902.
903.
893.
904.
894.

905. Zdeblick TA. A prospective, randomized study of lumbar fusion. Preliminary results. Spine. 1993;18:983-991. Bridwell KH, Sedgewick TA, O'Brien MF, Lenke LG, Baldus C. The role of fusion and instrumentation in the treatment of degenerative spondylolisthesis with spinal stenosis. J Spinal Disord. 1993;6:461-472. Mirza SK. Point of view: Commentary on the research reports that led to Food and Drug Administration approval of an artificial disc. Spine. 2005;30(14):1561-1564. Herron LD, Turner JA, Novell LA, Kreif SL. Patient selection for lumbar discectomy with a revised objective rating system. Clinical Orthopaedics & Related Research. 1996(325):148-155. Herron LD, Turner JA, Weiner P. Lumbar disc herniations: The predictive value of the Health Attribution Test (HAT) and the Minnesota Multiphasic Personality Inventory (MMPI). Journal of Spinal Disorders. 1988;1(1):2-8. Hagg O, Fritzell P, Ekselius L, Nordwall A, Swedish Lumbar Spine S. Predictors of outcome in fusion surgery for chronic low back pain. A report from the Swedish Lumbar Spine Study. European Spine Journal. 2003;12(1):22-33. Fritzell P, Hagg O, Nordwall A, Swedish Lumbar Spine Study G. Complications in lumbar fusion surgery for chronic low back pain: comparison of three surgical techniques used in a prospective randomized study. A report from the Swedish Lumbar Spine Study Group. European Spine Journal. 2003;12(2):178-189. Cowan JA, Jr., Dimick JB, Wainess R, Upchurch GR, Jr., Chandler WF, La Marca F. Changes in the utilization of spinal fusion in the United States. Neurosurgery. 2006;59: 15-20. David T. Long-term results of one-level lumbar arthroplasty. Spine. 2007;32:661-666. Punt IM, Visser VM, van Rhijn LW, et al. Complications and reoperations of the SB Charite lumbar disc prosthesis: experience in 75 patients. Eur Spine J. 2008;17:36-43. Putzier M, Funk JF, Schneider SV, et al. Charite total disc replacement--clinical and radiographical results after an average followup of 17 years. European Spine Journal. 2006;15(2):183-195. 916. van Ooij A, Oner FC, Verbout AJ. Complications of artificial disc replacement. Journal of Spinal Disorders & Techniques. 2003;16:369-383. Regan J, McAfee P, Blumenthal S, et al. Evaluation of surgical volume and the early experience with lumbar total disc replacement as part of the investigational device exemption study of the Charite Artificial Disc. Spine. 2006;31(19):2270-2276. Fritzell P, Hagg O, Jonsson D, Nordwall A, Swedish Lumbar Spine Study Group. Costeffectiveness of lumbar fusion and nonsurgical treatment for chronic low back pain in the Swedish Lumbar Spine Study: a multicenter, randomized, controlled trial from the Swedish Lumbar Spine Study Group. Spine. 2004;29(4):421-434. Rivero-Arias O, Campbell H, Gray A, Fairbank J, Frost H, Wilson-MacDonald J. Surgical stabilisation of the spine compared with a programme of intensive rehabilitation for the management of patients with chronic low back pain: cost utility analysis based on a randomised controlled trial. BMJ. 2005;330(7502):1239. Carragee EJ. Single-level posterolateral arthrodesis, with or without posterior decompression, for the treatment of isthmic spondylolisthesis in adults. A prospective, randomised study. J Bone Joint Surg. 1997;79-A:1175-1180. Christensen F, Hansen E, Laursen M, Thomsen K, Bunger C. Long-term functional outcome of pedicle screw instrumentation as a support for posterolateral spinal fusion. Spine. 2002;27:1269-1277. Ekman P, Moller H, Hedlund R. The longterm effect of posterolateral fusion in adult isthmic spondylolisthesis: A randomized controlled study. Spine Journal. 2005;5(1): 36-44. McGuire RA, Amundsen GM. The use of primary internal fixation in spondylolisthesis. Spine. 1993;18:1662-1672. Moller H, Hedlund R. Instrumented and noninstrumented posterolateral fusion in adult spondylolisthesis: a prospective randomized study: part 2. Spine. 2000;25(13):1716-1721. Moller H, Hedlund R. Surgery versus conservative management in adult isthmic spondylolisthesis: a prospective randomized study: part 1. Spine. 2000;25(13):1711-1715.
906.
917.
907.
918.
908.
909.
919.
910.
920.
911.
921.
912.
922.
913. 914.
923.
924.
915.
925.

926. Thomsen K, Christensen FB, Eiskjaer SP, Hansen ES, Furensgaard S, Bunger CE. 1997 Volvo Award winner in clinical studies: The effect of pedicle screw instrumentation on functional outcomes and fusion rates in posterolateral lumbar spinal fusion: a prospective, randomized clinical study. Spine. 1997;22(24):2813-2822. Videbaek TS, Christensen FB, Soegaard R, et al. Circumferential fusion improves outcome in comparison with instrumented posterolateral fusion: long-term results of a randomized clinical trial. Spine. 2006;31(25):2875-2880. Swan J, Hurwitz E, Malek F, et al. Surgical treatment for unstable low-grade isthmic spondyolisthesis in adults: a prospective controlled study of posterior instrumented fusion compared with combined anterioposterior fusion. The Spine Journal. 2006;6:606-614. Christensen FB, Hansen ES, Eiskjaer SP, et al. Circumferential lumbar spinal fusion with Brantigan cage versus posterolateral fusion with titanium Cotrel-Dubousset instrumentation: a prospective, randomized clinical study of 146 patients. Spine. 2002;27(23):2674-2683. Fischgrund J, Mackay M, Herkowitz H, Brower R, Montgomery D, Kurz L. 1997 Volvo Award winner in clinical studies. Degenerative lumbar spondylolisthesis with spinal stenosis: a prospective, randomized study comparing decompressive laminectomy and arthrodesis with and without spinal instrumentation. Spine. 1997;22(24): 2807-2812. Grob D, Humke T, Dvorak J. Degenerative lumbar spinal stenosis. Decompression with and without arthrodesis. J Bone Joint Surg Am. 1995;77(7):1036-1041. Herkowitz H, Kurz L. Degenerative lumbar spondylolisthesis with spinal stenosis. A prospective study comparing decompression with decompression and intertransverse process arthrodesis. J Bone Joint Surg Am. 1991;73-A(6):802-808. Kitchel S, Matteri R. Prospective randomized evaluation of PLIF in degenerative spondylolisthesis patients over 60 years old. Current Concepts Review. 2002. Postacchini F, Cinotti G, Perugia D, Gumina S. The surgical treatment of central lumbar stenosis. J Bone Joint Surg. 1993;75-B: 386-392. Zhao J, Wang X, Hou T, He S. One versus two BAK fusion cages in posterior lumbar interbody fusion to L4-L5 degenerative spondylolisthesis: A randomized, controlled prospective study in 25 patients with minimum two-year follow-up. Spine. 2002;27(24):2753-2757. Birkmeyer NJ, Weinstein J, Tosteson A, et al. Design of the Spine Patient Outcomes Research Trial (SPORT). Spine. 2002;27(12):1361-1372. Atlas S, Keller R, Wu Y, Deyo R, Singer D. Long-Term Outcomes of Surgical and Nonsurgical Management of Lumbar Spinal Stenosis: 8 to 10 Year Results from the Maine Lumbar Spine Study. Spine. 2005;30(8):936-943. Atlas SJ, Deyo RA, Keller RB, et al. The Maine Lumbar Spine Study, Part III: 1-Year Outcomes of Surgical and Nonsurgical Management of Lumbar Spinal Stenosis. Spine. 1996;21(15):1787-1794. Atlas S, Keller R, Robson D, Deyo R, Singer D. Surgical and Nonsurgical Management of Lumbar Spinal Stenosis: Four-Year Outcomes From the Maine Lumbar Spine Study. Spine. 2000;25(5):556-562. Zucherman JF, Hsu KY, Hartjen CA, et al. A multicenter, prospective, randomized trial evaluating the X STOP interspinous process decompression system for the treatment of neurogenic intermittent claudication. Twoyear follow-up results. Spine. 2005;30: 1351-1358. Hsu KY, Zucherman JF, Hartjen CA, et al. Quality of life of lumbar stenosis-treated patients in whom the X STOP interspinous device was implanted. Journal of Neurosurgery Spine. Dec 2006;5(6):500-507. Kuntz KM, Snider RK, Weinstein JN, Pope MH, Katz JN. Cost-effectiveness of fusion with and without instrumentation for patients with degenerative spondylolisthesis and spinal stenosis. Spine. 2000;25:1132-1139. Katz JN, Lipson SJ, Lew RA, et al. Lumbar laminectomy alone or with instrumented or noninstrumented arthrodesis in degenerative lumbar spinal stenosis: patient selection, costs, and surgical outcomes. Spine. 1997;22(10):1123--1131. Bernsmann K, Kramer J, Ziozios I, Wehmeier J, Wiese M. Lumbar micro disc surgery with
935.
927.
936.
937.
928.
938.
929.
939.
930.
940.
941.
931.
932.
942.
943.
933.
934.
944.

and without autologous fat graft. A prospective randomized trial evaluated with reference to clinical and social factors. Archives of Orthopaedic & Trauma Surgery. Sep 2001;121(8):476-480. Chatterjee SF, Foy PMF, Findlay GFF. Report of a Controlled Clinical Trial Comparing Automated Percutaneous Lumbar Discectomy and Microdiscectomy in the Treatment of Contained Lumbar Disc Herniation. Spine. 1995;20(6):734-738. de Tribolet N, Porchet F, Lutz TW, et al. Clinical assessment of a novel antiadhesion barrier gel: prospective, randomized, multicenter, clinical trial of ADCON-L to inhibit postoperative peridural fibrosis and related symptoms after lumbar discectomy. American Journal of Orthopedics. 1998;27(2):111-120. Gambardella G, Gervasio O, Zaccone C, Puglisi E. Prevention of recurrent radicular pain after lumbar disc surgery: a prospective study. Acta Neurochirurgica - Supplement. 2005;92:151-154. Geisler FH. [Prevention of peridural fibrosis: current methodologies]. Neurological Research. 1999;21(Suppl 1):S9-S22. Haines SJ, Jordan N, Boen JR, Nyman JA, Oldridge NB, Lindgren BR. Discectomy strategies for lumbar disc herniation: Results of the LAPDOG trial. Journal of Clinical Neuroscience. 2002;9:411-417. Henriksen L, Schmidt V, Eskesen V, Jantzen E. A controlled study of microsurgical versus standard lumbar discectomy. Br J Neurosurg. 1996;10:289-293. Hermantin F, Peters T, Quartararo L, Kambin P. A prospective, randomized study comparing the results of open discectomy with those of video-assisted arthroscopic microdiscectomy. Journal of Bone & Joint Surgery. 1999;81(7):958-965. Huang J-j, Song Y-m, Liu L-m. [Clinical study of prevention of epidural scar and adhesion with polylactic acid membrane]. Chinese Journal of Reparative & Reconstructive Surgery. Jan 2004;18(1):21-24. Huang T-J, Hsu RW-W, Li Y-Y, Cheng C-C. Less systemic cytokine response in patients following microendoscopic versus open lumbar discectomy. Journal of Orthopaedic Research. Mar 2005;23(2):406-411. Jensen T, Asmussen K, Berg-Hansen E, et al. First-time operation for lumbar disc herniation with or without free fat transplantation. Prospective triple-blind randomized study with reference to clinical factors and enhanced computed tomographic scan 1 year after operation. Spine. 1996;21(9):1072-1076. Kim K, Wang J, Robertson D, et al. Reduction of leg pain and lower-extremity weakness for 1 year with Oxiplex/SP gel following laminectomy, laminotomy, and discectomy. Neurosurgical Focus. 2004;17(1):ECP1. Lagarrigue J, Chaynes P. Comparative study of disk surgery with or without microscopy. A prospective study of 80 cases. Neurochirurgie. 1994;40(2):116-120. MacKay M, Fischgrund J, Herkowitz H, Kurz LT, Hecht B, Schwartz M. The effect of interposition membrane on the outcome of lumbar laminectomy and discectomy. Spine. 1995;20(16):1793-1796. Mayer HM, Brock M. Percutaneous endoscopic discectomy: surgical technique and preliminary results compared to microsurgical discectomy. J Neurosurg. 1993;78:216-225. Paul M, Hellinger J. Nd-YAG (1064nm) versus diode (940nm) PLDN: a prospective randomised blinded study. Published in Proceedings of the First Interdisciplinary World Congress on Spinal Surgery and Related Disciplines, Berlin (Germany), August 27 - September 1, 2000; 555-8. Richter H, Kast E, Tomczak R, Besenfelder W, Gaus W. Results of applying ADCON-L gel after lumbar discectomy: the German ADCON-L study. Journal of Neurosurgery. Oct 2001;95(2 Suppl):179-189. Thome C, Barth M, Scharf J, Schmiedek P. Outcome after lumbar sequestrectomy compared with microdiscectomy: a prospective randomized study. Journal of Neurosurgery. 2005;Spine. 2(3):271-278. Tullberg T, Isacson J, Weidenhielm L. Does microscopic removal of lumbar disc herniation lead to better results than the standard procedure? Results of a one-year randomized study. Spine. 1993;18(1):24-27. Weinstein JN, Lurie JD, Tosteson TD, et al. Surgical vs nonoperative treatment for lumbar disk herniation: the Spine Patient Outcomes Research Trial (SPORT) observational cohort. JAMA. 2006;296(20):2451-2459. Atlas SJ, Deyo RA, Keller RB, et al. The Maine Lumbar Spine Study, Part II: 1-Year
945.
955.
946.
956.
957.
947.
958.
948.
959.
949.
950.
960.
951.
961.
952.
962.
953.
963.
954.
964.

Outcomes of Surgical and Nonsurgical Management of Sciatica. Spine. 1996;21(15):1777-1786. Atlas S, Keller R, Chang Y, Deyo R, Singer D. Surgical and Nonsurgical Management of Sciatica Secondary to a Lumbar Disc Herniation: Five-Year Outcomes From the Maine Lumbar Spine Study. Spine. 2001;26(10):1179-1187. Atlas S, Keller R, Wu Y, Deyo R, Singer D. Long-Term Outcomes of Surgical and Nonsurgical Management of Sciatica Secondary to a Lumbar Disc Herniation: 10 Year Results from the Maine Lumbar Spine Study. Spine. 2005;30(8):927-935. Nykvist F, Hurme M, Alaranta H, Kaitsaari M. Severe sciatica: a 13-year follow-up of 342 patients. Eur Spine J. 1995;4(6):335-338. Steffen R, Wittenberg R, Kraemer J. Chemonucleolysis versus laser disc decompression--a prospective randomised trial. J Bone Joint Surg. 1997;79-B:247. Barth M, Weiss C, Thome C. Two-year outcome after lumbar microdiscectomy versus microscopic sequestrectomy: part 1: evaluation of clinical outcome. Spine. 2008;33(3):265-272. Haines SJ, Jordan N, Boen JR, Nyman JA, Oldridge NB, Lindgren BR. Discectomy strategies for lumbar disc herniation: Study design and implications for clinical research. Journal of Clinical Neuroscience. 2002;9: 440-446. Malter A, Larson E, Urban N, Deyo R. CostEffectiveness of Lumbar Discectomy for the Treatment of Herniated Intervertebral Disc. Spine. 1996;21(9):1048-1054. Stevenson RC, McCabe CJ, Findlay AM. An economic evaluation of a clinical trial to compare automated percutaneous lumbar discectomy with microdiscectomy in the treatment of contained lumbar disc herniation. Spine. 1995;20:739-742. Corts Giner JR. Estudio DS 103-282: relajante muscular en lumbalgia aguda o lumbao (estudio doble ciego de tizanidina + paracetamol vs. placebo + paracetamol). Rev Esp de Cir Ost. 1989:119-124. Sirdalud Ternelin Asia-Pacific Study Group. Efficacy and gastroprotective effects of tizanidine plus diclofenac versus placebo plus dicofenac in patients with painful muscle spasms. Current Therapeutic Research. 1998;59:13-22. 975. Tervo T, Petaja L, Lepisto P. A controlled clinical trial of a muscle relaxants analgesic combination in the treatment of acute lumbago. The British Journal of Clinical Practice. 1976;30:62-64. Borenstein D, Lacks S, Wiesel S. Cyclobenzarpine and naproxen versus naproxen alone in the treatment of acute low back pain and muscle spasm. Clinical Therapeutics. 1990;12:125-131. Wand BM, Bird C, McAuley JH, Dore CJ, MacDowell M, De Souza LH. Early intervention for the management of acute low back pain: A single-blind randomized controlled trial of biopsychosocial education, manual therapy, and exercise. Spine. 2004;29(21):2350-2356. Wright A, Lloyd-Davies A, Williams S, Ellis R, Strike P. Individual active treatment combined with group exercise for acute and subacute low back pain. Spine. 2005;30(11):1235-1241. He RY. Clnical observation on treatment of lumbago due to cold-dampness by warmacupuncture plus Chinese medicine. Chin Acupuncture Moxibustion. 1997;17:279-280. Borman P, Keskin D, Bodur H. The efficacy of lumbar traction in the management of patients with low back pain. Rheumatology International. 2003;23(2):82-86. Nachemson A. Evaluation of results in lumbar spine surgery. Acta Orthop Scand Suppl. 1993;252:130-133. Trescot AM, Chopra P, Abdi S, Datta S, Schultz DM. Systematic review of effectiveness and complications of adhesiolysis in the management of chronic spinal pain: An update. Pain Physician. 2007;10:129-146. Chopra P, Smith H, Deer T, Bowman R. Role of adhesiolysis in the management of chronic spinal pain: A systematic review of effectiveness and complications. Pain Physician. 2005;8(1):87-100. NICE (National Institute for Clinical Excellence). Interventional procedure overview of endoscopic division of epidural adhesions. IP Guidance Number IPG088. National Institute for Clinical Excellence; 2004. Heavner JE, Racz GB, Raj P. Percutaneous epidural neuroplasty. Prospective evaluation of 0.9% NaCl versus 10% NaCl with or
965.
976.
966.
977.
967.
978.
968.
969.
979.
980.
970.
981.
971.
982.
972.
983.
973.
984.
974.
985.

without hyaluronidase. Reg Anesth Pain Med. 1999;24:202-207. Manchikanti L, Pampati V, Fellows B, Rivera JJ, Beyer CD, Damron KS. Role of one day epidural adhesiolysis in management of chronic low back pain: A randomzied clinical trial. Pain Physician. 2001;4:153-166. Veihelmann A, Devens C, Trouillier H, Birkenmaier C, Gerdesmeyer L, Refior HJ. Epidural neuroplasty versus physiotherapy to relieve pain in patients with sciatica: a prospective randomized blinded clinical trial. J Orthop Sci. 2006;11:365-369. Manchikanti L, Pampati V, Bakhit CE, Pakanati RR. Non-endoscopic and endoscopic adhesiolysis in post lumbar laminectomy syndrome. A one-year outcome study and cost effective analysis. Pain Physician. 1999;2:52-58. Geurts JW, Kallewaard JW, Richardson J, Groen GG. Targeted methylprednisolone acetate/hyaluronidase/clonidine injection after diagnostic epiduroscopy for chronic sciatica: a prospective, 1-year follow-up study. Reg Anesth Pain Med. 2002;27:343-352. Hassenbusch SJ, Portenoy RK, Cousins M, et al. Polyanalgesic consensus conference 2003: an update on the management of pain by intraspinal drug delivery--report of an expert panel. Journal of Pain and Symptom Management. 2004;27(6):540-563. Staats PS, Yearwood T, Charapata SG, et al. Intrathecal ziconotide in the treatment of refractory pain in patients with cancer or AIDS. A randomzied controlled trial. JAMA. 2004;291:63-70. Rauck RL, Wallace MS, Leong MS, et al. A randomzied, double-blind, placeob-controlled study of intrathecal ziconotide in adults with severe chronic pain. Journal of Pain and Symptom Management. 2006;31(5):393-406. Kumar K, Malik S, Demeria D. Treatment of chronic pain with spinal cord stimulation versus alternative therapies: costeffectiveness analysis. Neurosurgery. 2002;51:106-116. Deer T, Chapple I, Classen A, et al. Intrathecal drug delivery for treatment of chronic low back pain: report from the National Outcomes Registry for Low Back Pain. Pain Medicine. 2004;5(1):6-13. Prager JP. Neuraxial medication delivery: the development and maturity of a concept for treating chronic pain of spinal origin. Spine. 2002;27(22):2593-2605; discussion 2606. de Lissovoy G, Brown RE, Halpern M, Hassenbusch SJ, Ross E. Cost-effectiveness of long-term intrathecal morphine therapy for pain associated with failed back surgery syndrome. Clinical Therapeutics. 1997;19(1):96-112. Kumar K, Hunter G, Demeria DD. Treatment of chronic pain by using intrathecal drug therapy compared with conventional pain therapies: a cost-effectiveness analysis. J Neurosurg. 1997;97:803-810. Timm KE. A randomized-control study of active and passive treatments for chronic low back pain following L5 laminectomy. The Journal of Orthopaedic and Sports Physical Therapy. 1994;20(6):276-286. Miller B, Gatchel R, Lou L, Stowell A, Robinson R, Polatin P. Interdisciplinary treatment of failed back surgery syndrome (FBSS): A comparison of FBSS and nonFBSS patients. Pain Practice. 2005;5: 190-202. Dario A, Fortini G, Bertollo D, Bacuzzi A, Grizzetti C, Cuffari S. Treatment of failed back surgery syndrome. Neuromodulation. 2001;4:105-110. North RB, Kidd D, Shipley J, Taylor RS. Spinal cord stimulation versus reoperation for failed back surgery syndrome: a cost effectiveness and cost utility analysis based on a randomized, controlled trial. Neurosurgery. Aug 2007;61(2):361-368; discussion 368-369. Rossignol M, Abenhaim L, Seguin P, et al. Coordination of primary health care for back pain. A randomized controlled trial. Spine. 2000;25(2):251-258. Meeuwesen L, Huyse FJ, Koopmans GT, Heimans JJ. Supervised integrated screening of low-back pain patients by a neurologist. A randomized clinical trial. General Hospital Psychiatry. 1996;18(6):385-394. Linton SJ, van Tulder MW. Preventative interventions for back and neck pain problems. Spine. 2001;26(7):778-787. Glomsrod B, Lonn JH, Soukup MG, Bo K, Larsen S. Active back school, prophylactic management for low back pain: Three-year follow-up of a randomized controlled trial. J Rehabil Med. 2001;33:26-30. Hides J, Jull G, Richardson C. Long-term effects of specific stabilizing exercises for
986.
996.
987.
997.
988.
998.
999.
989.
1000.
990.
1001.
991.
1002.
992.
1003.
993.
1004.
994.
1005.
995.
1006.

