www.acpsm.

org

Acute Management
of Soft Tissue Injuries
Protection, Rest, Ice, Compression, and Elevation Guidelines

www.acpsm.org

Management of acute soft tissue injury using Protection Rest Ice Compression
and Elevation: Recommendations from the Association of Chartered
Physiotherapists in Sports and Exercise Medicine (ACPSM)
Chris M Bleakley1, Philip D Glasgow2, Nicola Phillips2, Laura Hanna2, Michael J Callaghan2, Gareth W Davison3,
Ty J Hopkins3, Eamonn Delahunt3
1

Lead author and guarantor

ACPSM consensus panel
3
External authors and contributors
2

Acknowledgements to other ACPSM contributors: Lynn Booth, Nicola Combarro, Sian Knott, Chris McNicholl and Colin
Paterson for their assistance with literature searching, data extraction, and interpretation of outcomes.
This work was funded by the Association of Chartered Physiotherapists in Sports and Exercise Medicine (ACPSM). The
guidelines are endorsed by the Chartered Society of Physiotherapys Supporting Knowledge in Physiotherapy Practice
Programme (SKIPP), after peer review from their Good Practice Panel in October 2010.

CONTENTS
Chapter 1: Project methods
Chapter 2: What is the magnitude and depth of cooling associated with ice?
Chapter 3: Can PRICE decrease the inflammatory response after acute soft tissue injury?
Chapter 4: What effect does mechanical loading have on inflammation and soft tissue
healing after acute injury?
Chapter 5: Do the physiological effects of local tissue cooling affect function, sporting performance
and injury risk?
Chapter 6: Which components of PRICE are effective in the clinical management of acute
soft tissue injury?
Chapter 7: Executive summary
Chapter 8: Appendices

Chapter 1
Project methods
Background
The need for guidelines
Soft tissue injury is a common problem in sport, recreational and physical activities. Indeed, sprains, strains and contusions
are the most commonly encountered injuries in amateur and professional sporting disciplines including; professional
football,1-2 gaelic football,3 rugby,4 triathlon,5 and Australian rules football.6 Almost 8-12% of referrals from General
Practitioners for physiotherapy are for soft tissue injuries,7 and it is estimated that the annual cost of sport and exercise
related soft tissue injuries in the United Kingdom and the Netherlands is approximately 9.8 million pounds.8
Following an injury incident one of the main objectives of physiotherapeutic interventions is the restoration of full function
and the re-gaining of pre-injury status. However for a commonly encountered injury such as an ankle sprain it has been
reported that just between 36-85% of individuals report full recovery within 3 years; with one third suffering persistent
pain, instability and re-injury.9 Inadequate recovery after soft tissue injury could have a detrimental effect on future physical
activity levels, and general health. Currently, ankle sprains are one of the primary causes of post-traumatic ankle joint
osteoarthritis,10 and a history of ligament injury and/or joint laxity have been implicated in the aetiology of knee joint
osteoarthritis.11-14
The quality of acute stage management of a soft tissue injury is thought to be an important determinant for short and
long term recovery. In the early stages, soft tissue injuries are characterised by an acute inflammatory response. This is
manifested clinically by the presence of pain, redness, swelling and loss of function. Traditionally, the clinical management
of soft tissue injury has placed most emphasis on minimising or controlling acute inflammation. Protection, Rest, Ice,
Compression and Elevation (PRICE) remains one of the most popular approaches for the management of acute injuries
and in particular the control of inflammation. Some or all components of PRICE continue to be combined, based on the
type or severity of soft tissue injury.
Original ACPSM guidelines
Previous recommendations15 for using PRICE in the management of acute soft tissue injury were provided by the Associated
of Chartered Physiotherapists in Sports Medicine and endorsed by the Chartered Society of Physiotherapy (London).
These recommendations were made based on evidence published up to 1996, and included guidelines for the practical
application of each component of PRICE. Evidence from studies using physiological and clinical outcomes were used to
inform the recommendations. The strength of the recommendations were made based on study quality which was graded
from level three (poorly performed observation research) to level one [well performed randomised controlled trial (RCT)].
In the event that there was no empirical evidence, recommendations were made using consensus agreement from a panel.
The guidelines15 recommended all components of PRICE immediately after acute injury. Protection and rest were
recommended for at least 3 days post injury, with longer periods advised according to injury severity. Only isometric type
exercises were recommended during the acute stages. Compression was recommended immediately after injury, with
continuous use over the first 72 hours only. Guidance on its practical application included: ensuring uniform pressure,
proximal to distal application, with additional intermittent compression permitted. Elevation was recommended for as long
as possible in the first 72 hours, however, immediate restoration of the injured body part to a gravity dependant position
was not advised. The only recommendation based on evidence from an RCT, was to avoid concomitant use of compression
and elevation. Ice application was recommended in the immediate stages post injury. The optimal protocol was chipped
ice (with damp barrier) for durations of 20-30 minutes, repeated every 2 hours; variations were suggested according to
body fat levels, and the presence of a superficial nerve at the injury site. Athletes were advised not to return to competition
immediately after applying ice.
Updating the original guidelines
Making judgements about evidence and developing recommendations for clinical practice is complex and requires a
systematic and explicit approach. The original ACPSM guidelines were developed largely based on Thompsons model16
which followed a stepwise process of: identifying a topic; composing a developmental group with appropriate skills;
literature searching; critical appraisal and formation of recommendations. Recommendations were made and presented
based on a hierarchy of study quality derived from the Canadian Task Force Classifications.17