first-episode low back pain. Spine. 2001;26(11):E243-248. Kellett K, Kellett D, Nordholm L. Effects of an exercise program on sick leave due to back pain. Phys Ther. 1991;71(4):283-291; discussion 291-283. Soukup M, Glomsrod B, Lonn J, Bo K, Larsen S. The Effect of a Mensendieck Exercise Program as Secondary Prophylaxis for Recurrent Low Back Pain: A Randomized, Controlled Trial With 12-Month Follow-up. Spine. 1999;24(15):1585. Verbeek J, W vd, van Dijk F. Early Occupational Health Management of Patients with Back Pain: A Randomized Controlled Trial. Spine. 2002;27(17):1844-1850. Vleeming A, Albert HB, Ostgaard HC, Stuge B, Sturesson B, On behalf of the Cost B. Working Group on Guidelines for the Management of Acute Low Back Pain in Primary Care. European guidelines on the diagnosis and treatment of pelvic girdle pain. European Commission, Research Directorate-General, Cost Action B13, Working Group 4; 2004. Wedenberg K, Moen B, Norling A. A prospective randomized study comparing acupuncture with physiotherapy for low-back and pelvic pain in pregnancy. Acta obstetricia et gynecologica Scandinavica. 2000;79(5):331-335. Stuge B, Hilde G, Vollestad N. Physical therapy for pregnancy-related low back and pelvic pain: a systematic review. Acta Obstetricia et Gynecologica Scandinavica. 2003;82(11):983-990. Guerreiro da Silva JB, Nakamura MU, Cordeiro JA, Kulay L, Jr. Acupuncture for low back pain in pregnancy--a prospective, quasirandomised, controlled study. Acupuncture in Medicine. 2004;22(2):60-67. Kvorning N, Holmberg C, Grennert L, Aberg A, Akeson J. Acupuncture relieves pelvic and low-back pain in late pregnancy. Acta Obstetricia et Gynecologica Scandinavica. 2004;83(3):246-250. Kihlstrand M, Stenman B, Nilsson S, Axelsson O. Water-gymnastics reduced the intensity of back/low pain in pregnant women. Acta Obstetricia Gynecologica Scandinavica. 1999;78(3):180-185. Suputtitada A, Wacharapreechanont T, Chaisayan P. Effect of the "sitting pelvic tilt exercise" during the third trimester in primigravidas on back pain. Chotmaihet thangphaet. 2002;85(1). Noren L, Ostgaard S, Nielsen TF, Ostgaard HC. Reduction of sick leave for lumbar bak and posterior pelvic pain in pregnancy. Spine. 1997;22:2157-2160. Ostgaard HC, Zetherstrom G, Roos-Hansson E, Svanberg B. Reduction of back and posterior pelvic pain in pregnancy. Spine. 1994;19:894-900. Mantle MJ, Holmes J, Currey HL. Backache in pregnancy II. Prophylactic influence of back care classes. Rheumatol Rehabil. 1981;20:227-232. Dumas GA, Reid JG, Wolfe LA, Griffin MP. Exercise, posture, and back pain during pregnancy. Part 2. Exercise and back pain. Clin Biomechanics. 1995;10:104-109. Ostgaard HC, Zetherstrom G, Roos-Hansson E. Back pain in relation to pregnancy: a 6year follow-up. Spine. 1997;22(24):29452950. Field T, Hernandez-Reif M, Hart S, Theakston H, Schanberg S, Kuhn C. Pregnant women benefit from massage therapy. Journal of Psychosomatic Obstetrics and Gynaecology. 1999;20(1):31-38. Field T, Diego MA, Hernandex-Reif M, Schanberg S, Kuhn C. Massage therapy effects on depressed pregnant women. J Psychosom Obstet Gynecol. 2004;25:115122. Young G, Jewell D. Interventions for preventing and treating pelvic and back pain in pregnancy Cochrane Database of Systematic Reviews. 2005(1):Art. No.: CD001139. Thomas I, Nicklin J, Pollock H, Faulkner K. Evaluation of a maternity cushion (Ozzlo pillow) for backache and insomnia in late pregnancy. Australian & New Zealand Journal of Obstetrics & Gynaecology. 1989;29(2):133-138. Baldwin ML, Cote P, Frank JW, Johnson WG. Cost-effectiveness studies of medical and chiropractic care for occupational low back pain. a critical review of the literature. The Spine Journal. 2001;1(2):138-147. Soegaard R, Christensen FB. Health economic evaluation in lumbar spinal fusion: a systematic literature review anno 2005. Eur Spine J. 2006;15:1165-1173. Thomsen AB, Sorensen J, Sjogren P, Eriksen J. Economic evaluation of multidisciplinary
1007.
1017.
1008.
1018.
1019.
1009.
1020.
1010.
1021.
1022.
1011.
1023.
1012.
1024.
1013.
1025.
1014.
1026.
1015.
1027.
1016.
1028.

pain management in chronic pain patients: a qualitative systematic review. Journal of Pain & Symptom Management. 2001;22(2): 688-698. Uhlig T, Finset A, Kvien T. Effectiveness and cost-effectiveness of comprehensive rehabilitation programs. Current Opinion in Rheumatology. 2003;15(2):134-140. van der Roer N, Goossens ME, Evers SM, et al. What is the most cost-effective treatment for patients with low back pain? A systematic review. Best Practice & Research in Clinical Rheumatology. 2005;19(4):671-684. Ioannidis JP, Evans SJ, Gotzsche PC, et al. Better reporting of harms in randomized trials: an extension of the CONSORT statement. Ann Intern Med. 2004;141(10):781-788. van Tulder MW, Cherkin DC, Berman B, Lao L, Koes BW. Acupuncture for low-back pain. Cochrane Database of Systematic Reviews. 1999(2):Art. No.: CD001351. Lipetz J, Malanga G. Oral medications in the treatment of acute low back pain. Occupational Medicine. Occupational Medicine.1998;13(1):151-166. Ammendolia C, Furlan AD, Imamura M, Irvin E, van Tulder M. Evidence-informed management of chronic low back pain with needle acupuncture. The Spine Journal. 2008;8(1):160-172. Bronfort G, Haas M, Evans R, Kawchuk G, Dagenais S. Evidence-informed management of chronic low back pain with spinal manipulation and mobilization. The Spine Journal. 2008;8(1):213-225. Brox JI, Storheim K, Grotle M, Tveito TH, Indahl A, Eriksen HR. Evidence-informed management of chronic low back pain with back schools, brief education, and fearavoidance training. The Spine Journal. 2008;8(1):28-39. Brox JI, Storheim K, Grotle M, Tveito TH, Indahl A, Eriksen HR. Systematic review of back schools, brief education, and fearavoidance training for chronic low back pain. Spine Journal. 2008;8:948-958. Carreon LY, Glassman SD, Howard J. Fusion and nonsurgical treatment for symptomatic lumbar degenerative disease: a systematic review of Oswestry Disability Index and MOS Short Form-36 outcomes. Spine Journal. 2008;8:747-755. 1039. Chrubasik JE, Roufogalis BD, Chrubasik S. Evidence of effectiveness of herbal antiinflammatory drugs in the treatment of painful osteoarthritis and chronic low back pain. Phytotherapy Research. 2007;21: 675-683. Clarke JA, van Tulder MW, Blomberg SE, et al. Traction for low-back pain with or without sciatica. Cochrane Database of Systematic Reviews. 2007(2):Art. No.: CD003010. Dagenais S, Mayer J, Haldeman S, BorgStein J. Evidence-informed management of chronic low back pain with prolotherapy. The Spine Journal. 2008;8(1):203-212. Dagenais S, Mayer J, Wooley JR, Haldeman S. Evidence-informed management of chronic low back pain with medicine-assisted manipulation. The Spine Journal. 2008;8(1):142-149. Deshpande A, Furlan A, Mailis-Gagnon A, Atlas S, Turk D. Opioids for chronic low-back pain. Cochrane Database of Systematic Reviews. 2007(3):Art. No.: CD004959. Engers A, Jellema P, Wensing M, van der Windt DAWM, Grol R, van Tulder MW. Individual patient education for low back pain. Cochrane Database of Systematic Reviews. 2008(1):Art. No.: CD004057. Freeman BJ, Mehdian R. Intradiscal electrothermal therapy, percutaneous discectomy, and nucleoplasty: what is the current evidence? Current Pain & Headache Reports. 2008;12:14-21. Furlan AD, Imamura M, Dryden T, E. I. Massage for low-back pain. Cochrane Database of Systematic Reviews. 2008(4):Art. No.: CD001929. Gagnier JJ. Evidence-informed management of chronic low back pain with herbal, vitamin, mineral, and homeopathic supplements. The Spine Journal. 2008;8(1):70-79. Gatchel RJ, Mayer TG. Evidence-informed management of chronic low back pain with functional restoration. The Spine Journal. 2008;8(1):65-69. Gay RE, Brault JS. Evidence-informed management of chronic low back pain with traction therapy. The Spine Journal. 2008;8(1):234-242. Henchoz Y, Kai-Lik So A. Exercise and nonspecific low back pain: a literature review. Joint, Bone, Spine: Revue du Rhumatisme. 2008;75:533-539.
1029.
1040.
1030.
1041.
1031.
1042.
1032.
1043.
1033.
1044.
1034.
1045.
1035.
1046.
1036.
1047.
1048.
1037.
1049.
1038.
1050.

1051. Imamura M, Furlan AD, Dryden T, Irvin E. Evidence-informed management of chronic low back pain with massage. The Spine Journal. 07.16.08 2008;8(1):121-133. Khadilkar A, Odebiyi DO, Brosseau L, Wells GA. Transcutaneous electrical nerve stimulation (TENS) versus placebo for chronic low-back pain. Cochrane Database of Systematic Reviews. 2008(4):Art. No.: CD00308. Lawrence DJ, Meeker W, Branson R, et al. Chiropractic management of low back pain and low back-related leg complaints: a literature synthesis. Journal of Manipulative & Physiological Therapeutics. 2008;31: 659-674. Liddle SD, Gracey JH, Baxter GD. Advice for the management of low back pain: a systematic review of randomised controlled trials. Manual Therapy. 2007;12:310-327. Machado LAC, Kamper SJ, Herbert RD, Maher CG, McAuley JH. Analgesic effects of treatments for non-specific low back pain: a meta-analysis of placebo-controlled randomized trials. Rheumatology. 2009 May;48(5):520-7. Epub 2008 Dec 24. Macedo LG, Maher CG, Latimer J, McAuley JH. Motor control exercise for persistent, nonspecific low back pain: a systematic review. Physical Therapy. 2009;89:9-25. Martell BA, O'Connor PG, Kerns RD, et al. Systematic review: Opioid treatment for chronic back pain: Prevalence, efficacy, and association with addiction. Ann Intern Med. 2007;146:116-127. Mayer J, Mooney V, Dagenais S. Evidenceinformed management of chronic low back pain with lumbar extensor strengthening exercises. The Spine Journal. 2008;8(1): 96-113. Norlund A, Ropponen A, Alexanderson K. Multidisciplinary interventions: review of studies of return to work after rehabilitation for low back pain. Journal of Rehabilitation Medicine. 2009;41:115-121. Novak S, Nemeth WC. The basis for recommending repeating epidural steroid injections for radicular low back pain: a literature review. Archives of Physical Medicine & Rehabilitation. 2008;89:543-552. Pennick V. Interventions for preventing and treating pelvic and back pain in pregnancy. Cochrane Database of Systematic Reviews. 2007(2):Art. No.: CD001139. 1062. Perrot S, Javier RM, Marty M, Le Jeunne C, Laroche F. Is there any evidence to support the use of anti-depressants in painful rheumatological conditions? Systematic review of pharmacological and clinical studies. Rheumatology. 2008;47:1117-1123. Poiraudeau S, Rannou F, Revel M. Functional restoration programs for low back pain: a systematic review. Annales de Readaptation et de Medecine Physique. 2007;50:419-424. Poitras S, Brosseau L. Evidence-informed management of chronic low back pain with transcutaneous electrical nerve stimulation, interferential current, electrical muscle stimulation, ultrasound, and thermotherapy. The Spine Journal. 2008;8(1):226-233. Racz GB, Heavner JE, Trescot A. Percutaneous lysis of epidural adhesions-evidence for safety and efficacy. Pain Practice. 2008;8:277-286. Roelofs, P. D. D. M., Deyo RA, Koes BW, Scholten RJPM, van Tulder MW. Nonsteroidal anti-inflammatory drugs for low back pain. Spine. 2008;33:1766-1774. Roelofs PDDM, Deyo RA, Koes BW, Scholten RJPM, van Tulder MW. Nonsteroidal anti-inflammatory drugs for low back pain. Cochrane Database of Systematic Reviews. 2008(1):Art. No.: CD000396. Sahar T, Cohen MJ, Ne'eman V, et al. Insoles for prevention and treatment of back pain. Cochrane Database of Systematic Reviews. 2007(4):Art. No.: CD005275. Slade SC, Keating JL. Unloaded movement facilitation exercise compared to no exercise or alternative therapy on outcomes for people with nonspecific chronic low back pain: a systematic review. Journal of Manipulative & Physiological Therapeutics. 2007;30: 301-311. Standaert CJ, Weinstein SM, Rumpeltes J. Evidence-informed management of chronic low back pain with lumbar stabilization exercises. The Spine Journal. 07.16.08 2008;8(1):114-120. Stuber KJ, Smith DL. Chiropractic treatment of pregnancy-related low back pain: a systematic review of the evidence. Journal of Manipulative & Physiological Therapeutics. 2008;31:447-454. Urquhart DM, Hoving JL, Assendelft WWJJ, Roland M, van Tulder MW. Antidepressants for non-specific low back pain. Cochrane
1052.
1063.
1053.
1064.
1054.
1065.
1055.
1066.
1067.
1056.
1057.
1068.
1069.
1058.
1059.
1070.
1060.
1071.
1061.
1072.

Database of Systematic Reviews. 2008(1):Art. No.: CD001703. van Duijvenbode ICD, Jellema P, van Poppel MNM, van Tulder MW. Lumbar supports for prevention and treatment of low back pain. Cochrane Database of Systematic Reviews. 2008(2):CD001823. van Geen J, MJA E, Janssen ME, van Eijk JT. The long-term effect of multidisciplinary back training. Spine. 2007;32:249-255. Vlachojannis J, Roufogalis BD, Chrubasik S. Systematic review on the safety of Harpagophytum preparations for osteoarthritic and low back pain. Phytotherapy Research. 2008;22:149-152. Wai EK, Rodriguez S, Dagenais S, Hall H. Evidence-informed management of chronic low back pain with physical activity, smoking cessation, and weight loss. The Spine Journal. 2008;8(1):195-202. Williams NH, Hendry M, Lewis R, Russell I, Westmoreland A, Wilkinson C. Psychological response in spinal manipulation (PRISM): a systematic review of psychological outcomes in randomised controlled trials. Complement Ther Med. 2007;15:271-283. 1078. Williams RM, Westmorland MG, Lin CA, Schmuck G, Creen M. Effectiveness of workplace rehabilitation interventions in the treatment of work-related low back pain: a systematic review. Disability & Rehabilitation. 2007;29:608-624. Wolfer LR, Derby R, Lee J-E, Lee S-H. Systematic review of lumbar provocation discography in asymptomatic subjects with a meta-analysis of false-positive rates. Pain Physician. 2008;4:513-538. Yousefi-Nooraie R, Schonstein E, Heidari K, et al. Low level laser therapy for nonspecific low-back pain. Cochrane Database of Systematic Reviews. 2007(2):Art. No.: CD005107. Yuan J, Purepong N, Kerr DP, Park J, Bradbury I, McDonough S. Effectiveness of acupuncture for low back pain. Spine. 2008;33:E887-E900. Yuan J, Kerr D, Park J, Liu XH, McDonough S. Treatment regimens of acupuncture for low back pain--a systematic review. Complement Ther Med. 2008;16:295-304.
1073.
1079.
1074.
1075.
1080.
1076.
1081.
1077.
1082.

APPENDIX 1: SYSTEMATIC REVIEWS SEARCH STRATEGIES

ALL SYSTEMATIC REVIEWS SEARCHES 2005-2008 (updated yearly) Database: Ovid MEDLINE 1 ((ache$ or pain$) adj2 (low back or lower back or lumbar)).mp. 2 lbp.mp. 3 exp Back Pain/ 4 (1 or 2) and 3 5 low back pain/ 6 4 or 5 7 limit 6 to humans 8 limit 7 to "all adult (19 plus years)" 9 meta-analysis.mp. or exp Meta-Analysis/ 10 (cochrane or medline).tw. 11 search$.tw. 12 9 or 10 or 11 13 "Review Literature as Topic"/ or systematic review.mp. 14 12 or 13 15 8 and 14 16 from 15 keep ALL Database: Cochrane Database of Systematic Reviews 1 cochrane back group.gn. 2 back pain.ti. 3 1 or 2 4 from 3 keep ALL

APPENDIX 2: PRIMARY STUDIES SEARCH STRATEGIES

PRIMARY CARE SEARCHES 2005 Basic search strategy of Cochrane Central Register of Controlled Trials, through 4th Quarter 2005 1 randomized controlled trial.pt. 2 controlled clinical trial.pt. 3 Randomized Controlled Trials/ 4 Random Allocation/ 5 Double-Blind Method/ 6 Single-Blind Method/ 7 1 or 2 or 3 or 4 or 5 or 6 8 animal/ not human/ 9 7 not 8 10 clinical trial.pt. 11 exp clinical trials/ 12 (clinic$ adj25 trial$).tw. 13 ((singl$ or doubl$ or trebl$ or tripl$) adj (mask$ or blind$)).tw. 14 placebos/ 15 placebo$.tw. 16 random$.tw. 17 research design/ 18 (latin adj square).tw. 19 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 20 19 not 8 21 20 not 9 22 comparative study/ 23 exp evaluation studies/ 24 follow-up studies/ 25 prospective studies/ 26 (control$ or prospective$ or volunteer$).tw. 27 cross-over studies/ 28 22 or 23 or 24 or 25 or 26 or 27 29 28 not 8 30 29 not (9 or 21) 31 9 or 21 or 30 32 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 33 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 34 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 35 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 36 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 37 34 or 35 or 36 38 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 39 (32 or 33) and 38 40 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 41 37 or 39 or 40


PRIMARY CARE SEARCHES 2005 Unique intervention search steps (through 4th quarter 2005): Acetaminophen 42 Acetaminophen.mp. or exp ACETAMINOPHEN/ 43 paracetamol.mp. 44 42 or 43 45 41 and 44 46 from 45 keep ALL Aspirin 42 acetylsalicylic acid.mp. 43 aspirin$.mp. or exp ASPIRIN/ 44 42 or 43 45 41 and 44 46 from 45 keep ALL COX-2 42 rofecoxib.mp. 43 valdecoxib.mp. 44 celecoxib.mp. 45 etoricoxib.mp. 46 lumiracoxib.mp. 47 ((cox-2 or cyclooxygenase-2) adj5 inhib$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 48 42 or 43 or 44 or 45 or 46 or 47 49 41 and 48 50 from 49 keep ALL Opioids 42 (Opioid$ or Narcotic$).mp. 43 41 and 42 44 from 43 keep ALL
Basic search strategy of Ovid MEDLINE , 1966 to September Week 3 2005 1 randomized controlled trial.pt. 2 controlled clinical trial.pt. 3 Randomized Controlled Trials/ 4 Random Allocation/ 5 Double-Blind Method/ 6 Single-Blind Method/ 7 1 or 2 or 3 or 4 or 5 or 6 8 animal/ not human/ 9 7 not 8 10 clinical trial.pt. 11 exp clinical trials/ 12 (clinic$ adj25 trial$).tw. 13 ((singl$ or doubl$ or trebl$ or tripl$) adj (mask$ or blind$)).tw. 14 placebos/ 15 placebo$.tw. 16 random$.tw. 17 research design/ 18 (latin adj square).tw. 19 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 20 19 not 8 21 20 not 9 22 comparative study/


PRIMARY CARE SEARCHES 2005 23 exp evaluation studies/ 24 follow-up studies/ 25 prospective studies/ 26 (control$ or prospective$ or volunteer$).tw. 27 cross-over studies/ 28 22 or 23 or 24 or 25 or 26 or 27 29 28 not 8 30 29 not (9 or 21) 31 9 or 21 or 30 32 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 33 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 34 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 35 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 36 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 37 34 or 35 or 36 38 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 39 (32 or 33) and 38 40 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 41 37 or 39 or 40 Unique intervention search steps (1966 to September Week 3 2005): Aceteminophen 42 Acetaminophen.mp. or exp ACETAMINOPHEN/ 43 paracetamol.mp. 44 42 or 43 45 41 and 44 46 31 and 45 47 from 46 keep ALL Aspirin 42 acetylsalicylic acid.mp. 43 aspirin$.mp. or exp ASPIRIN/ 44 42 or 43 45 41 and 44 46 31 and 45 47 45 not 46 48 from 47 keep ALL Corticosteroids 42 Corticosteroid$.mp. or exp Adrenal Cortex Hormones/ 43 41 and 42 44 31 and 43 45 43 not 44 46 from 45 keep ALL COX-2 42 rofecoxib.mp. 43 valdecoxib.mp. 44 celecoxib.mp. 45 etoricoxib.mp. 46 lumiracoxib.mp.


PRIMARY CARE SEARCHES 2005 47 ((cox-2 or cyclooxygenase-2) adj5 inhib$).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 48 42 or 43 or 44 or 45 or 46 or 47 49 41 and 48 50 from 49 keep ALL Hydrotherapy 42 exp Hydrotherapy/ or Water therapy.mp. or Balneology/ 43 (balneotherapy or hydrotherapy).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 44 42 or 43 45 41 and 44 46 limit 45 to english language 47 limit 45 to abstracts 48 46 or 47 49 from 48 keep ALL Mattresses 42 exp "Bedding and Linens"/ or exp Beds/ or mattress$.mp. 43 41 and 42 44 from 43 keep ALL Opioids 42 Opioid$.mp. or exp Narcotics/ or narcotic$.mp. 43 41 and 42 44 31 and 43 45 43 not 44 46 from 45 keep ALL Superficial heat or cold 42 exp HEAT/tu[Therapeutic Use] 43 (((heat or heats or heated or heating or electrotherm$ or therm$ or warm$) adj3 (therap$ or treat$ or procedure$)) or thermother$).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 44 42 or 43 45 41 and 44 46 limit 45 to humans 47 31 and 46 48 46 not 47 49 from 48 keep 1-104 (104) Tramadol 42 tramadol.mp. or exp TRAMADOL/ 43 41 and 42 44 from 43 keep ALL Yoga 42 yoga.mp. or exp Yoga/ 43 41 and 42 44 31 and 43 45 43 not 44 46 from 45 keep ALL


PRIMARY CARE SEARCHES 2005 Basic search strategy of CINAHL (Cumulative Index to Nursing & Allied Health Literature), 1982 to September Week 1 2005 1 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 2 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 3 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 4 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 5 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, subject heading word, abstract, instrumentation] 6 3 or 4 or 5 7 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, subject heading word, abstract, instrumentation] 8 (1 or 2) and 7 9 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, subject heading word, abstract, instrumentation] 10 6 or 8 or 9 Unique intervention search steps (1982 to September Week 1 2005): Antiepileptic drugs (Gabapentin) 11 gabapentin.mp. 12 10 and 11 13 from 12 keep ALL Botulinum toxin 11 (botox or (Botuli$)).mp. {mp=title, subject heading word, abstract, instrumentation] 12 10 and 11 13 from 12 keep ALL Low level lasers 11 ((laser$ adj3 (therap$ or treat$)) or lllt).mp. or exp LASERS/tu[mp=title, subject heading word, abstract, instrumentation] 12 ((HeNe or IR or diode or infrared) adj laser$).mp.[mp=title, subject heading word, abstract, instrumentation] 13 (GaA1As or GaAs or Nd:YAG).mp.[mp=title, subject heading word, abstract, instrumentation] 14 ((low or lower or lowest or lowering) adj3 laser$).mp.[mp=title, subject heading word, abstract, instrumentation] 15 11 or 12 or 13 or 14 16 10 and 15 17 from 16 keep ALL Mattresses 11 exp "Bedding and Linens"/ or exp Beds/ or mattress$.mp. 12 10 and 11 13 from 12 keep ALL Superficial heat or cold 11 exp HEAT/tu[Therapeutic Use] 12 (((heat or heats or heated or heating or electrotherm$ or therm$ or warm$) adj3 (therap$ or treat$ or procedure$)) or thermother$).mp.[mp=title, subject heading word, abstract, instrumentation] 13 11 or 12 14 10 and 13 15 from 14 keep ALL


PRIMARY CARE SEARCHES 2005 Tramadol 11 tramadol.mp. or exp TRAMADOL/ 12 10 and 11 13 from 12 keep ALL Ultrasound 11 exp Ultrasonics/ or exp Ultrasonic Therapy/ 12 ((ultrasound$ or ultrason$) adj2 (treat$ or therap$)).mp.[mp=title, subject heading word, abstract, instrumentation] 13 11 or 12 14 10 and 13 15 from 14 keep ALL


PRIMARY CARE SEARCHES 2006 Basic search strategy of Cochrane Central Register of Controlled Trials, through 1st Quarter 2006 1 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 2 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 3 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 4 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 5 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 6 3 or 4 or 5 7 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 8 (1 or 2) and 7 9 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 10 6 or 8 or 9 Unique intervention search steps (through 1st Quarter 2006): Antidepressants 11 duloxetine.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 from 12 keep ALL Coordination of care 11 (coordinat$ or integrat$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 from 12 keep ALL Corticosteroids 11 (Corticosteroid$ or steroid or steroids).mp. or exp Adrenal Cortex Hormones/[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 from 12 keep ALL Cost benefit 11 ((economic$ or financ$ or cost or costs or costing) adj2 (benefit$ or effectiv$ or evaluat$ or analyz$ or analys$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 from 12 keep ALL Decision tools 11 ((rule$ or tool$) adj2 (decision$ or decid$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 ((decision$ or decid$) adj2 (support$ or confirm$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 14 10 and 13 15 (decision$ or decid$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 16 10 and 15 17 from 16 keep ALL


PRIMARY CARE SEARCHES 2006 Interdisciplinary 11 ((primary or family or general) adj2 (care or practitioner$ or physician$ or practice$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 (specialist$ or specialty or orthopedics or orthopedist$ or neurologist$ or neurosurgeon$ or surgeon$ or chiropractic$ or occupational therapist$ or physiotherapist$ or physical therapist$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 14 10 and 13 15 (referral$ or consult$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 16 (interdisciplin$ or multidisciplin$ or inter-disciplin$ or multi-disciplin$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 17 11 and 13 18 11 and 15 19 11 and 16 20 13 and 15 21 13 and 16 22 15 and 16 23 17 or 18 or 19 or 20 or 21 or 22 24 10 and 23 25 from 24 keep 1-113 (113) Intrathecal 11 ((fail$ or unsuccessful$ or ineffectiv$) adj2 (surger$ or surgic$ or operation$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 (fbss or flss).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 14 10 and 13 15 12 or 14 16 ((fail$ or unsuccessful$ or ineffectiv$) adj3 (back or lumbar)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 17 15 or 16 18 (reoperat$ or re-operat$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 19 10 and 18 20 (fail$ or unsuccessful$ or ineffectiv$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 21 19 and 20 22 17 or 21 23 from 22 keep 1-48 24 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 25 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 26 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 27 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 28 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 29 26 or 27 or 28 30 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 31 (24 or 25) and 30 32 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 33 29 or 31 or 32 34 intrathecal$.mp. 35 33 and 34 36 from 35 keep ALL


PRIMARY CARE SEARCHES 2006 Mattresses 11 exp "Bedding and Linens"/ or exp Beds/ or mattress$.mp. 12 10 and 11 13 from 12 keep ALL Pregnancy 11 (pregnanc$ or pregnant or prenatal$ or postpartum$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 from 12 keep ALL Self care 11 ((self or selves or themsel$) adj3 (care or look after)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 (patient$ adj3 (informed or information or informing or educat$ or teach$ or learn$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 13 11 or 12 14 10 and 13 15 from 14 keep ALL Superficial heat or cold 11 exp HEAT/tu[Therapeutic Use] 12 (((heat or heats or heated or heating or electrotherm$ or therm$ or warm$) adj3 (therap$ or treat$ or procedure$)) or thermother$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 13 11 or 12 14 10 and 13 15 from 14 keep ALL Tramadol 11 tramadol.mp. or exp TRAMADOL/ 12 10 and 11 13 from 12 keep ALL Basic search strategy of Ovid MEDLINE, 1966 to February Week 3 2006 1 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 2 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 3 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 4 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 5 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 6 3 or 4 or 5 7 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 8 (1 or 2) and 7 9 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 10 6 or 8 or 9


PRIMARY CARE SEARCHES 2006 Unique intervention search steps (1966 to February Week 3 2006): Coordination of care 11 ((coordinat$ or integrat$) adj3 (care or caring or cares or therap$)).mp. 12 10 and 11 13 from 12 keep ALL Cost Benefit 11 exp Cost-Benefit Analysis/ 12 10 and 11 13 ((economic$ or financ$ or cost or costs or costing) adj2 (benefit$ or effectiv$ or evaluat$)).mp. 14 10 and 13 15 12 or 14 16 limit 15 to english language 17 limit 15 to abstracts 18 16 or 17 19 from 18 keep ALL Decision tools 11 exp decision support techniques/ 12 Clinical decision rule$.mp. 13 ((rule$ or tool$) adj2 (decision$ or decid$)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 14 11 or 12 or 13 15 10 and 14 16 limit 15 to humans 17 limit 16 to english language 18 limit 16 to abstracts 19 17 or 18 20 from 19 keep ALL Diagnostic nerve blocks 11 ((nerve$ or nervous) adj2 block$).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 12 10 and 11 13 exp Diagnosis/ 14 12 and 13 15 (di or ra or us or pa).fs. 16 10 and 11 and 15 17 14 or 16 18 limit 17 to english language 19 limit 17 to abstracts 20 18 or 19 21 limit 20 to humans 22 from 21 keep ALL Multidisciplinary 11 exp Primary Health Care/ 12 exp Physicians, Family/ 13 exp Family Practice/ 14 11 or 12 or 13 15 exp Orthopedics/ 16 exp neurology/ 17 exp surgery/ 18 exp chiropractic/ 19 exp "Physical Therapy (Specialty)"/ 20 exp Occupational Therapy/


PRIMARY CARE SEARCHES 2006 21 15 or 16 or 17 or 18 or 19 or 20 22 exp "Referral and Consultation"/ 23 (interdisciplin$ or multidisciplin$ or inter-disciplin$ or multi-disciplin$).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 24 exp Complementary Therapies/ 25 exp "Outcome and Process Assessment (Health Care)"/ 26 Combined Modality Therapy/ 27 10 and 14 28 10 and 21 29 10 and (23 or 26) 30 27 and 28 31 27 and 29 32 28 and 29 33 30 or 31 or 32 34 27 and 22 35 28 and 22 36 29 and 22 37 34 or 35 or 36 38 27 and 24 39 28 and 24 40 29 and 24 41 38 or 39 or 40 42 33 or 37 or 41 43 limit 42 to english language 44 limit 42 to abstracts 45 43 or 44 46 from 45 keep ALL Pregnancy 11 exp postpartum period/ or exp pregnancy/ 12 10 and 11 13 (th or dt or dh or pc).fs. 14 12 and 13 15 exp Pregnancy Complications/ 16 10 and 15 17 13 and 16 18 14 or 17 19 limit 18 to english language 20 limit 18 to abstracts 21 19 or 20 22 from 21 keep ALL Special search strategies of Ovid MEDLINE, 1966 to February Week 3 2006 Self care 1 exp Self Care/ 2 exp Health Education/ 3 exp Back Pain/ 4 ((self or selves or themsel$) adj3 (care or look after)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 5 (patient$ adj3 (informed or information or informing or educat$ or teach$ or learn$)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 6 1 or 2 7 3 and 6 8 4 or 5 9 3 and 8 10 7 or 9 11 limit 10 to english language


PRIMARY CARE SEARCHES 2006 12 13 14 limit 10 to abstracts 11 or 12 from 13 keep ALL
Search strategies of PEDro (Physiotherapy Evidence Database/Centre of Evidence-Based Physiotherapy [Cochrane Collaboration]), through 4th quarter 2006 Hyrdotherapy 1 THERAPY = hyrdrotherapy, balneotherapy 2 PROBLEM = pain 3 BODY PART = lumbar spine, sacroiliac or pelvis 4 Keep ALL Low level lasers 1 ABSTRACT/TITLE = laser 2 THERAPY = none selected 3 PROBLEM = pain 4 BODY PART = lumbar spine, sacroiliac or pelvis 5 Keep ALL Mattresses 1 ABSTRACT/TITLE = mattress 2 THERAPY = none selected 3 PROBLEM = pain 4 BODY PART = lumbar spine, sacroiliac or pelvis 5 Keep ALL Superficial heat or cold 1 THERAPY = electrotherapies, heat or cold 2 PROBLEM = pain 3 BODY PART = lumbar spine, sacroiliac or pelvis 4 Keep ALL


PRIMARY CARE SEARCHES 2007 Basic search strategy of Cochrane Central Register of Controlled Trials, through 1st Quarter 2007 1 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 2 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 3 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 4 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 5 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 6 3 or 4 or 5 7 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 8 (1 or 2) and 7 9 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 10 6 or 8 or 9 Unique intervention search steps (through 1st Quarter 2007): Antiepileptic drugs 11 gabapentin.mp. 12 (anticonvulsant$ or anti-convulsant$).mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 13 topiramate.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 14 valproic acid.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 15 pregabalin.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 16 lamotrigine.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 17 carbamazepine.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 18 oxcarbazepine.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 19 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 20 10 and 19 21 from 20 keep ALL Lidocaine 11 lidocaine.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 12 10 and 11 13 from 12 keep ALL Basic search strategy of Ovid MEDLINE, 1950 to February Week 3 2007 1 randomized controlled trial.pt. 2 controlled clinical trial.pt. 3 Randomized Controlled Trials/ 4 Random Allocation/ 5 Double-Blind Method/ 6 Single-Blind Method/ 7 1 or 2 or 3 or 4 or 5 or 6 8 animal/ not human/ 9 7 not 8 10 clinical trial.pt. 11 exp clinical trials/ 12 (clinic$ adj25 trial$).tw. 13 ((singl$ or doubl$ or trebl$ or tripl$) adj (mask$ or blind$)).tw. 14 placebos/ 15 placebo$.tw. 16 random$.tw. 17 research design/


PRIMARY CARE SEARCHES 2007 18 (latin adj square).tw. 19 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 20 19 not 8 21 20 not 9 22 comparative study/ 23 exp evaluation studies/ 24 follow-up studies/ 25 prospective studies/ 26 (control$ or prospective$ or volunteer$).tw. 27 cross-over studies/ 28 22 or 23 or 24 or 25 or 26 or 27 29 28 not 8 30 29 not (9 or 21) 31 9 or 21 or 30 32 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ 33 (spine or coccyx or intervertebral disk$ or lumbar vertebrae or sacrum or spinal canal or back).tw. 34 spinal diseases/ or intervertebral disk displacement/ or spinal curvatures/ or kyphosis/ or lordosis/ or scoliosis/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 35 (spinal disease$ or spinal curvatur$ or kyphosis or lordosis or scoliosis or spinal osteophytosis or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 36 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 37 34 or 35 or 36 38 exp pain/ or pain$.mp. or ache.mp. or aches.mp. or aching.mp. or ached.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 39 (32 or 33) and 38 40 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 41 37 or 39 or 40 Unique intervention search steps (1950 to February Week 3 2007): Antiepileptic drugs (Gabapentin) 42 gabapentin.mp. 43 exp gamma-Aminobutyric Acid/ 44 exp Cyclohexanecarboxylic Acids/ 45 exp AMINES/ 46 43 and 44 and 45 47 topiramate.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 48 valproic acid.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 49 pregabalin.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 50 lamotrigine.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 51 carbamazepine.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 52 oxcarbazepine.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 53 (anticonvulsant$ or anti-convulsant$).mp.[mp=title, original title, abstract, name of substance word, subject heading word] 54 42 or 46 or 47 or 48 or 49 or 50 or 51 or 52 or 53 55 41 and 54 56 31 and 55 57 55 not 56 58 from 57 keep ALL


PRIMARY CARE SEARCHES 2007 Antidepressants 42 duloxetine.mp. 43 41 and 42 44 from 43 keep ALL 42 venlafaxine.mp. 43 41 and 42 44 from 43 keep ALL Lidocaine 42 exp Lidocaine/ 43 41 and 42 44 31 and 43 45 43 not 44 46 from 45 keep ALL


INTERVENTIONAL & SURGICAL SEARCHES 2008 Search strategy of Ovid MEDLINE, 1950 to July Week 2 2008 1 randomized controlled trial.mp. or exp Randomized Controlled Trial/ 2 randomized controlled trial.pt. 3 controlled clinical trial.mp. or exp Controlled Clinical Trial/ 4 controlled clinical trial.pt. 5 clinical trial.mp. or exp Clinical Trial/ 6 clinical trial.pt. 7 or/1-6 8 limit 7 to humans 9 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ or facet joint/ or zygapophysial joint/ or sacroiliac.mp.[mp=title, original title, abstract, name of substance word, subject heading word] 10 (spine or coccyx or intervertebral disk or lumbar vertebrae or sacrum or spinal canal or back or facet joint or zygapophysial joint or sacroiliac).tw. 11 spinal diseases/ or intervertebral disk displacement/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 12 (spinal disease$ or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 13 sciatica/ or radiculopathy/ 14 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp. 15 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp. 16 or/9-15 17 8 and 16 18 exp injections/ 19 ((spine$ or spinal$ or nerv$) adj7 block$).mp. 20 (prolotherar$ or sclerotherap$).mp. 21 trigger point injection$.mp. 22 medial branch block$.mp. 23 or/18-22 24 botox.mp. or Botulinum Toxin Type A/ 25 (disc$ adj3 (replac$ or prosthe$)).mp. 26 exp Intervertebral Disk/ 27 exp "Prostheses and Implants"/ 28 25 or (26 and 27) 29 ((intradisc$ or intradisk$) adj5 (electrotherm$ or annuloplast$)).mp. 30 ((intradisc$ or intradisk$) adj5 (radiofrequenc$ or thermocoagulat$)).mp. 31 nucleoplast$.mp. 32 chemonucleolysis.mp. or Intervertebral Disk Chemolysis/ 33 spinewand$.mp. 34 dekompress$.mp. 35 limit 34 to english language 36 Injections, Epidural/ 37 (epidural adj2 (corticosteroid$ or steroid$)).mp. 38 36 or 37 39 exp spinal cord/ 40 exp electric stimulation therapy/ 41 (electric$ adj7 stimulat$).mp. 42 (39 and 40) or 41 43 23 or 24 or 28 or 29 or 30 or 31 or 32 or 33 or 35 or 38 or 42 44 17 and 43 45 from 44 keep ALL


INTERVENTIONAL & SURGICAL SEARCHES 2008 Search strategy of Cochrane Central Register of Controlled Trials, through 2nd Quarter 2008 1 randomized controlled trial.mp. or exp Randomized Controlled Trial/ 2 randomized controlled trial.pt. 3 controlled clinical trial.mp. or exp Controlled Clinical Trial/ 4 controlled clinical trial.pt. 5 clinical trial.mp. or exp Clinical Trial/ 6 clinical trial.pt. 7 or/1-6 8 limit 7 to humans[Limit not valid; records were retained] 9 spine/ or coccyx/ or intervertebral disk/ or lumbar vertebrae/ or sacrum/ or spinal canal/ or exp back/ or facet joint/ or zygapophysial joint/ or sacroiliac.mp.[mp=title, original title, abstract, mesh headings, heading words, keyword] 10 (spine or coccyx or intervertebral disk or lumbar vertebrae or sacrum or spinal canal or back or facet joint or zygapophysial joint or sacroiliac).tw. 11 spinal diseases/ or intervertebral disk displacement/ or spinal osteophytosis/ or hyperostosis, diffuse idiopathic skeletal/ or spinal stenosis/ or spondylitis/ or spondylolisthesis/ or spondylolysis/ 12 (spinal disease$ or hyperostosis or spinal stenosis or spondyliti$ or spondylolisthesis or spondylolysis).tw. 13 sciatica/ or radiculopathy/ 14 exp BACK INJURIES/ or ((back or lumbar or sacrum or sacral) adj2 (wound$ or injur$ or trauma$)).mp. 15 exp back pain/ or ((back or lumbar or sacrum or sacral) adj2 (pain$ or ache$ or aching)).mp. 16 or/9-15 17 8 and 16 18 exp injections/ 19 ((spine$ or spinal$ or nerv$) adj7 block$).mp. 20 (prolotherar$ or sclerotherap$).mp. 21 trigger point injection$.mp. 22 medial branch block$.mp. 23 or/18-22 24 botox.mp. or Botulinum Toxin Type A/ 25 (disc$ adj3 (replac$ or prosthe$)).mp. 26 exp Intervertebral Disk/ 27 exp "Prostheses and Implants"/ 28 25 or (26 and 27) 29 ((intradisc$ or intradisk$) adj5 (electrotherm$ or annuloplast$)).mp. 30 ((intradisc$ or intradisk$) adj5 (radiofrequenc$ or thermocoagulat$)).mp. 31 nucleoplast$.mp. 32 chemonucleolysis.mp. or Intervertebral Disk Chemolysis/ 33 spinewand$.mp. 34 dekompress$.mp. 35 limit 34 to english language[Limit not valid; records were retained] 36 Injections, Epidural/ 37 (epidural adj2 (corticosteroid$ or steroid$)).mp. 38 36 or 37 39 exp spinal cord/ 40 exp electric stimulation therapy/ 41 (electric$ adj7 stimulat$).mp. 42 (39 and 40) or 41 43 23 or 24 or 28 or 29 or 30 or 31 or 32 or 33 or 35 or 38 or 42 44 17 and 43 45 from 44 keep ALL