Recently, the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) working group have
developed a refined template for the development of guidelines in healthcare.18 This was produced to provide a more
systematic, explicit and transparent approach. Our methods will therefore follow the sequential process set out by GRADE.
This has many similarities to methods used in the original guidelines, but with the following updates: we will consider
the importance of outcomes and weight them prior to making recommendations; we will judge the overall quality of
available evidence on: study design, study quality [based on Cochrane risk of bias tool19 (sequence generation, allocation
concealment, assessor blinding, incomplete outcome data)], directness (ie. the extent to which the people, interventions
and outcomes in the studies, are similar to those of interest) and consistency (the similarities of estimates across studies);
finally, we will make judgements on the strength of recommendations based on the overall quality of the evidence, costs
and benefits/harms.
Purpose
The purpose of the current study is to review the clinical effectiveness and pathophysiological rationale for using PRICE
for acute soft tissue injury. We will focus on research published after 1996. In accordance with the GRADE guidelines,18
relevant evidence will be synthesized to provide an update for the original ACPSM recommendations.15 It is anticipated
that the results of the recommendations will target physicians and healthcare professionals (particularly those involved in
sports medicine), first aiders, and adults self managing an acute soft tissue injury.
Overview of methods
In January 2009, a number of ACPSM members volunteered to form a working group responsible for developing the
guidelines. This initial working group was composed of a number of health and allied health professionals (physiotherapy,
physiology and biochemistry), with a diversity of expertise (content experts, front line clinicians, academics). All members
stated that they had no conflicts of interests in participating in the development of the guidelines. One member of the
group (CMB) was assigned to co-ordinate the research.
Our methods were based on the recommendations developed by the GRADE working group.18 The following milestones
were made and addressed in the following order: 1) develop specific clinical questions; 2) literature searching; 3) address
importance of outcomes; 4) extract data/grade quality of evidence; 5) develop final recommendations.
1). Developing specific clinical questions
We devised five clinical questions (or clinically related questions) following group consensus in March 2009. Our
primary aim was to update the clinical evidence for using PRICE in the management of acute soft tissue injuries. Our
first clinical question was:
- Which components of PRICE are effective in the clinical management of acute soft tissue injury?
Our secondary aim was to update the pathophysiological rationale for using PRICE in the management of acute soft tissue
injury. The following 4 questions were considered to be relevant to the clinical effectiveness of PRICE:
- What is the magnitude and depth of cooling associated with ice?
- Can PRICE decrease the inflammatory response after acute soft tissue injury?
- What effect does mechanical loading have on inflammation and healing after acute soft tissue injury?
- Do the physiological effects of local tissue cooling affect function, sporting performance and injury risk?
2). Literature searching
A series of literature searches were undertaken on the following databases; MEDLINE, EMBASE, and the Cochrane
database (Ovid SP). The following limitations were imposed: January 1996-October 2009. The main search strategy
is detailed in Appendix Table 1-3. Each search was undertaken independently by at least two researchers following
standardised written guidelines on the search strategy and study inclusion criteria (CMB; SK). Supplementary searches
were undertaken based on related article searches on Pubmed (http://www.ncbi.nlm.nih.gov/pubmed/), citation tracking
of original and review articles (n=350) and a convenience sample of text books (Appendix Table 4a), we also checked the
results of literature searches from related reviews, previously completed by the lead author (Appendix Table 4b). Studies
were included or excluded (with reasons) according to their relevance to each clinical question (Appendix Table 5).