APPENDIX 3. SYSTEMATIC REVIEWS QUALITY RATING SYSTEM

Criteria for Assessing Scientific Quality of Research Reviews* Criteria 1. Were the search methods reported? Were the search methods used to find evidence (original research) on the primary questions stated? Yes if the review states the databases used, date of most recent searches, and some mention of search terms. 2. Was the search comprehensive? Was the search for evidence reasonably comprehensive? Yes if the review searches at least 2 databases and looks at other sources (such as reference lists, hand searches, queries experts). 3. Were the inclusion criteria reported? Were the criteria used for deciding which studies to include in the overview reported? 4. Was selection bias avoided? Was bias in the selection of studies avoided? Yes if the review reports how many studies were identified by searches, numbers excluded, and gives appropriate reasons for excluding them (usually because of pre-defined inclusion/exclusion criteria). 5. Were the validity criteria reported? Were the criteria used for assessing the validity of the included studies reported? 6. Was validity assessed appropriately? Was the validity of all the studies referred to in the text assessed using appropriate criteria (either in selecting studies for inclusion or in analyzing the studies that are cited)? Yes if the review reports validity assessment and did some type of analysis with it (e.g. sensitivity analysis of results according to quality ratings, excluded lowquality studies, etc.) 7. Were the methods used to combine studies reported? Were the methods used to combine the findings of the relevant studies (to reach a conclusion) reported? "Yes" for studies that did qualitative analysis if there is some mention that quantitative analysis was not possible and reasons that it could not be done, or if 'best evidence' or some other grading of evidence scheme used. The purpose of this index is to evaluate the scientific quality (i.e. adherence to scientific principles) of research overviews (review articles) published in the medical literature. It is not intended to measure literary quality, importance, relevance, originality, or other attributes of overviews. The index is for assessing overviews of primary (original) research on pragmatic questions regarding causation, diagnosis, prognosis, therapy, or prevention. A research overview is a survey of research. The same principles that apply to epidemiological surveys apply to overviews: a question must be clearly specified, a target population identified and accessed, appropriate information obtained from that population in an unbiased fashion, and conclusions derived, sometimes with the help of formal statistical analysis, as is done in meta-analyses. The fundamental difference between overviews and epidemiological studies is the unit of analysis, not the scientific issues that the questions in this index address. Since most published overviews do not include a methods section, it is difficult to answer some of the questions in the index. Base your answers, as much as possible, on information provided in the overview. If the methods that were used are reported incompletely relative to a specific question, score it as cant tell, unless there is information in the overview to suggest either the criterion was or was not met. Operationalization of Criteria

APPENDIX 3. SYSTEMATIC REVIEWS QUALITY RATING SYSTEM

Criteria for Assessing Scientific Quality of Research Reviews* Criteria 8. Were the findings combined appropriately? Were the findings of the relevant studies combined appropriately relative to the primary question the overview addresses? "Yes" if the review performs a test for heterogeneity before pooling, does appropriate subgroup testing, appropriate sensitivity analysis, or other such analysis. 9. Were the conclusions supported by the reported data? Were the conclusions made by the author(s) supported by the data and/or analysis reported in the overview? 10. What was the overall scientific quality of the overview? How would you rate the scientific quality of this overview? Operationalization of Criteria For Question 8, if no attempt has been made to combine findings, and no statement is made regarding the inappropriateness of combining findings, check No. If a summary (general) estimate is given anywhere in the abstract, the discussion, or the summary section of the paper, and it is not reported how that estimate was derived, mark No even if there is a statement regarding the limitations of combining the findings of the studies reviewed. If in doubt, mark Cant tell. For an overview to be scored as Yes in Question 9, data (not just citations) must be reported that support the main conclusions regarding the primary question(s) that the overview addresses. The score for Question 10, the overall scientific quality, should be based on your answers to the first nine questions. The following guidelines can be used to assist with deriving a summary score: If the Cant tell option is used one or more times on the preceding questions, a review is likely to have minor flaws at best and it is difficult to rule out major flaws (i.e. a score of 4 or lower). If the No option is used on Question 2, 4, 6 or 8, the review is likely to have major flaws (i.e. a score of 3 or less, depending on the number and degree of the flaws) Scoring: 1 Extensive Flaws 2 Each Question is scored as Yes, Partially/Cant tell or No 5 Minor Flaws 6
Major Flaws
3 4
Minimal Flaws 7
* Operationalization of Oxman criteria50, adapted from Furlan et al51

APPENDIX 4. RANDOMIZED CONTROLLED TRIALS QUALITY RATING SYSTEM

Criteria List for Methodological Quality Assessment* Criteria A. Was the method of randomization adequate? Operationalization of Criteria A random (unpredictable) assignment sequence. An example of adequate methods is a computer generated random number table and use of sealed opaque envelopes. Methods of allocation using DOB, date of admission, hospital numbers, or alternation should not be regarded as appropriate. Assignment generated by an independent person not responsible for determining the eligibility of the patients. This person has no information about the persons included in the trial and has no influence on the assignment sequence or on the decision about eligibility of the patient. In order to receive a yes, groups have to be similar in baseline regarding demographic factors, duration or severity of complaints, percentage of patients with neurologic symptoms, and value of main outcome measure(s). Score Yes/No/Dont Know
B. Was the treatment allocation concealed?
Yes/No/Dont Know
C. Were the groups similar at baseline regarding the most important prognostic factors? "Yes", if similar: Age & gender Description of type of pain Intensity, duration or severity of pain D. Was the patient blinded to the intervention? E. Was the care provider blinded to the intervention? F. Was the outcome assessor blinded to the intervention? G. Were cointerventions avoided or similar? H. Was the compliance acceptable in all groups?
Yes/No/Dont Know
The reviewer determines if enough information about the blinding is given in order to score a yes: Use the author's statement on blinding, unless there is a differing statement/reason not to (no need for explicit information on blinding). Cointerventions should either be avoided in the trial design or similar between the index and control groups. The reviewer determines if the compliance to the interventions is acceptable, based on the reported intensity, duration, number and frequency of sessions for both the index intervention and control intervention(s).
Yes/No/Dont Know Yes/No/Dont Know Yes/No/Dont Know Yes/No/Dont Know Yes/No/Dont Know

APPENDIX 4. RANDOMIZED CONTROLLED TRIALS QUALITY RATING SYSTEM

Criteria List for Methodological Quality Assessment* Criteria I. Was the drop-out rate described and acceptable? 15% drop out rate is acceptable. Operationalization of Criteria The number of participants who are included in the study but did not complete the observation period or were not included in the analysis must be described and reasons given. If the percentage of withdrawals and drop-outs does not exceed 15% and does not lead to substantial bias, a yes is scored. Timing of outcome assessment should be identical for all intervention groups and for all important outcome assessments. All randomized patients are reported/analyzed in the group they were allocated to by randomization for the most important moments of effect measurement (minus missing values) irrespective of noncompliance and cointerventions. Score Yes/No/Dont Know
J. Was the timing of the outcome assessment in all groups similar? K. Did the analysis include an intention-to-treat analysis? Yes if less than 5% of randomized patients excluded.
Yes/No/Dont Know
Yes/No/Dont Know
This list includes only the internal validity criteria (n=11) that refer to characteristics of the study that might be related to selection bias (criteria A and B), performance bias (criteria D, E, G, and H), attrition bias (criteria I and K and detection bias (criteria F and J). The internal validity criteria should be used to define methodologic quality in the meta-analysis.
* Adapted from methods developed by the Cochrane Back Review Group56

APPENDIX 5: QUALITY RATINGS OF INCLUDED SYSTEMATIC REVIEWS

Methods for combining studies
Author, year de Graaf, 2006269 Devill, 2000

270
Search methods
Comprehensive
Inclusion criteria
Bias avoided
Validity criteria
Validity assessed
Appropriately combined
Conclusions supported
Overall quality
KQ 1a PREDICTIVE FEATURES OF HISTORY AND PHYSICAL EXAM Yes Yes Yes Yes Yes Partial Yes (1 database: published before 1995) Partial (1 electronic database) Yes Yes Partial Yes Yes Can't tell Can't tell (no info why studies excluded) Can't tell Yes Yes No Yes Yes No Partial (no sensitivity analysis) Partial (no sensitivity analysis) Partial Yes Yes Yes Yes Yes (listed ranges) Yes (some analysis of low qual studies) Can't tell Yes Yes Yes 6/7 7/7 4/7
Jarvik, 2002268
van den Hoogen, 1995
271
Yes
Yes
Yes
Yes
Yes
5/7
Vroomen, 1999272 KQ 1b PROGNOSIS Borge, 2001296 Crook, 2002

284 285
Yes
Yes
Yes
Yes
Can't tell
5/7
Yes Partial Yes Yes

287
No Yes Yes Yes Yes Yes Yes Partial Yes Yes Yes Partial Partial
Partial Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Partial Yes
Yes Yes Can't tell Yes Can't tell Can't tell Cant tell Cant tell Cant tell Cant tell Can't tell Yes Yes
No Yes Yes Yes Yes Yes No Yes Yes Yes Yes No Yes
No Yes Partial Yes Yes Yes No Partial Partial Yes Yes No Yes
No No No Partial Yes Yes No No Yes Yes Yes No Yes
Can't tell Can't tell Can't tell Can't tell Yes Can't tell Cant tell Cant tell Yes Yes Yes Can't tell Yes
Can't tell Can't tell Can't tell Can't tell Yes Partial Cant tell Cant tell Partial Yes Yes Can't tell Yes
2/7 4/7 3/7 4/7 6/7 4/7 3/7 3/7 5/7 5/7 5/7 2/7 6/7
Dionne, 2001 Fayad, 2004
286
Hartvigsen, 2004 Kuijer, 2006

288 289
Yes Yes Yes
Linton, 2000
McIntosh, 2000 Pengel, 2003

11
290
Yes Yes Yes Yes Yes
Pincus, 2002291 Pincus, 2006 Shaw, 2001

292 293 294
Steenstra, 2005
Yes


Methods for combining studies No
Author, year Truchon, 2000295 de Graaf, 2006269 Hoffman, 1991 Jarvik, 2002268 Pullman, 2000321 van den Hoogen, 1995271
320
Search methods Yes
Comprehensive Yes
Inclusion criteria Yes
Bias avoided Yes
Validity criteria No
Validity assessed No
Appropriately combined Cant' tell
Conclusions supported Can't tell
Overall quality 3/7
KQ 2a and 2b DIAGNOSTIC TESTING Yes Partial Yes Yes Yes Yes Yes Partial Partial Yes (1 database: published before 1995) Yes Yes Yes Yes Yes Partial Partial Yes Yes Yes Can't tell Can't tell Can't tell (no info why studies excluded) Cant tell Yes Yes Yes Yes No No Yes Yes Yes No No Partial (no sensitivity analysis) Partial Yes Yes No Yes Yes Yes Yes (listed ranges) Can't tell Yes (some analysis of low qual studies) Yes Yes Yes Yes Yes Yes Can't tell Yes 6/7 6/7 4/7 2/7 5/7
KQ 3 ADVICE TO REMAIN ACTIVE Hagen, 2002359, 360 Hagen, 2005

64, 65
Partial Yes Yes
Yes Yes Yes
Yes Yes Yes
Yes Yes Yes
Yes Yes Yes
Yes Yes Yes
6/7 7/7 7/7
KQ 3 - ADVICE TO REST IN BED Hagen, 200564, 65 KQ 3 LUMBAR SUPPORTS Jellema, 2001384, 385 French, 2006398 KQ 4 ACETAMINOPHEN Schnitzer, 2004411 van Tulder, 2000412, 413 Cherkin, 2003555 Yes Yes Partial (databases only) Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes No Yes No Yes 4/7 7/7 Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
KQ 3 SUPERFICIAL HEAT OR COLD Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
KQ 4 ACUPUNCTURE AND DRY NEEDLING Yes Yes Yes Yes No No Yes No Yes 4/7


Methods for combining studies No Yes Yes
Author, year Ernst, 2001556 Furlan, 2005

69, 70 68
Search methods Yes Yes Yes
Comprehensive Yes Yes Yes
Inclusion criteria Yes Yes Yes
Bias avoided Can't tell Yes Yes
Validity criteria No Yes Yes
Validity assessed No Yes Yes
Appropriately combined Can't tell Yes Yes Partial (combined tricyclic and non-tricyclic antidepressants trials; sensitivity analysis showed no differences)
Conclusions supported Can't tell Yes Yes
Overall quality 3/7 7/7 7/7
Manheimer, 2005
KQ 4 ANTIDEPRESSANTS
Salerno, 2002479
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
6/7
Schnitzer, 2004411
Yes
Partial (databases only)
Yes
Yes
Yes
Yes
No
Yes
Partial (used rates of improvement from baseline as one criteria for evaluating effectiveness) Yes
5/7
Staiger, 2003480 KQ 4 BACK SCHOOLS Elders, 2000588 Heymans, 2005586, 587 Maier-Riehle, 2001589 van der Hulst, 2005590 KQ 4 BENZODIAZEPINES van Tulder, 2003488, 489
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
7/7
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Can't tell (excluded studies) Yes
No Yes No Yes
Yes Yes Partial Partial
Partial Yes Yes Yes
Can't tell Yes Yes Yes
Can't tell Yes Yes Yes
3/7 7/7 4/7 6/7
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
7/7


Author, year KQ 4 EXERCISE Clare, 2004616 Hayden, 2005615 Hayden, 2005613, 614 Kool, 2004
617
Search methods
Comprehensive
Inclusion criteria
Bias avoided
Validity criteria
Validity assessed
Overall quality
Yes Yes Yes Yes Yes

619
Yes Yes Yes Yes Yes Yes Yes
Yes Yes Yes Can't tell (excluded 14 mid-to-high quality studies) Can't tell Yes Yes
Yes Partial (no sensitivity analysis) Yes Partial Partial Yes Yes
Yes Yes Yes No No Yes Yes
Yes Yes (no heterogeneity) Yes Can't tell Yes Yes Yes
Yes Yes Yes Can't tell Yes Yes Yes
6/7 7/7 7/7 7/7 3/7 7/7 4/7
Liddle, 2004618 Machado, 2006 McNeely, 2003

620
Yes Yes
KQ 4 FUNCTIONAL RESTORATION (PHYSICAL CONDITIONING, WORK CONDITIONING, OR WORK HARDENING) Schonstein, 2003302, 303 KQ 4 HERBAL THERAPIES Gagnier, 2007545, 546 Guzman, 2001643, 644 Karjalainen, 2001 Tveito, 2004
645 299, 300
Yes
Partial
Yes
Can't tell
Yes
Yes
Yes
Yes
Yes
6/7
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
7/7
KQ 4 INTERDISCIPLINARY REHABILITATION (MULTIDISCIPLINARY REHABILITATION) Yes Yes Partial Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Partial Yes Partial Yes Yes Yes Yes Yes Yes Yes Yes Yes 6/7 7/7 5/7
KQ 4 MASSAGE Cherkin, 2003555 Furlan, 2002

700, 701
Yes Yes
Yes Yes
Yes Yes
Yes Partial
No Yes
No Yes
Yes Yes
No Yes
Yes Cant tell
5/7 6/7
KQ 4 NEUROREFLEXOTHERAPY Urrutia, 2004582 Yes Yes Yes Cant tell Yes Yes Yes Yes Yes 6/7


Author, year
Search methods
Comprehensive Partial (databases only) Yes
Inclusion criteria
Bias avoided
Validity criteria
Validity assessed
Overall quality
KQ 4 NON-SELECTIVE NON-STEROIDAL ANTI-INFLAMMATORY DRUGS (NSAIDS) Schnitzer, 2004411 Van Tulder, 2000412, 413 Yes Yes Yes Yes Yes Yes Cant tell (insufficient detail of excluded studies) Yes Yes Yes Yes No Yes Yes Yes Partial Yes 5/7 7/7
Vroomen, 2000100
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
5/7
KQ 4 PSYCHOLOGICAL THERAPIES Hoffman, 2007722 Ostelo, 2005301 Yes Yes Yes Yes Yes Yes Partial Yes Yes Yes Yes Partial (no sensitivity analysis) Yes Yes Yes Yes Yes Yes 6/7 6/7
KQ 4 SKELETAL MUSCLE RELAXANTS Browning, 2001506 Schnitzer, 2004411 van Tulder, 2003488, 489 Vroomen, 2000 Pittler, 2006745 KQ 4 SPINAL MANIPULATION Assendelft, 200366, 67 Avery, 2004
753 754 100
Yes Yes Yes Yes
Yes Partial (data bases only) Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Cant tell
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes No Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
7/7 5/7 7/7 5/7
KQ 4 SPA THERAPY AND BALNEOTHERAPY Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Cant Tell
Yes No Yes Yes
Yes Partial Yes Yes
Yes No Yes Yes
Yes Partial Partial Yes
Yes Partial Partial Yes
7/7 2/7 4/7 6/7
Bronfort, 2004 Brown, 2005
750


Methods for combining studies Yes Yes Yes Yes Yes No Yes Yes No Yes No Yes Yes
Author, year Cherkin, 2003555 Ernst, 2001 Ernst, 2003

759 755 752 751
Search methods Yes Yes Yes Yes Yes Yes Yes Yes Partial Yes Yes Yes Yes
Comprehensive Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Partial (data bases only)
Inclusion criteria Yes Yes Yes Yes Yes No Yes Yes No Yes Yes Yes Yes
Bias avoided Yes Yes Yes Yes Yes Cant tell Yes Yes No Can't tell Can't tell Can't Tell Can't Tell
Validity criteria No No Yes Yes Yes No Yes No No Yes No Yes Yes
Validity assessed No No Yes Yes Yes No Yes No No Yes No Yes Yes
Appropriately combined Yes Cant tell Partial Partial Partial Cant tell Partial Partial Partial No Can't tell Yes Partial
Conclusions supported Yes Cant tell Yes Yes Yes Cant tell Yes Can't tell Partial No Can't tell Yes Partial
Overall quality 5/7 3/7 5/7 7/7 5/7 2/7 5/7 4/7 1/7 3/7 2/7 5/7 4/7
Ferreira, 2002 Ferreira, 2003 Gay, 2005

756 763
Kent, 2005
Licciardone, 2005757 Meeker, 2002760 Oliphant, 2004

761 762
Stevinson, 2002 Vroomen, 2000
100 758
Woodhead, 2005
KQ 4 TRAMADOL Schnitzer, 2004411 Yes Yes Yes Yes Yes No Yes Yes 5/7
KQ 4 TRANSCUTANEOUS ELECTRICAL NERVE STIMULATION (TENS) (see also Assendelft, 2003 (spinal manipulation)66, 67; Clarke, 2005 (traction)676, 677; French, 2006 (superficial heat or cold)398; Furlan, 2002 (massage)700, 701; Manheimer, 2005 (acupuncture68) Khadilkar, 2005
698, 699
Yes
Yes
Yes
Yes
Yes
Yes (size prohibited analysis) Yes Yes Partial Yes
Yes
Yes
Yes
7/7
KQ 4 TRACTION Clarke, 2006676, 677 Harte, 2003678 Philadelphia Panel, 2001 Vroomen, 2000
100 399
Yes Yes Partial Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Can't tell Yes Yes Can't tell
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Partial Yes
Yes Yes Yes Yes
6/7 7/7 5/7 5/7


Author, year
Search methods
Comprehensive No (no ML search) Yes
Inclusion criteria
Bias avoided
Validity criteria
Validity assessed
Overall quality
KQ 5 DECISION TOOLS FOR TARGETING TREATMENT Hestboek, 2000795 Najm, 2003791 Yes Yes Yes Yes Yes Yes Yes Yes Partial Partial (no sensitivity analysis) Partial (no sensitivity analysis) Yes Partial No (not reported) Partial Partial Partial No No No No Can't tell Partial Can't tell Yes 4/7 5/7
Seffinger, 2004792 van der Wurff, 2000793, 794 Boswell, 2003836 Hansen, 200783 Sehgal, 2007837 Datta, 2007834 Buenaventura, 2007809 Cohen, 2005806 Willems, 2004
810
Yes Yes Yes Yes Yes Yes Yes Partial Yes
Yes Yes Yes Yes Yes Yes Yes Partial Yes
Yes Yes Yes Yes Yes Yes Yes No Yes
Yes Can't tell Can't tell Can't tell Yes Yes Can't tell Can't tell Can't tell
Yes Yes Yes Yes Yes Yes Yes No No
Yes No No Partial No No Yes No No
Yes Yes No Can't tell No No No No No
Yes Yes No Can't tell No No No No Can't tell
6/7 5/7 2/7 2/7 2/7 2/7 3/7 1/7 2/7
KQ 7 DIAGNOSTIC INTRA-ARTICULAR FACET JOINT BLOCK, SACROILIAC JOINT BLOCK, OR MEDIAL BRANCH BLOCK
KQ 7 DIAGNOSTIC SELECTIVE NERVE ROOT BLOCKS KQ 7 PROVOCATIVE DISCOGRAPHY
KQ 8 CHEMONUCLEOLYSIS Gibson, 200781, 82 Abdi, 200771 Armon, 2007

74 77
Yes
Yes
Yes
Yes
Yes
Partial
Yes
Yes
Yes
6/7
KQ 8 EPIDURAL STEROID INJECTIONS Yes Yes Yes Yes Partial Yes Yes Yes Cant tell Yes Yes Yes Partial Partial Partial Yes Partial Yes No Partial No No Partial Cant tell Yes Partial Yes 3/7 4/7 4/7
DePalma, 2005


Methods for combining studies Yes Yes Partial Yes Can't tell Yes
Author, year Luijsterburg, 200784 Nelemans, 2001 Resnick, 2005 Staal, 2008
94 97 92 86
Search methods Yes Yes Yes Yes No Yes
Comprehensive Yes Yes Partial Yes No Yes
Inclusion criteria Yes Yes Partial Yes No Can't tell
Bias avoided Yes Yes Can't tell Yes No Yes
Validity criteria Yes Yes No Yes No Yes
Validity assessed Yes Yes No Yes Partial Yes
Appropriately combined Yes Yes Can't tell Yes Can't tell Yes
Conclusions supported Yes Yes Can't tell Yes No Yes
Overall quality 7/7 7/7 2/7 7/7 1/7 5/7
Tonkovich-Quaranta, 2000 Vroomen, 2000 Boswell, 200775 Resnick, 200592 Slipman, 2003 Staal, 2008
94 93 100
KQ 8 FACET JOINT INJECTION AND MEDIAL BRANCH BLOCK Yes Yes Partial Yes Yes Partial Partial Yes Yes Partial Yes Yes Can't tell Can't tell Can't tell Yes Yes No No Yes Partial No Partial Yes Yes Partial Partial Yes No Can't tell Can't tell Yes No Can't tell Can't tell Yes 3/7 2/7 3/7 7/7
KQ 8 INTRADISCAL ELECTROTHERMAL THERAPY (IDET) Andersson, 200672 Appleby, 2006 Gibson, 2005 NICE, 2004
87 99 73 79, 80
Yes Partial Yes Yes Yes
Partial No Yes Yes Partial
Yes Partial Yes Yes Yes
Can't tell Can't tell Yes Yes Yes
Partial No Yes No Yes
Partial No Partial Yes Yes
Yes Yes Yes No Partial
No No Yes Can't tell Yes
No No Yes Can't tell Yes
2/7 1/7 6/7 4/7 6/7
Urrutia, 2007
KQ 8 INTRADISCALSTEROID INJECTION Gibson, 200781, 82 KQ 8 LOCAL INJECTIONS Abdi, 2005170 Resnick, 2005 Staal, 2008
94 92
Yes
Yes
Yes
Yes
Yes
Partial
Yes
Yes
Yes
6/7
Yes Yes Yes
Yes Partial Yes
Yes Partial Yes
Yes Can't tell Yes
Partial No Yes
Yes No Yes
No Partial Yes
No Can't tell Yes
Yes Can't tell Yes
3/7 2/7 7/7


Methods for combining studies Yes No No Yes Partial
Author, year Gibson, 200579, 80 NICE, 200487 NICE, 200488 Niemisto, 2003 Urrutia, 2007
99 90, 91
Search methods Yes Yes Yes Yes Yes
Comprehensive Yes Yes Yes Yes Partial
Inclusion criteria Yes Yes Yes Yes Yes
Bias avoided Yes Yes Can't tell Yes Yes
Validity criteria Yes No No Yes Yes
Validity assessed Partial Yes Partial Yes Yes
Appropriately combined Yes Can't tell Yes Yes Yes
Conclusions supported Yes Can't tell Yes Yes Yes
Overall quality 6/7 4/7 4/7 7/7 6/7
KQ 8 PERCUTANEOUS INTRADISCAL RADIOFREQUENCY THERMOCOAGULATION (PIRFT) AND COBLATION NUCLEOPLASTY
KQ 8 PROLOTHERAPY Dagenais, 200776 Boswell, 200775 Geurts, 2001

78
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
7/7
KQ 8 RADIOFREQUENCY DENERVATION Yes Yes Yes Yes Partial Yes Yes Yes Partial Partial Yes Yes Yes Partial Yes Can't tell Yes Yes Can't tell Can't tell Yes Yes Yes No No Partial Yes Yes No Partial Yes Yes Yes Partial Partial No Yes Yes Can't tell Can't tell No Yes Yes Can't tell Can't tell 3/7 7/7 7/7 2/7 3/7
Niemisto, 200390, 91 Resnick, 2005

92 93
Slipman, 2003
KQ 8 SACROILIAC JOINT STEROID INJECTION Hansen, 200783 Maillis-Gagnon, 200485 Taylor, 2006
95, 96 98
Yes
Yes
Yes
Can't tell
Yes
Partial
Yes
Yes
Yes
5/7
KQ 8 (and KQ 11) SPINAL CORD STIMULATION Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Can't tell Can't tell Yes Yes Yes Yes Yes Partial Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7 5/7 5/7
Turner, 2004
KQ 9 SURGERY FOR ISTHMIC SPONDYLOLISTHESIS Gibson, 200579, 80 Kwon, 2005

216
Yes Partial
Yes No
Yes Yes
Yes Can't tell
Yes No
Partial No
Yes Partial
Yes Can't tell
Yes Can't tell
6/7 1/7


Methods for combining studies Yes Yes Partial
Author, year Andersson, 200672 Bono, 2004

210 212
Search methods Yes Yes Yes
Comprehensive Partial Partial Yes
Inclusion criteria Yes Yes Yes
Bias avoided Can't tell Can't tell Yes
Validity criteria Partial No Yes
Validity assessed Partial No Yes
Appropriately combined No Can't tell Yes Partial (combined RCTs and observational studies) Can't tell Can't tell Yes Partial Yes Partial Can't tell Can't tell Can't tell
Conclusions supported No Can't tell Partial Partial (incomplete analyses of potential confounders) Yes Can't tell Yes Partial Yes Yes Can't tell Can't tell Can't tell
Overall quality 2/7 3/7 6/7
KQ 9 SURGERY FOR NON-RADICULAR LOW BACK PAIN WITH COMMON DEGENERATIVE CHANGES
de Kleuver, 2003
Fenton, 2007230
Yes
Cant tell
Yes
Partial
No
No
Yes
3/7
Freeman, 2006213 Geisler, 2004 Gibson, 2005

214 79, 80 215
Yes Yes Yes Yes Yes Yes Yes Yes Yes
Partial No Yes Yes Partial Yes Yes Partial Partial
No Yes Yes Yes Yes Yes Partial Partial Yes
Partial Can't tell Yes Yes Can't tell Can't tell Can't tell Can't tell Can't tell
No No Yes Yes Yes No No No No
No No Partial Yes Partial Partial Partial Partial Partial
Partial No Yes Yes No No Can't tell Can't tell Can't tell
4/7 2/7 6/7 5/7 5/7 4/7 2/7 2/7 3/7
Ibrahim, 2008 Mirza, 2007
218
NICE, 2004221 Resnick, 2005224 Resnick, 2005 Resnick, 2005 Boult, 2000211 Gibson, 2007 NICE, 2005 NICE, 2003 NICE, 2003
81, 82 222 219 220 227 225 226
KQ 9 SURGERY FOR RADICULOPATHY WITH HERNIATED LUMBAR DISC Yes Yes Yes Yes Yes Yes Partial Yes Yes Yes Yes Partial Yes Yes Yes Yes Yes Partial Can't tell Yes Can't tell Cant tell Can't tell Can't tell No Yes No No No No No Partial Partial Partial Partial Partial Yes Yes No No No No Yes Yes Yes NA NA Can't tell Yes Yes Yes Yes Yes Can't tell 4/7 6/7 4/7 4/7 4/7 2/7
Resnick, 2005


Author, year Aalto, 2006254 Bono, 2004

210 79, 80
Search methods
Comprehensive
Inclusion criteria
Bias avoided
Validity criteria
Validity assessed
Overall quality
KQ 9 SURGERY FOR SPINAL STENOSIS WITH OR WITHOUT DEGENERATIVE SPONDYLOLISTHESIS Yes Yes Yes Yes Yes
228 229 225
Yes Partial Yes Yes Yes Partial Partial Partial Partial
Yes Yes Yes Yes Yes Partial Partial Partial Partial
Partial Can't tell Yes Partial Can't tell Can't tell Can't tell Can't tell Can't tell
Yes No Yes Yes No No No No No
Yes No Partial Partial Partial Partial Partial Partial Partial
No Yes Yes Yes No No Can't tell Can't tell Can't tell (inconsistency between text and tables)
Can't tell Cant tell Yes Yes Yes Can't tell Can't tell Can't tell Can't tell
Can't tell Cant tell Yes Yes Yes Can't tell Can't tell Can't tell Can't tell
4/7 3/7 6/7 5/7 4/7 2/7 2/7 2/7 2/7
Gibson, 2005 Martin, 2007 NICE, 2005
217
223
Resnick, 2005 Resnick, 2005 Resnick, 2005
Yes Yes Yes Yes
Resnick, 2005226
KQ 10 ACUPUNCTURE COMBINED WITH OTHER NON-INVASIVE INTERVENTIONS Furlan, 200569, 70. van Tulder, 2003488, 489 Hayden 2005613, 614 Furland, 2002700, 701 Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
KQ 10 COMBINATIONS OF MEDICATIONS Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
KQ 10 EXERCISE COMBINED WITH OTHER INTERVENTIONS Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
KQ 10 MASSAGE COMBINED WITH OTHER INTERVENTIONS Yes Yes Yes Partial Yes Yes Yes Yes Can't tell 6/7
KQ 10 PSYCHOLOGICAL THERAPIES COMBINED WITH OTHER INTERVENTIONS Ostelo, 2005301 Yes Yes Yes Yes Yes Partial (no sensitivity analysis) Yes Yes Yes 6/7


Author, year Assendelft, 200466, 67 Clarke, 2006676, 677
Search methods
Comprehensive
Inclusion criteria
Bias avoided
Validity criteria
Validity assessed
Overall quality
KQ 10 SPINAL MANIPULATION COMBINED WITH OTHER INTERVENTIONS Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
KQ 10 TRACTION COMBINED WITH OTHER INTERVENTIONS Yes Yes Yes Can't tell Yes Yes Yes Yes No (incorrect classification of trial as randomized) Yes Yes Yes Yes 6/7
KQ 11 ADHESIOLYSIS AND FORCEFUL EPIDURAL INJECTIONS
Trescot, 2007
982
Yes
Yes
Yes
Yes
Yes
Partial
Yes
Partial
3/7
KQ 11 SPINAL CORD STIMULATION Mailis-Gagnon, 200485 Taylor, 2006

95, 96 98
Yes Yes Yes
Yes Yes Yes
Yes Yes Yes
Yes Can't tell Can't tell
Yes Yes Yes
Yes Yes Partial
Yes Yes Yes
Yes Yes Yes
5/7 5/7 5/7
Turner, 2004
KQ 13 ADVICE TO STAY ACTIVE Hilde, 2002360 KQ 13 BACK SCHOOLS Heymans, 2004586, 587 Linton, 2001
1004
Partial
Yes
Yes
Don't know
Yes
Yes
Yes
Yes
Yes
6/7
Yes Partial
Yes Yes
Yes Yes
Yes Can't tell
Yes No
Yes No
Yes No
Yes Don't know
Yes Yes
7/7 3/7
KQ 13 EXERCISE Hayden, 2005613, 614 Linton, 2001

1004
Yes Partial
Yes Yes
Yes Yes
Yes Can't tell
Yes No
Yes No
Yes No
Yes Don't know
Yes Yes
7/7 3/7
KQ 13 LUMBAR SUPPORTS Jellema, 2001385 Furlan, 200569, 70 Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7
KQ 14 ACUPUNCTURE DURING PREGNANCY Yes Yes Yes Yes Yes Yes Yes Yes Yes 7/7


Methods for combining studies No
Author, year Stuge, 20031012 Stuge, 20031012 Stuge, 20031012 Young, 20051024
Search methods Yes
Comprehensive Yes
Inclusion criteria Yes
Bias avoided Yes
Validity criteria Yes
Validity assessed Yes
Appropriately combined Yes
Conclusions supported Yes
Overall quality 6/7
KQ 14 MASSAGE DURING PREGNANCY Yes Yes Yes Yes Yes Yes No Yes Yes 6/7
KQ 14 PHYSICAL THERAPY DURING PREGNANCY Yes Yes Yes Yes Yes Yes No Yes Yes 6/7
KQ 14 SUPPORTIVE DEVICES DURING PREGNANCY Yes Partial Yes Can't tell No Partial No Yes Yes 4/7

APPENDIX 6: LIST OF EXCLUDED SYSTEMATIC REVIEWS
Author, year, title KQ 1a DIAGNOSIS Deyo, 1992273 What can the history and physical examination tell us about low back pain? Jarvik, 2003275 Imaging of adults with low back pain in the primary care setting
Reason for exclusion
Outdated Not clear if systematic methods used Reports same results as another included 268 systematic review (Jarvik, 2002 )
Rebain, 2002274 Does not evaluate diagnostic accuracy of A systematic review of the passive straight leg raising test as a diagnostic aid straight leg raise test for low back pain (1989 to 2000) KQ 2a and 2b DIAGNOSTIC TESTING Boos, 1996316 Clinical efficacy of imaging modalities in the diagnosis of low-back pain disorders Deyo, 1992273 What can the history and physical examination tell us about low back pain? Outdated
Outdated
Geisser, 2005322 Only evaluates ability of surface A meta-analytic review of surface electromyography among persons with low electromyography to distinguish persons with back pain and normal, healthy controls low back pain from persons without low back pain Jarvik, 2003275 Imaging of adults with low back pain in the primary care setting Kardaun, 1989324 CT, myelography, and phlebography in the detection of lumbar disk herniation: an analysis of the literature Reports same results as another included 268 systematic review (Jarvik, 2002 ) Outdated
Kent, 1992325 Outdated Diagnosis of lumbar spinal stenosis in adults: a metaanalysis of the accuracy of CT, MR, and myelography Littenberg, 1995317 Clinical efficacy of SPECT bone imaging for low back pain Outdated
Mohseni-Bandpei, 2000323 Only evaluates the ability of surface Application of surface electromyography in the assessment of low back pain: electromyography to distinguish persons with low back pain from persons without low back a literature review pain KQ 2a and 2b IMAGING Systematic reviews of tests to diagnose serious underlying conditions Boos, 1996316 Clinical efficacy of imaging modalities in the diagnosis of low-back pain disorders Deyo, 1992273 What can the history and physical examination tell us about low back pain? Jarvik, 2003275 Imaging of adults with low back pain in the primary care setting Littenberg, 1995317 Clinical efficacy of SPECT bone imaging for low back pain Does not clearly use systematic methods to synthesize the literature Outdated Not clear if systematic methods used Reports same results as another included 268 systematic review (Jarvik, 2002 ) Outdated

Author, year, title KQ 3 ADVICE TO REST IN BED Allen, 1999342 Bed rest: a potentially harmful treatment needing more careful evaluation Hagen, 2000341 The Cochrane review of bed rest for acute low back pain and sciatica Koes, 1994343 Efficacy of bed rest and orthoses for low back pain Maher, 1999344 Prescription of activity for low back pain: what works? Scheer, 1995345 Randomized controlled trials in industrial low back pain relating to return to work. Part 1. Acute interventions van der Weide, 1997346 Vocational outcome of intervention for low-back pain van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions Waddell, 1997347 Systematic reviews of bed rest and advice to stay active for acute low back pain KQ 3 LUMBAR SUPPORTS Koes, 1994343 Efficacy of bed rest and orthoses for low back pain Scheer, 1997386 Randomized controlled trials in industrial low back pain. Part 3. Subacute/chronic pain interventions van Poppel, 2000387 Mechanisms of action of lumbar supports KQ 3 SUPERFICIAL HEAT-COLD Philadelphia Panel, 2001399 Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for low back pain KQ 4 ACETAMINOPHEN Deyo, 1996415 Drug therapy for back pain. Which drugs help which patients? van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 ACUPUNCTURE Ernst, 1997558 Life-threatening adverse reactions after acupuncture? A systematic review Ernst, 1998559 Acupuncture for back pain. A meta-analysis of randomized controlled trials. Outdated Outdated Outdated Outdated Outdated
Outdated Not specific for low back pain Updated Cochrane review available (Hagen, 64, 65 ) 2004 Outdated Outdated Outdated
Outdated Outdated
Outdated
Does not evaluate clinical outcomes from use of lumbar supports
Outdated Systematic methods not reported Outdated

Author, year, title Ezzo, 2000 Is acupuncture effective for the treatment of chronic pain? A systematic review Patel, 1989562 A meta-analysis of acupuncture for chronic pain Smith, 2000564 Teasing apart quality and validity in systematic reviews: an example from acupuncture trials in chronic neck and back pain Strauss, 1999560 Acupuncture and the treatment of chronic low-back pain: a review of the literature ter Riet, 1990561 Acupuncture and chronic pain: a criteria-based meta-analysis van Tulder, 1999557, 1032 The effectiveness of acupuncture in the management of acute and chronic low back pain. A systematic review within the framework of the Cochrane Collaboration Back Review Group KQ 4 ANTIDEPRESSANTS Fishbain, 2000485 Evidence-based data on pain relief with antidepressants Goodkin, 1989482 Antidepressants for the relief of chronic pain: do they work? Onghena, 1992483 Antidepressant-induced analgesia in chronic non-malignant pain: a metaanalysis of 39 placebo-controlled studies Turner, 1993484 Do antidepressant medications relieve chronic low back pain? van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 BACK SCHOOLS Cohen, 1994592 Group education interventions for people with low back pain Di Fabio, 1995593 Efficacy of comprehensive rehabilitation programs and back school for patients with low back pain: a meta-analysis Keijsers, 1991594 Validity and comparability of studies on the effects of back schools Koes, 1994343 The efficacy of back schools: a review of randomized clinical trials Nentwig, 1999596 Effectiveness of the back school. A review of the results of evidence-based evaluation Scheer, 1995345 Randomized controlled trials in industrial low back pain relating to return to work. Part 1. Acute interventions Outdated Outdated
563
Reason for exclusion Outdated
Outdated Not specific for low back pain Outdated
Outdated
Outdated Not specific for low back pain Updated Cochrane review available (Furlan, 200569, 70)
Not specific for low back pain Outdated Not specific for low back pain Outdated Not specific for low back pain Outdated Outdated
Outdated Outdated Outdated German language Outdated

Author, year, title Scheer, 1997 Randomized controlled trials in industrial low back pain. Part 3. Subacute/chronic pain interventions Turner, 1996597 Educational and behavioral interventions for back pain in primary care van der Weide, 1997346 Vocational outcome of intervention for low-back pain van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions van Tulder, 1999591 Back schools for non-specific low-back pain KQ 4 BENZODIAZEPINES Deyo, 1996415 Drug therapy for back pain. Which drugs help which patients? van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 EXERCISE Beckerman, 1993623 Efficacy of physiotherapy for musculoskeletal disorders: what can we learn from research? Cleland, 2002627 The role of therapeutic exercise in treating instability-related lumbar spine pain: a systematic review Colle, 2002622 Impact of quality scales on levels of evidence inferred from a systematic review of exercise therapy and low back pain Faas, 1996624 Exercises: which ones are worth trying, for which patients, and when? Hilde, 1998625 Effect of exercise in the treatment of chronic low back pain: a systematic review, emphasizing type and dose of exercise Koes, 1991626 Physiotherapy exercises and back pain: a blinded review Maher, 1999344 Prescription of activity for low back pain: what works? Ostelo, 2003790 Rehabilitation following first-time lumbar disc surgery. A systematic review within the framework of the Cochrane Collaboration Scheer, 1995345 Randomized controlled trials in industrial low back pain relating to return to work. Part 1. Acute interventions
386
Outdated Outdated Outdated
Updated Cochrane review available 586, 587 (Heymans, 2004 Outdated Systematic methods not reported Outdated
Outdated Not specific for low back pain Systematic methods not used for synthesizing results Instability-related lumbar spine pain not a clearly recognized entity Only included trials identified by an outdated 621 Cochrane review (van Tulder, 2000 ) Outdated Outdated
Outdated Outdated Post-surgery patients
Outdated