3). Addressing the importance of outcomes

Outcomes were differentiated according to their importance to patients with an acute soft tissue injury. Members of the
working group were presented with a summary of the outcomes and a nine point scale was used to gain consensus on
their importance. The upper end of the scale, 7-9 identified outcomes of critical importance to an individual with an acute
soft tissue injury; 4-6 represents outcomes that are important but not critical for decision making, and rating of 1-3 were
not important for decision making. As these guidelines covered a wide range of both interventions and injuries, we subgrouped outcomes for ease of discussion. The hierarchy of outcomes according to their importance to subjects with acute
soft tissue injuries is shown in Figure 1.

Clinical outcomes from human subjects with acute,

closed soft tissue injury (eg. pain, function, strength)

(Critical)
9
8
7

Physiological/biochemical outcomes from human

subjects with acute, closed soft tissue injury
(eg. blood flow, inflammatory biomarkers,
temperature reduction)

Clinical/physiological/biochemical outcomes
from healthy subjects (eg. strength, blood flow,
temperature reduction)

Physiological/biochemical outcomes from injured

animal models

Any outcomes derived from studies using

interventions in vitro

(Important not critical)

6
5
4

(Not important)
3
2
1

4) Data extraction / grading the quality of evidence

Members of the ACPSM working group were each assigned to small specialist working groups relating to each of the 5
clinical questions. The scientific evidence/data relevant to each clinical question was then collected, extracted, graded
and summarised following a standardised written protocol. A separate review article was then written up for each clinical
question.20-24 Full methodological details are described in each chapter within the main document. Each review was
sent for evaluation by relevant external professionals or experts at this stage (GD, TH, ED), and modifications made if
applicable. This process was completed for each specialist working group by using either group meetings or iterative
discussion by e-mail.
5) Developing final recommendations
Final recommendations were made in September 2010. Each member of the working group attended this one day
meeting to review the process by which final recommendations would be made. Prior to this meeting each member was
provided with a full written copy of the relevant evidence (Chapter 2-7 of the main document).20-24

The strength of the recommendations (for or against the intervention) was graded as strong (definitely do: indicating
judgement that most well informed people will make the same choice); weak (probably do: indicating judgment that a
majority of well informed people will make the same choice, but a substantial minority will not, ie. different patients, in
different clinical contexts, with different values and preferences, will likely make different choices), or uncertain (indicating
that the panel made no specific recommendation for or against interventions). The strength of the recommendation
represents the degree of confidence that we felt the desirable effects outweighed the undesirable effects of an intervention;
this judgement was based on the quality of evidence (we primarily focused on evidence derived from outcomes that were
of critical importance, but were appropriate we also considered those deemed to be important but not critical, as per Table
1), with desirable effects considered to be beneficial health outcomes, low burden and low costs. Undesirable effects were
associated harms and side effects, more burdens and expense.18 Typically a strong recommendation is based on high or
moderate quality evidence on a critically important outcome. Exceptionally, a strong recommendation can be made based
on low quality evidence; panel members were given an example of when this may be appropriate (Figure 2).
To explore the range and distribution of the opinions held by every member of the working group, we implemented the
GRADE grid25 (Figure 2). This approach allows each member of the consensus panel to record their views about the
balance between the benefits and disadvantages of each component of PRICE, based on their analysis of the available
evidence. At the end of the meeting each member of the panel were provided with full restatements of the proposition,
a summary presentation of the relative evidence from the primary author (CMB), and a further review of the potential
sources of disagreement. They were then presented with a number of specific clinical scenarios relating to the 5 clinical
questions, and were asked to devise recommendations for each one. They each had two weeks to formulate his/her
judgements on the quality of the evidence, and the strength of the recommendations. Grids were returned to the primary
reviewer (CMB) before the final write up and publication.
Figure 2
GRADE grid for recording panellists views in the development of final recommendations
GRADE score
1