Author, year, title Scheer, 1997 Randomized controlled trials in industrial low back pain. Part 3. Subacute/chronic pain interventions van der Weide, 1997346 Vocational outcome of intervention for low-back pain van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions
386
Outdated Outdated
van Tulder, 2000621 Updated Cochrane review available (Hayden, 613, 614 ) Exercise therapy for low back pain. A systematic review within the framework 2005 of the Cochrane Collaboration Back Review Group KQ 4 HERBAL THERAPIES Gagnier, 2004547 Outdated Harpgophytum procumbens for osteoarthritis and low back pain: A systematic review KQ 4 INTERDISCIPLINARY REHABILITATION (MULTIDISCIPLINARY REHABILITATION) Di Fabio, 1995593 Efficacy of comprehensive rehabilitation programs and back school for patients with low back pain: a meta-analysis KQ 4 LOW LEVEL LASER Beckerman, 1992662 The efficacy of laser therapy for musculoskeletal and skin disorders: a criteria-based meta-analysis of randomized clinical trials Beckerman, 1993623 Efficacy of physiotherapy for musculoskeletal disorders: what can we learn from research? Outdated Not specific for low back pain Not specific for low back pain Outdated
Bjordal, 2003663 Not specific for low back pain A systematic review of low level laser therapy with location-specific doses for pain from chronic joint disorders de Bie, 1998 Efficacy of 904 nm laser therapy in the management of musculoskeletal disorders: a systematic review
664
Not specific for low back pain
Gam, 1993665 Outdated The effect of low-level laser therapy on musculoskeletal pain: a meta-analysis Not specific for low back pain KQ 4 MASSAGE Ernst, 1999 Massage therapy for low back pain: a systematic review Ernst, 2003735 The safety of massage therapy
734
Outdated Not specific for low back pain Includes 2 case reports of serious adverse events following massage in patients with low back pain Outdated
Philadelphia Panel, 2001399 Philadelphia Panel evidence-based clinical practice guidelines on selected rehabilitation interventions for low back pain

Author, year, title KQ 4 MODIFIED WORK Hlobil, 2005741 Effectiveness of a return-to-work intervention for subacute low-back pain Krause, 1998742 Modified work and return to work: a review of the literature Tveito, 2004645 Low back pain interventions at the workplace: a systematic literature review KQ 4 MUSCLE RELAXANTS Deyo, 1996415 Drug therapy for back pain. Which drugs help which patients? van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 NONSTEROIDAL ANTI-INFLAMMATORY DRUGS Deyo, 1996415 Drug therapy for back pain. Which drugs help which patients? Koes, 1997448 Efficacy of non-steroidal anti-inflammatory drugs for low back pain: a systematic review of randomized clinical trials van der Weide, 1997346 Vocational outcome of intervention for low-back pain van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 OPIOIDS Bartleson, 2002508 Evidence for and against the use of opioid analgesics for chronic nonmalignant low back pain: a review Brown, 1996509 Chronic opioid analgesic therapy for chronic low back pain Deyo, 1996415 Drug therapy for back pain. Which drugs help which patients? Furlan, 2006510 Opioids for chronic noncancer pain: A meta-analysis of effectiveness and side effects Kalso, 2004511 Opioids in chronic non-cancer pain: systematic review of efficacy and safety KQ 4 PSYCHOLOGICAL THERAPIES Morley, 1999724 Systematic review and meta-analysis of randomized controlled trials of cognitive behavior therapy and behavior therapy for chronic pain in adults, excluding headache Outdated
Did not evaluate benefits or harms associated with modified work Outdated Did not evaluate benefits or harms associated with modified work Outdated Systematic methods not reported Outdated
Outdated Systematic methods not reported Outdated
Outdated Outdated
Did not clearly use systematic methods
Not a systematic review Outdated Systematic methods not reported Not specific to low back pain
Not specific to low back pain

Author, year, title Scheer, 1997 Randomized controlled trials in industrial low back pain. Part 3. Subacute/chronic pain interventions Turner, 1996597 Educational and behavioral interventions for back pain in primary care van der Weide, 1997346 Vocational outcome of intervention for low-back pain van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions van Tulder, 2000723 Behavioral treatment for chronic low back pain. A systematic review within the framework of the Cochrane Back Review Group KQ 4 SPINAL MANIPULATION Abenhaim, 1992764 Twenty years of randomized clinical trials of manipulative therapy for back pain: a review Anderson, 1992765 A meta-analysis of clinical trials of spinal manipulation Assendelft, 1992766 The efficacy of chiropractic manipulation for back pain: blinded review of relevant randomized clinical trials Assendelft, 1995767 The relationship between methodological quality and conclusions in reviews of spinal manipulation Assendelft, 1996768 The effectiveness of chiropractic for treatment of low back pain: an update and attempt at statistical pooling Assendelft, 1996769 The effectiveness of chiropractic for treatment of low back pain: an update and attempt at statistical pooling Beckerman, 1993623 Efficacy of physiotherapy for musculoskeletal disorders: what can we learn from research? Brox, 1999770 Is exercise therapy and manipulation effective in low back pain? di Fabio, 1992771 Efficacy of manual therapy Ernst, 2000777 Does spinal manipulation have specific treatment effects? Ernst, 2001778 Spinal manipulation: a systematic review of sham-controlled, double-blind, randomized clinical trials Ernst, 2004779 Cerebrovascular complications associated with spinal manipulation Koes, 1991773 Outdated
386
Outdated Outdated Outdated Not specific for low back pain
Updated Cochrane review available (Ostelo, 301 2005 )
Outdated Outdated
Outdated Review of reviews Outdated
Outdated
Outdated
Outdated Norwegian language Outdated Not specific for low back pain/lumbar manipulation Not specific for low back pain/lumbar manipulation Cervical manipulation only Outdated

Author, year, title Spinal manipulation and mobilization for back and neck pain: a blinded review Koes, 1996772 Spinal manipulation for low back pain. An updated systematic review of randomized clinical trials Mohseni-Bandpei, 1998774 Spinal manipulation in the treatment of low back pain: a review of the literature with particular emphasis on randomized controlled clinical trials Ottenbacher, 1985775 Efficacy of spinal manipulation/mobilization therapy. A meta-analysis Scheer, 1995345 Randomized controlled trials in industrial low back pain relating to return to work. Part 1. Acute interventions Shekelle, 1992776 Spinal manipulation for low-back pain van der Weide, 1997346 Vocational outcome of intervention for low-back pain van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 SYSTEMIC STEROIDS Deyo, 1996415 Drug therapy for back pain. Which drugs help which patients? Lipetz, 19981033 Oral medications in the treatment of acute low back pain Rozenberg, 1998188 Glucocorticoid therapy in common lumbar spinal disorders KQ 4 TRANSCUTANEOUS ELECTRICAL NERVE STIMULATION (TENS) Beckerman, 1993623 Efficacy of physiotherapy for musculoskeletal disorders: what can we learn from research? Outdated Outdated
Outdated
Outdated Outdated
Outdated Outdated Outdated
Systematic methods not reported Not a systematic review Not a systematic review
Brosseau, 2002702 Updated Cochrane review available 698, 699 ) Efficacy of the transcutaneous electrical nerve stimulation for the treatment of (Khadilkar, 2005 chronic low back pain. A meta-analysis Flowerdew, 1997705 A review of the treatment of chronic low back pain with acupuncture-like transcutaneous electrical nerve stimulation and transcutaneous electrical nerve stimulation Gadsby, 2000704 Transcutaneous electrical nerve stimulation and acupuncture-like transcutaneous electrical nerve stimulation for chronic low back pain Outdated
Updated Cochrane review available (Khadilkar, 2005698, 699)
Milne, 2001703 Updated Cochrane review available 698, 699 ) Transcutaneous electrical nerve stimulation (TENS) for chronic low-back pain (Khadilkar, 2005

Author, year, title van Tulder, 1997 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 TRACTION Beckerman, 1993623 Efficacy of physiotherapy for musculoskeletal disorders: what can we learn from research van der Heijden, 1995682 The efficacy of traction for back and neck pain: a systematic, blinded review of randomized clinical trial methods van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions KQ 4 ULTRASOUND Beckerman, 1993623 Efficacy of physiotherapy for musculoskeletal disorders: what can we learn from research? Gam, 1995717 Ultrasound therapy in musculoskeletal disorders: a meta-analysis Robertson, 2001716 A review of therapeutic ultrasound: effectiveness studies van der Windt, 1999715 Ultrasound therapy for musculoskeletal disorders: a systematic review Sehgal, 2005835 Diagnostic utility of facet (zygapophysial) joint injections in chronic spinal pain: A systematic review of evidence KQ 7 DIAGNOSTIC SACROILIAC JOINT BLOCK McKenzie-Brown, 2005186 A systematic review of sacroiliac joint interventions KQ 7 DIAGNOSTIC SELECTIVE NERVE ROOT BLOCK Everett, 2005833 A systematic review of diagnostic utility of selective nerve root blocks KQ 7 PROVOCATIVE DISCOGRAPHY Shah, 2005 807 Discography as a diagnostic test for spinal pain: A systematic and narrative review KQ 8 BOTULINUM TOXIN Difazio, 2002175 A focused review of the use of botulinum toxins for low back pain KQ 8 CHEMONUCLEOLYSIS Gibson, 1999176 The Cochrane review of surgery for lumbar disc prolapse and degenerative Outdated
193
Outdated
Outdated
Not specific for low back pain
Not specific for low back pain Not specific for low back pain Not specific for low back pain
KQ 7 DIAGNOSTIC INTRA-ARTICULAR FACET JOINT BLOCKS AND MEDIAL BRANCH BLOCK Updated review available (Sehgal, 2007837)
Updated review available (Hansen, 200783)
Updated review available (Datta, 2007834)
Updated review available (Buenaventura, 809 2007 )
Not a systematic review
Updated Cochrane review available (Gibson, 200781, 82)

Author, year, title lumbar spondylosis Gibson, 2000177 Surgery for lumbar disc prolapse Scheer, 1996190 Randomized controlled trials in industrial low back pain relating to return to work. Part 2. Discogenic low back pain Stevens, 1997191 Efficacy of lumbar discectomy and percutaneous treatments for lumbar disc herniation KQ 8 EPIDURAL STEROID INJECTION Abdi, 2005170 Role of epidural steroids in the management of chronic spinal pain: A systematic review of effectiveness and complications
Updated Cochrane review available (Gibson, 81, 82 ) 2007 Outdated
Outdated
Updated review available (Abdi, 200771)
Boswell, 2003172 Updated review available (Abdi, 200771) Epidural steroids in the management of chronic spinal pain and radiculopathy Cannon, 2000173 Lumbosacral epidural steroid injections Haselkorn, 1995178 Epidural steroid injections and the management of sciatica: a meta-analysis Kepes, 1985181 Treatment of backache with spinal injections of local anesthetics, spinal and systemic steroids. A review Koes, 1995184 Efficacy of epidural steroid injections for low-back pain and sciatica: a systematic review of randomized clinical trials Koes, 1999183 Epidural steroid injections for low back pain and sciatica: An updated systematic review of randomized clinical trials Nelemans, 1999187 Injection therapy for subacute and chronic benign low-back pain Rozenberg, 1998188 Glucocorticoid therapy in common lumbar spinal disorders Not a systematic review Outdated Published as abstract only Not a systematic review
Outdated
Outdated
Updated Cochrane review available 86 (Nelemans, 2001 ) Not a systematic review
Rozenberg, 1999189 Outdated Efficacy of epidural steroids in low back pain and sciatica. A critical appraisal by a French task force of randomized trials Scheer, 1996190 Randomized controlled trials in industrial low back pain relating to return to work. Part 2. Discogenic low back pain Scheer, 1997386 Randomized controlled trials in industrial low back pain. Part 3. Subacute/chronic pain interventions van Tulder, 1997193 Conservative treatment of acute and chronic nonspecific low back pain. A systematic review of randomized controlled trials of the most common interventions Outdated
Outdated (only 1 study of injections included in this review) Outdated

Author, year, title Watts, 1995 Outdated A meta-analysis on the efficacy of epidural corticosteroids in the treatment of sciatica KQ 8 FACET JOINT STEROID INJECTION Boswell, 2005171 Therapeutic facet joint interventions in chronic spinal pain: A systematic review of effectiveness and complications Nelemans, 1999187 Injection therapy for subacute and chronic benign low-back pain KQ 8 INTRADISCAL ELECTROTHERMAL THERAPY (IDET) Chou, 2005174 Intradiscal electrothermal annuloplasty Gibson, 1999176 The Cochrane review of surgery for lumbar disc prolapse and degenerative lumbar spondylosis Gibson, 2000255 Surgery for degenerative lumbar spondylosis Heary, 2001179 Intradiscal electrothermal annuloplasty: the IDET procedure Wetzel, 2002196 Intradiscal electrothermal therapy used to manage chronic discogenic low back pain. New directions and interventions KQ 8 INTRADISCAL STEROID INJECTION Gibson, 1999176, 177 The Cochrane review of surgery for lumbar disc prolapse and degenerative lumbar spondylosis Gibson, 2000177 Surgery for lumbar disc prolapse KQ 8 LOCAL INJECTIONS Nelemans, 1999187 Injection therapy for subacute and chronic benign low-back pain KQ 8 PROLOTHERAPY Kim, 2004182 Critical review of prolotherapy for osteoarthritis, low back pain, and other musculoskeletal conditions: A physiatric perspective Yelland, 2004197 Prolotherapy injections for chronic low-back pain KQ 8 RADIOFREQUENCY DENERVATION Boswell, 2005171 Therapeutic facet joint interventions in chronic spinal pain: A systematic review of their role in chronic spinal pain management and complications Hooten, 2005180 Radiofrequency Neurotomy for Low Back Pain: Evidence-Based Procedural Guidelines
194
75 Updated review available (Boswell, 2007 )
Updated Cochrane review available (Nelemans, 200186) Doesn't clearly use systematic methods Updated Cochrane review available (Gibson, 200579, 80) Updated Cochrane review available (Gibson, 79, 80 ) 2005 Not a systematic review Not a systematic review
Updated Cochrane review available (Gibson, 200781, 82) Updated Cochrane review available (Gibson, 81, 82 ) 2007 Updated Cochrane review available 86 (Nelemans, 2001 ) Systematic methods not clearly used
Updated Cochrane review available (Dagenais, 200776) Updated review available (Boswell, 200775)
Focused on technical aspects; did not evaluate efficacy

Author, year, title Manchikanti, 2002 Medial Branch Neurotomy in Management of Chronic Spinal Pain: Systematic Review of the Evidence KQ 8 SACROILIAC JOINT STEROID INJECTION McKenzie-Brown, 2005186 A systematic review of sacroiliac joint interventions Turner, 1995192 Spinal cord stimulation for chronic low back pain: a systematic literature synthesis
185
Reason for exclusion Updated review available (Boswell, 200775)
Updated review available (Hansen, 200783)
KQ 8 SPINAL CORD STIMULATION FOR BACK PAIN WITHOUT FAILED BACK SURGERY SYNDROME Outdated
Wetzel, 2000195 Did not use systematic methods Treatment of chronic pain in failed back surgery patients with spinal cord stimulation: a review of current literature and proposal for future investigation KQ 9 SURGERY Surgery for non-radicular low back pain with common degenerative changes Gibson, 1999176 The Cochrane review of surgery for lumbar disc prolapse and degenerative lumbar spondylosis Gibson, 2000255 Surgery for degenerative lumbar spondylosis Turner, 1992260 Patient outcomes after lumbar spinal fusions Turner, 1993261 Meta-analysis of the results of lumbar spine fusion Surgery for isthmic spondylolithesis Gibson, 1999176 The Cochrane review of surgery for lumbar disc prolapse and degenerative lumbar spondylosis Gibson, 2000177 Surgery for degenerative lumbar spondylosis Gibson, 1999176 The Cochrane review of surgery for lumbar disc prolapse and degenerative lumbar spondylosis Gibson, 2000255 Surgery for degenerative lumbar spondylosis Mardjetko, 1994257 Degenerative lumbar spondylolisthesis. A meta-analysis of literature 19701993 Updated Cochrane review available (Gibson, 200579, 80) Updated Cochrane review available (Gibson, 200579, 80) Updated Cochrane review available (Gibson, 200579, 80) Updated Cochrane review available (Gibson, 200579, 80) Outdated Updated Cochrane review available (Gibson, 79, 80 ) 2005 Updated Cochrane review available (Gibson, 79, 80 ) 2005 Outdated Outdated
Surgery for spinal stenosis with or without degenerative spondylolisthesis
Niggemeyer, 1997258 Outdated Comparison of surgical procedures for degenerative lumbar spinal stenosis: a meta-analysis of the literature from 1975 to 1995 Turner, 1992259 Surgery for lumbar spinal stenosis. Attempted meta-analysis of the literature Surgery for radiculopathy with herniated disc Outdated

Author, year, title Gibson, 1999 The Cochrane review of surgery for lumbar disc prolapse and degenerative lumbar spondylosis Gibson, 2000177 Surgery for lumbar disc prolapse Hoffman, 1993256 Surgery for herniated lumbar discs: a literature synthesis Scheer, 1996190 Randomized controlled trials in industrial low back pain relating to return to work. Part 2. Discogenic low back pain Stevens, 1997191 Efficacy of lumbar discectomy and percutaneous treatments for lumbar disc herniation
176
Reason for exclusion Updated Cochrane review available (Gibson, 81, 82 ) 2007 Updated Cochrane review available (Gibson, 81, 82 ) 2007 Outdated Outdated
Outdated

APPENDIX 7: QUALITY RATINGS OF TRIALS

Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Yes (balanced allocation method) Yes Yes Yes Don't Know Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes (balanced allocation method) Yes Yes Don't know Don't know Yes No Yes No No Yes No No Can't tell Yes Dont know Dont know
Author, year
Randomization Yes (balanced allocation method) Yes Yes Yes Don't Know Yes Don't know Yes Yes Yes Don't know Don't know Don't know Yes Yes Dont know Yes
Concealed treatment allocation Yes (balanced allocation method) Yes Yes Don't know Don't know Dont know Don't know Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Baseline group similarity
Patient blinded
Care provider blinded
Score
KQ 1c IDENTIFICATION AND TREATMENT OF YELLOW FLAGS Gatchel, 2003

306
Don't know Yes Yes Yes Yes Yes Yes No Yes Yes Yes No Yes Yes Yes Dont know Dont know NA NA NA NA NA NA
Don't know Yes Yes Yes No No Don't know Yes No Yes No No Don't know Yes Yes Yes Dont know
Yes Yes No Yes Yes
Yes Yes Yes Yes Yes Yes Don't know Dont know Yes Yes Yes Yes Yes
Yes Yes No Yes No No No Yes No Yes Yes Yes No Yes No No Dont know
6/9 7/9 7/9 6/9 4/9 5/8 3/8 6/8 5/8 7/8 6/8 4/8 4/8 8/9 7/9 4/9 3/9
George, 2003308 Hay, 2005

304
Jellema, 2005305 Von Korff, 2005 Deyo, 1987330 Djais, 2005

331 307
KQ 2d IMAGING
Gilbert, 2004334 Jarvik, 1997 Jarvik, 2003 Kerry, 2002

338 337
NA
Kendrick, 2001332
333 335, 336
Modic, 2005
KQ 3 ADVICE TO REMAIN ACTIVE Pengel, 2007362 Frost 2004 Little 2001

361
Yes Don't know Dont know Dont know
Yes Yes Yes Dont know
363
Stankovic, 1990, 1995364, 365


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis
Author, year
Randomization
Concealed treatment allocation
Baseline group similarity Dont know Yes Yes Don't know No Yes Yes Yes Yes Yes Dont know Yes Dont know No (pop. w/ previous LBP) Dont know Yes
Patient blinded
Score
KQ 3 ADVICE TO RESTRICT EARLY MORNING FLEXION Snook, 1998377, 378 No No No NA Dont know Dont know Dont know No Yes Yes 2/10
KQ 3 LAY-FACILITATED GROUPS FOR SELF-CARE Haas, 2005379 Von Korff, 1998380 KQ 3 MATTRESSES Atherton, 1983396 Garfin, 1981397 Kovacs, 2003395 Burton, 1999374 Cherkin, 1996368 Cherkin, 1998367. Cherkin, 2001 Hazard, 2000
369
Dont know Dont know
Yes Dont know
NA
NA
Dont know Dont know
Dont know Dont know
No Yes
No Yes
Yes Yes
No Yes
3/9 5/9
No No Yes Yes Dont know Dont know Yes Dont know Yes Dont know
No No Yes Yes Dont know Yes Dont know Yes Yes Yes
No No Yes
No No Yes
No No Yes Yes Yes Yes Yes Yes No
Yes Don't know Yes Dont know Dont know No Yes Dont know Don't know Dont know
Yes Don't know Yes Yes Yes Yes Yes Yes Yes Yes
Yes No Yes No Yes Yes Yes Yes No No
Yes Don't know Yes Yes Yes Yes Yes Yes Yes Yes
No No Yes No Yes Yes Yes Yes No Dont know
4/11 0/11 11/11 6/9 6/9 7/9 8/9 5/9 5/9 4/9
KQ 3 SELF-CARE BOOKS
370
Linton, 2000309 Little 2001363
NA
NA
Yes
Roberts, 2002
372
Yes
Don't know
Yes
Dont know
Don't know
No
Yes
No
4/9
Roland, 1989373 Sherman, 2005371
No Yes
No Yes
Dont know Yes
Dont know Yes
Yes No
No Yes
Yes Yes
No Yes
2/9 8/9


Author, year
Randomization
Baseline group similarity Don't know Don't know Yes Dont know Dont know Yes Yes Yes Yes Yes Dont know Yes Yes
Patient blinded
Score
KQ 3 SELF-CARE E-MAIL DISCUSSION GROUPS Lorig, 2002375 Don't know Don't know NA NA No Don't know No No Yes Yes 2/9
KQ 3 - SELF-CARE EXERCISE VIDEOTAPE Miller, 2004376 Yes Yes NA NA Don't know Don't know No No Yes No 3/9
KQ 3 SELF-HELP TOOLS FOR BACK SURGERY DECISIONS Deyo, 2000382 KQ 4 ACETAMINOPHEN Doran, 1975393 Hackett, 1988420 KQ 4 ACUPRESSURE Hsieh, 2004578 Hsieh, 2006579 KQ 4 ACUPUNCTURE Brinkhaus, 2006566 Thomas, 2006567 Witt, 2006568 Yes Yes Yes Yes Yes Yes Yes No No NA Dont know No Dont know Dont know No (phys. therapy & manipulation) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes 8/10 7/10 8/10 Yes Yes Yes Yes No No NA Yes Yes Don't know Don't know Yes Don't know No No Yes Yes No No 6/10 5/10 Don't know Don't know Don't know Don't know No Yes No Dont know No Dont know Don't know Don't know Don't know Don't know Yes Yes Yes Yes No No 2/11 3/11 Yes Yes NA NA Don't know Dont know Yes Yes Yes No 6/9
KQ 4 ANTIEPILEPTIC DRUGS Khoromi, 2005503 McCleane, 2001501 Muehlbacher, 2006

504
Yes Yes Yes
Yes Don't know Dont know
Yes Yes Yes
Yes Yes Yes
Yes Yes Yes
Don't know Yes Don't know
Dont know Don't know Don't know
No Yes Yes
Yes Yes Yes
Yes No Dont know
7/11 8/11 7/11


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Don't know Yes Don't know Yes Yes No
Author, year Yildirim, 2003502 KQ 4 ASPIRIN Evans, 1980416
Randomization Don't know
Concealed treatment allocation Don't know
Baseline group similarity Don't know Don't know
Patient blinded Don't know
Care provider blinded Don't know
Score 3/11
Don't know
Don't know
No
No
Yes
Yes
Don't know
Yes
Yes
No
4/11
KQ 4 BRIEF EDUCATIONAL INTERVENTION Indahl, 1995 and 1998603, 604 Karjalainen, 2003 and 2004609, 610 Molde Hagen, 2000 and 2003607, 608 Niemisto, 2003 and 2005611, 612 KQ 4 EXERCISE UK BEAM Trial, 2004629 Don't know Don't know Yes NA NA No Don't know No No Yes No 2/9 No Yes No Don't know Yes Yes Yes Yes Yes Yes Don't know Yes NA NA Don't know Yes Don't know Yes Don't know Yes Yes Yes Yes Yes Yes Yes 4/9 8/9 Yes No Dont know Yes Dont know Yes Yes No Yes Yes Yes Yes 7/9 7/9
KQ 4 FUNCTIONAL RESTORATION (PHYSICAL CONDITIONING, WORK CONDITIONING, OR WORK HARDENING) Gatchel, 2003
306
Yes (balanced allocation method) Yes No Don't know
Yes (balanced allocation method) No No Dont know
Don't know
NA
NA
Don't know
Yes
Don't know
Yes
Yes
Yes
6/9
KQ 4 HYDROTHERAPY McIlveen, 1998638 Sjogren, 1997

639
No Yes Yes (small sample sizes) NA NA
No No No
Dont know Don't know Don't know
Dont know Dont know Dont know
Yes Yes Yes
Yes Yes Yes
No No Yes
3/9 4/9 2/9
Yozbatiran, 2004640


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Yes NA Yes Don't know Yes No Yes Yes No Yes Yes Yes Don't know Yes Yes No Don't know Yes Dont know Dont know Dont know Don't know Don't know Don't know Yes Don't know Yes Yes Yes Yes Yes Yes Don't know Don't know Don't know Don't know Don't know Yes No No No Yes Yes Don't know No Don't know Yes Yes Yes Yes Yes Yes Don't know Yes Yes Yes Yes Yes Yes Yes No No Yes Don't know Don't know Don't know N Yes
Author, year Hurley, 2001661 Hurley, 2004

659
Randomization Dont know Yes Yes Yes Don't know Yes Don't know No Don't know Yes
Concealed treatment allocation Yes Yes No Don't know Don't know Don't know Don't know Don't know Don't know Yes
Baseline group similarity No Yes Yes Yes Yes Yes Don't know Don't know Yes Don't know Yes Don't know Don't know Yes No Yes Don't know Yes
Patient blinded NA No No Yes No Yes Yes No Yes Yes
Score 5/10 7/10 4/10 8/11 3/11 6/11 5/11 1/11 6/11 10/11
KQ 4 INTERFERENTIAL THERAPY
Werners, 1999660 Basford, 1999666 Gur, 2003670 Klein, 1990667 Longo, 1988671 Monticone, 2004672 Soriano, 1998668 Toya, 1994
669
KQ 4 LOW LEVEL LASER THERAPY
KQ 4 OPIOIDS Allan, 2005513 Baratta, 1976514 Gostick, 1989515 Hale, 1997426 Hale, 1999516 Hale, 2005
512
Don't know Don't know Don't know Don't know Don't know Yes Don't know Don't know
Yes Don't know Don't know Don't know Don't know Yes Don't know Don't know
No Yes Yes Yes Yes Yes No No
Yes Don't know Yes No No Don't know Don't know Don't know
Don't know Don't know Don't know Don't know Don't know Don't know Don't know Don't know
No Yes No No Yes No Yes No
Yes Yes Yes Yes Yes Yes Yes Yes
No No No No No No Yes No
4/11 5/11 5/11 5/11 5/11 7/11 3/11 2/11
Jamison, 1998517 Salzman, 1999518


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Dont know (concomitant NSAIDs allowed & tracked, with no diff between gps in use stated, but no data provided)
Author, year
Randomization
Baseline group similarity Dont know (no data on pain intensity or duration)
Patient blinded
Score
Wiesel, 1980353
Dont know
Dont know
Yes
Yes
Yes
Dont know
No
Dont know
No
3/11
KQ 4 PERCUTANEOUS ELECTRICAL NERVE STIMULATION (PENS) Ghoname, 1999712 Weiner, 2003713 Yokoyama, 2004711 Yes Yes Don't know Don't know Don't know Don't know Don't know Yes Yes Don't know Don't know No Yes (to sham diathermy) No No No No No Don't know Yes Don't know Don't know Don't know Don't know Don't know Don't know Don't know No No Yes Yes Yes Yes Don't know Don't know No 2/11 4/11 3/11
KQ 4 SHORT WAVE DIATHERMY Gibson, 1985

673
Dont know
Dont know
Dont know Dont know Yes Yes Don't know No Yes
No
Yes
Dont know
Dont know
Yes
Yes
No
4/11
Rasmussen, 1979674 Sweetman, 1993675 KQ 4 SPA THERAPY Constant, 1995748 Constant, 1998747 Guillemin, 1994746 Konrad, 1992
696
Dont know Yes Don't know Don't know Don't know Don't know
Dont know No No Yes Dont know Dont know
No No
Dont know Dont know Yes
Dont know Dont know Don't know Don't know Don't know Don't know
Dont know Dont know Don't know Dont know Don't know Don't know
Yes Yes Yes Yes Yes Yes
Yes Yes Yes Yes Yes Yes
Yes Yes Yes Yes Yes No
3/11 5/11 5/9 5/9 4/9 4/9
NA
NA
Yes Yes Yes


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Yes Don't know Don't know Yes Yes Yes
Author, year Yurtkuran, 1997749
Randomization Yes
Concealed treatment allocation Don't know
Baseline group similarity Don't know
Patient blinded NA
Care provider blinded NA
Score 5/9
KQ 4 SPINAL MANIPULATION Hurwitz, 2002

780, 781
Yes Yes Don't know
Yes Yes Dont know
Yes Yes Yes
NA NA NA
NA NA NA
Don't know Yes No
Yes (not applicable effectiveness study) No Don't know
No Yes No
Yes Yes No
Yes Yes Yes
Yes Yes No
7/9 5/9 2/9
Santilli, 2006782 UK BEAM Trial, 2004629 Finckh, 2006536 Friedman, 2006

539
KQ 4 SYSTEMIC CORTICOSTEROIDS Yes Yes Yes Don't know Yes Yes Don't know Don't know Yes Yes Don't know Yes Don't know Don't know Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Don't know Don't know Yes Yes Don't know Don't know Yes Yes Yes Yes Yes Yes Don't know Yes Yes Yes Yes No 10/11 11/11 6/11 6/11
Haimovic, 1986537 Porsman, 1979538 KQ 4 TRAMADOL Raber, 1999531 Sorge, 1997532 KQ 4 - ULTRASOUND Ansari, 2006718 Nwuga, 1983719 Roman, 1960720 KQ 4 YOGA
Don't know Don't know
Yes Don't know
Yes Don't know
Yes Don't know
Yes Don't know
No Don't know
Dont' know Don't know
No Don't know
4/11 5/11
Dont know No Dont know
Don't know No Don't know
No Don't know Dont know
Yes Yes Yes
No Don't know Dont know
Don't know Yes Don't know
Don't know Don't know Don't know
Dont know Don't know Dont know
No (5/15) Don't know Don't know
Yes Yes Dont know
No Don't know Don't know
2/11 3/11 1/11


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Don't know NA NA Yes Yes Don't know Yes Yes Don't know No Yes No Yes No Yes Yes Yes Yes Yes No
Author, year Galantino, 2004642 Sherman, 2005 Williams, 2005

371 641
Randomization Don't know Yes Yes
Concealed treatment allocation Don't know Yes Don't know
Baseline group similarity No Yes No Don't know Yes
Patient blinded
Score 2/9 8/9 5/9
KQ 5 DECISION TOOLS AND OTHER METHODS OF PREDICTION Brennan, 2006800 Childs, 2004796 Yes Yes Dont know Yes NA NA Yes Don't know Dont know Don't know Yes Yes Yes 26% vs. 15% attending <50% of sessions No Yes Yes Yes Yes Yes 5/9 7/9
Fritz, 2003799
Yes
Don't know
Yes
NA
NA
Yes
Don't know
Yes
Yes
Yes
7/9
KQ 6 PRIMARY CARE REFERRAL AND MULTIDISCIPLINARY OUTCOMES Hurwitz, 2002

780, 781
Yes
Yes
Yes
NA
NA
Yes (not applicable Don't know effectiveness study)
No
Yes
Yes
Yes
7/9
REFER TO APPENDIX 8 FOR KQ 7 QUALITY SCORES ON DIAGNOSTIC ACCURACY TRIALS KQ 8 BOTULINUM TOXIN INJECTIONS Foster, 2001105 Yes Yes Dont know No Dont know Yes No Yes Yes AQ Dont know Yes Yes Yes Yes 9/11
KQ 8 CHEMONUCLEOLYSIS Bromley, 1984127 Burton, 2000103 Dabezies, 1988133 Fraser, 1982136 Yes No Don't know Don't know Yes No Yes Don't know Yes NA Yes Yes Yes NA Yes Don't know Don't know Yes Yes Yes Don't know Dont know Don't know Don't know Yes Yes Yes Yes Yes No No No Yes Yes Yes No Yes No No Yes 8/11 3/9 7/11 5/11


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Dont know Yes Dont know Yes Dont know Dont know Don't know Dont know Don't know Dont know Yes Yes Yes Yes Yes Yes Yes No Yes No Yes Yes Yes No Yes No Yes Dont know Yes Yes
Author, year Hoogland, 2006110 Javid, 1983144 Krugluger, 2000873 Schwestschenau, 1976162 Wittenberg, 2001122 Ackerman, 2007101 Arden, 2005
123
Randomization No Yes Dont know Don't know No
Concealed treatment allocation No Yes Dont know Don't know Dont know
Baseline group similarity Dont know Don't know No Yes Yes
Patient blinded Dont know Yes Dont know Yes Dont know
Care provider blinded Dont know Yes Dont know Don't know Dont know
Score 3/11 9/11 2/11 6/11 4/11
KQ 8 EPIDURAL STEROID INJECTIONS Yes Yes No Yes Don't know Yes

130
Don't know Yes No Don't know Don't know Dont know Yes Don't know Yes Don't know Dont know
Yes Yes Don't know Don't know Don't know Dont know Yes Don't know Yes Yes Yes
Yes Yes Don't know Yes Yes No Yes Yes Yes Yes No
No Yes Don't know No Don't know No Yes Yes No Yes No
Yes Yes Don't know Yes Yes Dont know Yes Yes Dont know Yes Yes
Yes Yes Don't know Don't know Yes Yes Yes Don't know Dont know Yes Dont know
Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes
Yes No Don't know No No Yes No Yes Yes No Yes
Yes (for pain relief) Yes No Don't know Yes Yes Yes Don't know Yes Yes Yes
Yes Yes Dont know Yes Yes Yes Yes Dont know Yes No Yes
9/11 9/11 1/11 5/11 6/11 5/11 10/11 5/11 7/11 7/11 6/11
Beliveau, 1971125 Breivik, 1976126 Bush, 1991128 Buttermann, 2004842 Carette, 1997
Yes Don't know Dont know Don't know Dont know
Cuckler, 1985132 Dashfield, 2005843 Dilke, 1973

135
Fukusaki, 1988106


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Don't know Yes Yes Yes Don't know Don't know Yes Don't know Dont know No Dont know Don't know Don't know Dont know Yes Yes Yes Yes Yes Yes Yes No Yes Yes No Don't know Don't know Yes No (varied for long-term f/u) No Yes No (varied for long-term f/u) Yes Yes Yes Dont know No Yes No Dont know Dont know Yes
Author, year Helliwell, 1985143 Jeong, 2007111 Karpinnen, 2001145 Klenerman, 1984147 Kraemer, 1997a148 Kraemer, 1997b148 Manchikanti, 2004115
Randomization Don't know Dont know Yes Dont know Don't know Don't know Yes
Concealed treatment allocation Don't know Don't know Yes Dont know Don't know Yes Dont know
Baseline group similarity Don't know Dont know Yes Dont know Don't Know Don't know Yes No (unequal distribution, no baseline pain data) Yes Dont know Yes Yes Don't know
Patient blinded Yes Yes Yes Dont know Don't know Yes Yes
Care provider blinded Don't know No Yes Dont know Don't know Yes No
Score 2/11 4/11 10/11 2/11 2/11 5/11 8/11
Mathews, 1987151
Don't know
Don't know
Yes
Don't know
Yes
Yes
Yes
No
Don't know
Dont know
4/11
Ng, 2005152 Ridley, 1988158 Riew, 2000

159 161
Yes Yes Dont know No Don't know
Yes Dont know Yes Don't know Don't know
Yes Dont know Yes Yes Yes
Yes Dont know Yes No No
Yes Yes Yes Yes Yes
Yes Dont know Dont know Don't know Yes
Yes Yes Yes Yes Yes
Yes Yes Yes Don't know No
Yes Yes Yes Yes No
Yes Dont know Yes Dont know Dont know
11/11 5/11 9/11 5/11 4/11
Rogers, 1992
Snoek, 1977164


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Yes Dont know Yes Yes Dont know No (for long term f/u) Yes Yes Yes No Yes Yes
Author, year WilsonMacDonald, 2005167 Zahaar, 1991169
Randomization Yes
Concealed treatment allocation Yes
Baseline group similarity Yes Don't know
Patient blinded Yes
Care provider blinded Yes
Score 9/11
Don't know
Don't know
Yes
No
Don't know
Don't know
Yes
No
Yes
3/11
KQ 8 FACET JOINT INJECTION OR THERAPEUTIC MEDIAL BRANCH BLOCK Carette, 1991129 Fuchs, 2005
859
Yes Yes Don't know No No
Don't know Dont know Don't know No No
Yes Yes Yes No Dont know No Yes
Yes No Don't know No No
Yes No Don't know No No
Don't know Yes Yes No Dont know
No Dont know Don't know Dont know Dont know
Yes Yes Yes Yes Yes
Yes No Don't know Yes No
Yes Yes Dont know Yes No
7/11 6/11 4/11 3/11 2/11
Lilius, 1989150, 860 Manchikanti, 2001862 Nash, 1989116
KQ 8 INTRADISCAL ELECTROTHERMAL THERAPY (IDET) Freeman, 2005138 Pauza, 2004 Buttermann, 2004104 Graham, 1975109 Khot, 2004112 Simmons, 1992119
157
Don't know Yes
Yes Don't know
Yes Yes
Yes No
Yes Yes
Don't know Yes
Yes Yes
Yes Yes
Yes Yes
Yes No
8/11 8/11
KQ 8 INTRADISCAL STEROID INJECTIONS Yes No Dont know Dont know Dont know No Dont know Dont know Yes Dont know Yes Dont know Yes No Dont know Yes Yes No No No Yes Yes Yes Dont know Yes No Dont know Dont know Dont know Yes Yes Yes Yes No Yes No Dont know No Dont know Yes Yes No Yes No Yes 5/11 4/11 4/11 6/11
KQ 8 LOCAL INJECTIONS Collee, 1991131 Don't know Don't know Yes Yes Yes Yes Yes Don't know Yes Dont know 7/11


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Yes Don't know Don't know Yes Yes Don't know Yes Don't know Yes No No Yes Yes Yes Yes Yes Dont know Yes
Author, year Garvey, 1989140 Hameroff, 1981142 Sonne, 1985165
Randomization Yes Don't know Don't know
Concealed treatment allocation Don't know Don't know Don't know
Baseline group similarity Don't know Don't know Don't know
Patient blinded Yes Don't know Don't know
Care provider blinded Yes Don't know Don't know
Score 8/11 2/11 4/11
KQ 8 PERCUTANEOUS INTRADISCAL RADIOFREQUENCY THERMOCOAGULATION (PIRFT) Barendse, 2001124 KQ 8 PROLOTHERAPY Dechow, 1999134 Klein, 1993
146 151
Yes
Yes
Yes
Yes
Yes
Yes
Don't know
Yes
Yes
Yes
Yes
10/11
Yes Yes Don't know Yes Yes
Don't know Don't know Don't know Don't know Yes
Yes Yes No Yes Yes Don't know Yes Yes No Yes
Yes Yes Yes No (manipulation) Yes Don't know Yes Yes Yes No
Yes Yes Don't know No Yes Don't know Yes Don't know Yes Dont know
Yes Yes Yes Yes Yes
Don't know Yes Yes No Yes
Yes Don't know Yes Don't know Yes
Yes Yes No Yes Yes
9/11 9/11 4/11 6/11 11/11
Mathews, 1987
Ongley, 1987156 Yelland, 2004168 Gallagher, 1994139 Geurts, 2003108 Leclaire, 2001149 Nath, 2008117 Oh, 2004118
KQ 8 RADIOFREQUENCY DENERVATION Don't know Yes Don't know Yes Dont know Don't know Yes Yes Dont know Dont know Yes Yes Yes Yes Dont know Don't know Yes Yes Dont know Dont know Yes Yes Yes Yes Yes No Yes Yes Yes Yes (none reported) Yes Yes Yes Yes Yes No Yes Yes Yes Yes 3/11 11/11 9/11 8/11 5/11


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Don't know Yes Yes Don't know Yes Yes Yes Yes Yes No No Yes Yes Yes Yes Yes Dont know Yes
Author, year Tekin, 2007120 van Kleef, 1999166 van Wijk, 2005 Luukkainen, 2002114 Brox, 2003245 Brox, 2006
244 246 121
Randomization Yes Yes Yes
Concealed treatment allocation Don't know Don't know Yes
Baseline group similarity Yes Don't know Yes
Patient blinded Don't know Yes Yes
Care provider blinded Don't know Yes Yes
Score 5/11 7/11 11/11
KQ 8 SACROILIAC JOINT STEROID INJECTION Dont know Dont know Yes Yes No Yes Dont know Yes Dont know Yes Yes 6/11
KQ 9 FUSION FOR NON-RADICULAR LOW BACK PAIN WITH COMMON DEGENERATIVE CHANGES Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Don't know NA NA NA NA Yes Yes No Dont know Don't know Dont know Dont know Dont know Dont know Don't know Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Don't know 8/9 8/9 6/9 7/9 5/9
Fairbank, 2005
Fritzell, 2001247 Hallett, 2007263
KQ 9 DISK REPLACEMENT SURGERY FOR NON-RADICULAR LOW BACK PAIN WITH DEGENERATIVE DISC DISEASE Blumenthal, 2005252 Zigler, 2007253 Yes Don't know Yes Yes Yes Yes Don't know Yes Yes Yes Yes NA NA No No NA No No Dont Know Don't know Yes Yes Yes Yes Yes Yes Yes No 7/10 5/10
KQ 9 SURGERY FOR ISTHMIC SPONDYLOLISTHESIS Inamdar, 2006233 Kim, 2006235 Swan, 2006928 Videbaek, 2006927 Moller, 2000
925
Don't know Don't know No Don't know No
Don't know Don't know No Don't know No
Don't know Don't know Yes Don't know Dont know
Don't know Don't know Don't know Don't know Dont know
Yes Yes Yes Yes Dont know
Don't know No Yes (93% at 2 years) No Yes
Yes Yes Yes Yes Yes
Don't know No No (93%) No Yes
2/9 3/9 5/9 3/9 4/9


Author, year Amundsen, 2000243 Anderson, 2006231 FernandezFairen, 2007232 Inamdar, 2006233 Kim, 2006235 Malmivaara, 2007236
Randomization
Baseline group similarity
Patient blinded
Score
KQ 9 SURGERY FOR SPINAL STENOSIS WITH OR WITHOUT DEGENERATIVE SPONDYLOLISTHESIS Yes Don't know Yes Don't know Don't know Yes Dont know Yes Don't know Don't know Don't know Yes Yes Yes Yes Don't know Yes No Don't know (Baseline characteristics for intervention group not reported) Dont know Dont know (prior epidural/ 64 vs 48%) Don't know Dont know Don't know Dont know Don't know Don't know No Dont know Don't know Dont know Don't know Don't know No Yes Yes Yes Yes Yes Yes Yes Yes Don't know Don't know No Yes Yes Yes Yes Yes Yes Yes Yes Don't know Yes Don't know No Yes 6/9 5/9 5/9 2/9 3/9 6/9
NA
NA Don't know Don't know No Yes Yes Yes 4//9
Weinstein, 2007241
Yes
Don't know
Weinstein, 2008250
Yes
Dont know
Dont know
Yes
No
Yes
Yes
Yes
5/9
Zucherman, 2004251
Dont know
Yes
Dont know
Dont know
Dont know
No
Yes
Dont know
2/9
KQ 9 SURGERY FOR RADICULOPATHY WITH HERNIATED LUMBAR DISC Katayama, 2006234 Don't know Don't know NA NA Don't know Don't know Yes (assumed) No (not described) Don't know Don't know 1/9