Balance between
desirable and
undesirable
consequences
of intervention

Desirable
clearly outweigh
undesirable

Desirable
probably
outweigh
undesirable

Trade-offs
equally
balanced or
uncertain

Undesirable
probably
outweigh
desirable

Undesirable
clearly
outweigh
desirable

Recommendation

Strong
definitely do it

Weak
probably
do it

No specific
recommendation

Weak
probably dont
do it

Strong
definitely
dont do it

For each intervention below, please mark an X in the cell that best corresponds to your assessment of the
available evidence, in terms of benefits versus disadvantages for its use in the management of acute soft
tissue injury
Protection
Rest
Ice
Compression
Elevation
Strong recommendations are more likely to be warranted with: larger differences between the desirable and undesirable
effects, less variability/uncertainty in values and preferences, lower costs and higher quality evidence.
Note: Example of a strong recommendation from low quality evidence: There is low quality evidence, but it is based
on a large number of observational studies, with clear evidence and consistency of effect on at least one critically
important outcome after acute soft tissue injury. The intervention is of low cost (in comparison to other related or
available treatment options), can be applied with little burden (to the patient and therapist), it is comfortable and easily
tolerable, has a low risk of minor side effects, and is not time consuming (for patient or therapist).

References
1. Hawkins RD, Hulse MA, Wilkinson C, Hodson A, Gibson M. The association of football medical research programme:
an audit of injuries in professional football. British Journal of Sports Medicine 2001; 35: 43-47.
2. Woods C, Hawkins R, Hulse M, Hodson A. The football association medical research programme: an audit of injuries
in professional football-analysis of preseason injuries. British Journal of Sports Medicine 2002; 36: 436-441.
3. Cromwell F, Walsh J, Gormley J. A pilot study examining injuries in elite gaelic footballers. British Journal of Sports Medicine
2000; 34: 104-108.
4. Gissane C, Jennings D, Kerr K, White JA. A pooled analysis of injury incidence in rugby league football. Sports Medicine
2002; 32: 211-216.
5. Egermann M, Brocai D, Lill CA, Schmitt H. Analysis of injuries in long distance triathletes. International Journal of Sports
Medicine 2003; 24: 271-276.
6. Gabbe B, Finch C, Wajswelner H, Bennell K. Australian football: injury profile at the community level. Journal of Science and
Medicine in Sports 2002; 5: 149-160.
7. Hackett GL, Bundred P, Hutton JL, OBrien J, Stanley IM. Management of joint and soft tissue injuries in three general
practices: value of onsite physiotherapy. British Journal of General Practice 1993; 2: 61-64.
8. Nicholl JP, Coleman P, Williams BT. The epidemiology of sports and exercise related injury in the United Kingdom. British
Journal of Sports Medicine 1995;29: 232-238.
9. Van Rijn RM, van Os AG, Bernsen RMD, Luijsterburg PA, Koes BW, Bierma-Zeinstra SMA. What is the clinical course of acute
ankle sprains? A systematic literature review. American Journal of Sports Medicine 2008;121:324-31.
10. Valderrabano V, Hintermann B, Horisberger M, Fung TS. Ligamentous posttraumatic ankle osteoarthritis. American Journal
of Sports Medicine 2006;34:612-20.
11. Baker P, Coggon D, Reading I, Barrett D, McLaren M, Cooper C. Sports injury, occupational physical activity, joint laxity, and
meniscal damage. Journal of Rheumatology 2002;29(3):557-63.
12. Issa SN, Sharma L. Epidemiology of osteoarthritis: an update. Current Rheumatology Reports. 2006;8(1):7-15.
13. Dekker J, van Dijk GM, Veenhof C. Risk factors for functional decline in osteoarthritis of the hip or knee. Current Opinion in
Rheumatology. 2009;21(5):520-4.
14. Hunter DJ, Sharma L, Skaife T. Aligment and osteoarthritis of the knee. Journal of Bone and Joint Surgery (American).
2009;91 Suppl 1:85-9.
15. Kerr KM, Daley L, Booth L for the Association of Chartered Physiotherapists in Sports Medicine (ACPSM). Guidelines for the
management of soft tissue (musculoskeletal) injury with protection, rest, ice, compression and elevation (PRICE) during the
first 72 h. London: Chartered Society of Physiotherapy, 1998.
16. Thomson R, Lavender M, Madhok R. How to ensure that guidelines are effective. British Medical Journal, 1995; 311, 237241.
17. Canadian Task Force. Canadian task force classifications. Canadian Medical Association Journal 1979; 121, 11931254.
18. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. for the GRADE Working Group. GRADE:
an emerging consensus on rating quality of evidence and strength of recommendations. British Medical Journal.
2008;336(7650):924-6.
19. Higgins JPT, Altman DG (editors). Chapter 8: Assessing risk of bias in included studies. Section 8.5. In: Higgins JPT, Green S
(editors), Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.2 [updated September 2009]. Available
from www.cochrane-handbook.org.
20. Bleakley CM, Glasgow PD, Philips P, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists
in Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and
Elevation, 2011. Chapter 2: What is the magnitude and depth of cooling associated with ice?
21. Bleakley CM, Glasgow PD, Philips P, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists
in Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and
Elevation, 2011. Chapter 3: Can PRICE decrease the inflammatory response after acute soft tissue injury?
22. Bleakley CM, Glasgow PD, Philips P, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists
in Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and
Elevation, 2011. Chapter 4: What effect does mechanical loading have on inflammation and soft tissue healing after
acute injury?
23. Bleakley CM, Glasgow PD, Philips P, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists
in Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and
Elevation, 2011. Chapter 5: Do the physiological effects of local tissue cooling affect function, sporting performance
and injury risk?
24. Bleakley CM, Glasgow PD, Philips P, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists
in Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and
Elevation, 2011. Chapter 6: Which components of PRICE are effective in the clinical management of acute soft tissue injury?
25. Jaeschke R, Guyatt GH, Dellinger P, Schnemann H, Levy MH, Kunz R, et al. for the GRADE working group. Use of
GRADE grid to reach decisions on clinical practice guidelines when consensus is elusive British Medical Journal 2008; 337: a744.