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Dont know Dont know Dont know Dont know Yes Yes Yes No Yes Yes No No
Author, year Hoogland, 2006110

970
Randomization No
Concealed treatment allocation No Yes
Baseline group similarity Dont know Dont know Yes (duration of symptoms 77 vs. 60 days) Yes Dont know Yes
Patient blinded
Score 3/9 3/9
Haines, 2002
949,
Dont know
Osterman, 2006237
Yes
Yes
NA
NA
Dont know
No (exercise)
No
Yes
Yes
Yes
6/9
Peul, 2007238, 239 Weber, 1983248 Weinstein, 2006249
Yes Dont know Yes
Yes Dont know Yes
No No Dont know
Don't know Dont know Yes
Yes Yes No
Yes Dont know Yes
Yes Yes Yes
Yes Yes No
7/9 4/9 6/9
KQ 10 COMBINATION MEDICATION THERAPIES Jamison, 1998517 Don't know Don't know Don't know Yes Yes Yes Yes NA NA No No No Don't know Don't know Yes Ys Yes 3/11
KQ 10 COMBINATION THERAPIES FOR SPINAL STENOSIS Whitman, 2006630 Pengel, 2007362 UK BEAM Trial, 2004629 Hurwitz, 2002780,
781
Yes Yes Dont know Yes
Yes Yes Dont know Yes
NA
NA
Yes Yes No Dont know
Dont know Yes Dont know Yes (NA:effectiveness study) Dont know
Yes Yes No No
Yes Yes No Yes
Yes Yes Yes Yes
Yes Dont know No Yes
8/9 8//9 2/9 7/9
KQ 10 EXERCISE OR SPINAL MANIPULATION COMBINED WITH OTHER INTERVENTIONS
KQ 10 SELF-CARE ADVICE COMBINED WITH OTHER INTERVENTIONS Cherkin, 1996368 Dont know Dont know Yes NA NA Yes Yes Yes Yes Ys 6/9


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Yes NA NA Yes Yes No Yes Dont know No Yes Yes Yes No No Dont know No No Dont know Dont know No Don't know Dont know No Yes Yes Dont know Yes Dont know Dont know Dont know No Don't know Dont know Dont know Dont know Dont know Dont know Dont know Yes Yes Dont know Dont know Yes Yes Yes Yes Dont know Yes No No No No No Yes No Yes Yes No Yes No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Dont know Yes No Yes No Yes Yes No Dont know No
Author, year Hurley, 2001661 Little, 2001363 Wand, 2004977 Wright, 2005
978
Randomization Dont know Dont know Yes Yes Don't know Dont know No Yes Dont know Dont know No
Concealed treatment allocation Yes Yes Yes Yes Yes Dont know No Dont know Dont know Dont know Dont know
Baseline group similarity No Dont know Yes Yes Yes Dont know No Yes No Yes Dont know Yes Yes Don't know
Patient blinded
Score 5/9 4/9 6/9 3/9 7/11 2/11 3/11 8/11 3/11 5/11 2/11
KQ 11 FAILED SURGERY, ADHESIOLYSIS AND FORCEFUL EPIDURAL INJECTIONS Dashfield, 2005843 Heavner, 1999985 Manchikanti, 2001986 Manchikanti, 2004115 Meadeb, 2001848 Revel, 1996850 Veihelmann, 2006987 Timm, 1994998 Kumar, 2007113 North, 2005153
KQ 11 FAILED SURGERY, NON-INVASIVE Dont know Yes Yes Dont know Yes Yes NA NA Dont know Don't know NA NA No Dont know Yes Yes Dont know No Yes Dont know Yes Yes Yes Yes Yes Dont know No No 2/9 6/9 6/9 KQ 11 FAILED SURGERY, SPINAL CORD STIMULATION
KQ 12 COORDINATION OF CARE/ SECONDARY PREVENTION Meeuwesen, 19961003 No No Yes NA NA Dont know Dont know No No Yes No 2/9


Timing of Compliance outcome Outcome Cointerventions acceptable Drop-out rate assessment in Intention to all groups assessor avoided or described and treat in all groups similar blinded similar acceptable analysis Yes Dont know Dont know Yes Yes No
Author, year Rossignol, 20001002 Molde Hagen, 2003608 KQ 13 EXERCISE Hides, 20011006 Kellett, 19911007 Soukup, 1999 Stankovic, 1995364
1008
Randomization Yes
Concealed treatment allocation Yes
Baseline group similarity No
Patient blinded NA
Care provider blinded NA
Score 4/9
KQ 13 ADVICE TO STAY ACTIVE No Yes Dont know No Dont know No Dont know NA NA NA NA NA NA Dont know Dont know Dont know Yes Yes Yes 4/9
Dont know Dont know Dont know Yes
Dont know Dont know Dont know Yes
Yes Dont know No Dont know
Yes Dont know Dont know Dont know
Dont know No Yes Dont know
Yes No Yes Dont know
Yes Yes Yes Yes
No No No Dont know
4/9 1/9 3/9 3/9
KQ 13 OCCUPATIONAL INTERVENTION Verbeek, 2002

1009
Yes
Yes
Yes
NA
NA
Yes
Yes
No (24% c/o in control group)
Yes
Yes
Yes
4/9
KQ 14 ACUPUNCTURE DURING PREGNANCY Guerreiro da Silva, 20041013 Kvorning, 20041014 Suputtitada, 20021016 No Yes No Yes No No No NA No Yes No Yes No Yes No 5/10 Dont know Dont know Yes Yes Yes No 4/10
KQ 14 PHYSICAL THERAPY DURING PREGNANCY Dont know Dont know Yes NA NA Dont know Dont know Dont know Yes Yes No 3/9
KQ 14 MASSAGE DURING PREGNANCY Field, 20041023 Dont know Dont know Dont know NA NA No Dont know Dont know No Yes No (not stated) 1/9

APPENDIX 8: QUALITY RATINGS OF DIAGNOSTIC ACCURACY TRIALS

KQ 7 DIAGNOSTIC INTRA-ARTICUALR FACET JOINT BLOCK AND MEDIAL BRANCH BLOCK/OUTCOMES Drop-out Rate Described Cointerventions Compliance and Avoided or Acceptable in Similar All Groups Acceptable Don't know YES Dont know Timing of Outcome Assessment in All Groups Similar YES
Author, year Birkenmaier, 2007838
Randomization Yes
Concealed Treatment Allocation Dont know
Baseline Group Similarity Don't know
Patient Blinded Dont know
Care provider Blinded NO
Outcome Assessor Blinded Dont know
Intention to Treat Analysis Don't know
Score 3/11
KQ 7 DISCOGRAPHY/POSITIVE RATES IN PERSONS WITHOUT SIGNIFICANT LOW BACK PAIN Consecutive series or random Author, year Prospective subset Carragee, 1999815 Carragee, 2000813 Carragee, 2000814 Carragee, 2002812 Carragee, 2006816 Derby, 2005817 Walsh, 1990811 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Evaluates patients with a spectrum of symptoms No Yes Yes Yes Yes Yes No Adequate description of discography technique Yes Yes Yes Yes Yes Yes Yes Use of current discography technique Yes Yes Yes Yes Yes Yes Yes Adequate description of criteria for positive test Yes Yes Yes Yes Yes Yes Yes Appropriate definition for positive test Yes Yes Yes Yes Yes Dont know Yes Statistical analysis Investigator not of predictors for aware of clinical positive tests symptoms Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Dont know Yes
Score 8/9 9/9 9/9 9/9 8/9 7/9 8/9

KQ 7 DISCOGRAPHY/PREDICTORS OF POSITIVE RESPONSES Adequate description Evaluates patients of with a spectrum of discography Prospective symptoms technique Yes No Dont know Yes Dont know Dont know Yes Dont know Yes No No No Adequate descriptio n of criteria for positive test Yes Dont know Yes Yes
Author/year Block, 1996818 Heggeness, 1997819 Manchikanti, 2001820 Ohnmeiss, 1995821
Consecutive series or random subset Yes Yes Yes Yes
Use of current discography technique Yes Dont know Yes Dont know
Appropriate definition for positive test Yes Dont know Yes Yes
Statistical analysis of predictors for positive tests Yes Yes Yes Yes
Investigator not aware of clinical symptoms Yes Dont know Dont know Dont know
Score 7/9 2/9 6/9 5/9
KQ 7 DISCOGRAPHY/OUTCOMES Concealed treatment Author, year Randomization allocation Carragee, 2006823 Madan, 2002822 NA NA NA NA Baseline group similarity Yes Don't know Care provider blinded NA NA Timing of Outcome Compliance Drop-out rate outcome assessor Cointerventions acceptable in described and assessment in all blinded avoided or Similar all groups acceptable groups similar Yes Don't know Don't know Don't know Yes Yes Yes Yes Yes Yes Intention to treat analysis Yes Yes
Patient blinded NA NA
Score 5/9 4/9

APPENDIX 9. SYSTEMATIC REVIEWS PUBLISHED TOO RECENTLY TO BE INCLUDED IN THIS EVIDENCE REVIEW
Author, year Intervention Ammendolia, 20081034 Bronfort, 20081035 Brox, 20081036 Brox, 20081037 Carreon, 20081038 Chrubasik, 20071039 Clarke, 20071040 Dagenais, 20081041 Dagenais, 20081042 Deshpande, 20071043 Engers, 20081044 Freeman, 20081045 Furlan, 20081046 Gagnier, 20081047 Gatchel, 20081048 Gay, 20081049 Henchoz, 20081050 Imamura, 20081051 Khadilkar, 20081052 Lawrence, 20081053 Liddle, 20071054 Machado, 2008 e-published ahead of 1055 print Macedo, 20091056 Martell, 20071057 Mayer, 20081058 Norlund, 20091059 Novak, 20081060 Pennick, 20071061 Perrot, 20081062 Poiraudeau, 20071063 Poitras, 20081064 Racz, 20081065 Roelofs, 20081066, 1067 Sahar, 20071068 Slade, 20071069 Standaert, 20081070 Stuber, 20081071 Urquhart 20081072 van Duijvenbode, 20081073 van Geen, 20071074 Vlachojannis, 20081075 Wai, 20081076 Williams, 20071077 Williams, 20071078 Wolfer, 20081079 Yousefi-Nooraie, 20071080 Yuan, 20081081 Yuan, 20081082
Intervention Acupuncture Spinal manipulation Bach schools, brief education, and fear-avoidance training Back schools, brief education, and fear-avoidance training Surgery for non-radicular low back pain Herbal therapy Traction Prolotherapy Medication-assisted manipulation Opioids Individual patient education IDET, percutaneous discectomy, and nucleoplasty Massage Herbal supplements Cognitive-behavioral therapy Traction Exercise therapy Massage TENS Spinal manipulation Advice Various non-interventional, non-surgical therapies Motor control exercise therapy Opioids Lumbar extensor strengthening exercises Interdisciplinary rehabilitation Epidural steroid injection Interventions for back and pelvic pain during pregnancy Antidepressants Functional restoration TENS, interferential therapy, electrical muscle stimulation, ultrasound, and thermotherapy Adhesiolysis Non-steroidal anti-inflammatory drugs Insoles Unloaded movement facilitation exercise therapy Lumbar stabilization exercises Spinal manipulation during pregnancy Antidepressants Lumbar supports Interdisciplinary rehabilitation Herbal therapy Physical activity, smoking cessation, and weight loss Spinal manipulation Workplace rehabilitation Provocative discography Low level laser therapy for nonspecific low back pain Acupuncture Acupuncture

APPENDIX 10: GLOSSARY

Acupressure Acupuncture Acute low back pain Back school An intervention consisting of manipulation with the fingers instead of needles at specific acupuncture points. An intervention consisting of the insertion of needles at specific acupuncture points. Low back pain present less than four weeks duration (sometimes grouped with subacute low back pain as symptoms present for less than 3 months). An intervention consisting of an education and a skills program, including exercise therapy, in which all lessons are given to groups of patients and supervised by a paramedical therapist or medical specialist. The use of auditory and visual signals reflecting muscle tension or activity in order to inhibit or reduce the muscle activity. Individualized assessment and education about low back pain problems without supervised exercise therapy or other specific interventions. Compression (usually due to a massive, centrally herniated disc) on nerve roots from the lower cord segments, often resulting in urinary retention or incontinence from loss of sphincter function, bilateral motor weakness of the lower extremities, and saddle anesthesia. Treatment of herniated discs with intradiscal injections of an enzyme extracted from papaya (chymopapain). Chymopapain acts by digesting the jelly-like inner portion of the disc known as the nucleus pulposus, while at the same time, leaving the outer portion, the annulus fibrosis, essentially intact. Low back pain present more than 3 months. An intervention that involves working with cognitions to change emotions, thoughts, and behaviors. A measure of the difference in outcome between intervention groups. Commonly expressed as a risk ration (relative risk), odds ratio, or risk difference for binary outcomes and as difference in means for continuous outcomes. May be referred to as treatment effect. Either a supervised exercise program or formal home exercise regimen. Exercise therapy can range from programs aimed at general physical fitness or aerobic exercise to programs more specifically aimed at muscle strengthening, flexibility, or stretching, or different combinations of these elements. Injection of long lasting steroid ("cortisone") in the facet joints part of the bony structure in the back. The steroid injected reduces the inflammation and/or swelling of tissue in the joint space. This may in turn reduce pain, and other symptoms caused by inflammation or irritation of the joint and surrounding structure. An intervention that involves simulated or actual work tests in a supervised environment in order to enhance job performance skills and improve strength, endurance, flexibility, and cardiovascular fitness in injured workers. Herniation of the nucleus pulposus of an intervertebral disc through its fibrous outer covering. An intervention that combines and coordinates physical, vocational, and behavioral components and is provided by multiple health care professionals with different clinical backgrounds. The intensity and content of interdisciplinary therapy varies widely. The superficial application of a medium frequency alternating current modulated to
Biofeedback Brief educational interventions Cauda equina syndrome
Chemonucleolysis
Chronic low back pain Cognitive behavioral therapy or treatment (CBT) Effect size
Exercise
Facet joint injection
Functional restoration (also referred to as work hardening or work conditioning) Herniated disc Interdisciplinary therapy (also referred to as multidisciplinary therapy) Interferential therapy


produce low frequencies up to 150 Hz. Low-level laser therapy Massage Neurogenic claudication Neuroreflexotherapy The superficial application of lasers at wavelengths between 632 and 904 nm. Optimal treatment parameters (wavelength, dosage, dose-intensity) are uncertain. Soft tissue manipulation using the hands or a mechanical device through a variety of specific methods. Symptoms of leg pain (and occasionally weakness) on walking or standing, relieved by sitting or spinal flexion, associated with spinal stenosis. A technique from Spain characterized by the temporary implantation of staples superficially into the skin over trigger points in the back and referred tender points in the ear. Pain occurring primarily in the back with no signs of a serious underlying condition such as cancer, infection, fracture, or cauda equine syndrome. The proportion of people with a positive test who have the disease. A technique that involves the deliberate tensing and relaxation of muscles, in order to facilitate the recognition and release of muscle tension. A procedure that uses a dextrose (sugar water) solution, which is injected into the ligament or tendon where it attaches to the bone. This causes a localized inflammation in these weak areas which then increases the blood supply and flow of nutrients and stimulates the tissue to repair itself. Includes biofeedback (the use of auditory and visual signals reflecting muscle tension or activity to inhibit or reduce the muscle activity), progressive relaxation (a technique that involves the deliberate tensing and relaxation of muscles to facilitate the recognition and release of muscle tension), and standard cognitive-behavioral and operant therapy. Dysfunction of a nerve root associated with pain, sensory impairment, weakness, or diminished deep tendon reflexes in the nerve root distribution. Pain radiating down the leg below the knee in the distribution of the sciatic nerve, suggesting nerve root compromise due to mechanical pressure or inflammation. The proportion of people who truly have a designated disorder who are so identified by the test. The test may consist of, or include, clinical observations. The proportion of truly diseased persons in the screened population who are identified as diseased by the screening testthat is, the true-positive rate. An inactive treatment or procedure that is intended to mimic as closely as possible a therapy in a clinical trial. Therapeutic elevation of the temperature of deep tissues by application of shortwave electromagnetic radiation with a frequency range from 10 to 100 MHz. One of several alternative modalities used in Europe to relieve LBP and other chronic ailments. Specific thermal techniques are used, including mineral water bathing, thermal techniques, and a temporary change of lifestyle. Most patients are treated while staying in spa resorts. The proportion of people who are truly free of a designated disorder who are so identified by the test. The test may consist of, or include, clinical observations. The proportion of truly non-diseased persons who are identified as such by the screening test; that is, the true-negative rate.
Nonspecific back pain Positive predictive value Progressive relaxation Prolotherapy
Psychological therapies
Radiculopathy Sciatica Sensitivity
Sham therapy
Shortwave diathermy Spa treatment
Specificity


Spinal manipulation Manual therapy in which loads are applied to the spine using short or long lever methods. High velocity thrusts are applied to a spinal joint beyond its restricted range of movement. Spinal mobilization, or low-velocity, passive movements within or at the limit of joint range, is often used in conjunction with spinal manipulation. Narrowing of the spinal canal that may result in bony constriction of the cauda equina and the emerging nerve roots. An effect size measure for continuous variables, computed as the difference between two means divided by the variability of that difference. A procedure in which the hip is flexed with the knee extended in order to passively stretch the sciatic nerve and elicit symptoms suggesting nerve root tension. A positive test is usually considered reproduction of the patients sciatica when the leg is raised between 30 and 70 degrees. Reproduction of the patients sciatica when the unaffected leg is lifted is referred to as a positive crossed straight leg raise test. Low back pain present between 4 weeks and 3 months. A diagnostic procedure that images the infrared radiation (heat) emitted from body surfaces. Use of a small battery-operated device to provide continuous electrical impulses via surface electrodes, with the goal of providing symptomatic relief by modifying pain perception. An intervention distinguished from traditional exercise therapy by the utilization of specific body positions, breathing techniques, and emphasis on mental focus. Many styles of yoga are practiced, each emphasizing different postures and techniques.
Spinal stenosis Standardized mean difference Straight leg raise test
Subacute low back pain Thermography Transcutaneous electrical nerve stimulation (TENS) Yoga

Guidelines For Evaluation and Management of LBP

Uploaded by

Copyright:

Available Formats

Guidelines For Evaluation and Management of LBP

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Guidelines For Evaluation and Management of LBP

Uploaded by

Copyright:

Available Formats

C L I N I C A L G U I D E L I N E F O R T H E E VA L U AT I O N A N D M A N A G E M E N T O F

Evaluation and Management of Low Back Pain

roger Chou, Md laurie hoyt huffman, MS

American Pain Society, Publisher Glenview, il

TABLE OF CONTENTS Introduction

Purpose of evidence review ......................................................................1 Background

Previous guidelines...................................................................................2 Scope of evidence review

TABLE OF CONTENTS Results

SCOPE OF EVIDENCE REVIEW

Non-invasive interventions Medications

Other non-invasive interventions

Interdisciplinary (multidisciplinary) rehabilitation

Invasive, non-surgical interventions

Invasive diagnostic tests

Inclusion and exclusion criteria

Data extraction and synthesis Systematic reviews

Individual trials on efficacy and safety of interventions

Observational studies of treatment efficacy

Studies of invasive diagnostic tests

Assessing research applicability and clinical relevance

Assessing magnitude of benefits or harms

Cauda equina syndrome

Herniated disc or radiculopathy

Recommendations and findings from other guidelines

Author, year Borge, 2001296

19 studies of prognostic factors for low back pain

18 studies of prognostic factors for low back pain 54

American Pain Society

Author, year Hartvigsen, 2004287

McIntosh, 290 2000

16 studies of prognostic factors for low back pain 9

6 studies of prognostic factors for low back pain

American Pain Society

Author, year Pincus, 2002291

Pincus, 2006292 Shaw, 2001293

American Pain Society

Author, year Steenstra, 2005294

Truchon, 295 2000

*Using Oxman criteria, maximum score 7 on a 1 to 7 scale

American Pain Society

Recommendations and findings from other guidelines

Efficacy of interventions for identifying and treating yellow flags

Author, year Hay, 2005304

Quality score* 7/9

Author, year Gatchel, 2003306

Quality score* 6/9

Table 4. Trials of fear-avoidance based interventions

Author, year George, 2003308

Von Korff, 2005307

Quality score* 7/9

Recommendations and findings of other guidelines

Key Questions 2a 2d Diagnostic testing

Accuracy of imaging for diagnosing cancer

Sensitivity 0.6 0.83-0.93 0.74-0.98 0.87-0.93

Specificity 0.95-0.995 0.90-0.97 0.64-0.81 0.91-0.93

Source: Jarvik and Deyo, 2002268

Accuracy of imaging for diagnosing vertebral infection

Technique Plain radiography MRI Radionuclide scanning