Chapter 2
What is the magnitude and depth of cooling associated with ice application?
What is known in this area: The basic premise of cold therapy is to cool injured tissue to achieve analgesia and
lower local cell metabolism. There is much confusion around how much cooling is adequate, and how this can
be achieved clinically. Current best evidence suggests that to optimise the clinical effectiveness of icing, we must
achieve a critical level of tissue cooling. These are skin temperatures of <13C for cold induced analgesia, and
tissue temperatures of 5-15C for metabolic reduction.
Aim: To update the pathophysiological rationale for using PRICE in the management of soft tissue injury.
Clinical question: What is the magnitude and depth of cooling associated with ice application?
Objective: To review recent literature to determine the rate and magnitude of tissue temperature reduction with
ice. Values were compared with current recommended threshold temperatures deemed necessary for optimal cold
induced analgesia (skin temperature <13C) and metabolic reduction (tissue temperature 5C-15C) after injury.
What this review adds: Current evidence confirms that icing dosage should be guided by the circumstances of
injury, and clinical rationale. Optimal levels of analgesia can be achieved clinically using crushed ice for durations
of 5-15 minutes. Other modes of icing may require longer durations to cool skin temperature to optimal levels,
and there is further evidence that skin temperature is a poor predictor of deep tissue temperature. Based on
healthy human models, it is difficult to induce large decreases in intra-muscular or joint temperature; particularly in
circumstances of deep tissue injury, areas of higher levels of body fat, or when dry barriers are used at the cooling
interface. Reaching currently accepted threshold temperatures for metabolic reduction (5C-15C) seem unlikely.
The lowest reported superficial muscle temperature (1cm sub-adipose) after icing was 21C, in a lean athletic
population. No study has considered temperature reductions after a closed soft tissue injury to the joint. Based on a
surgical model, untreated joints increase in temperature after surgery, however ice seems to minimise the extent of
this increase, or in some cases, cool slightly below baseline temperature (maximum decrease of 4C).
Limitations and future study recommendations: Most evidence on skin and muscle temperature reductions is
derived from healthy human subjects. Joint cooling models are derived from post surgical models of the knee and
shoulder. No studies have assessed the magnitude of or depth of temperature reductions obtainable in smaller
joints with less surrounding adipose tissue (eg. wrist, elbow, ankle), in the absence of tourniquet and post operative
dressings. Our conclusions are based on the assumption that in order to induce a clinically meaningful effect on
pain reduction or cell metabolism, critical levels of skin temperature (<12C) and tissue temperature (5C-15C),
must be achieved. Although these values currently represent the best available evidence, they are not definitive.
These values may be subject to change based on emerging understanding of the biochemical and physiological
events surrounding acute injury and inflammation.

Background
Ice is a popular intervention for soft tissue injury within the first aid, sporting and post surgical settings. Its simple premise
is to extract heat from the body tissue to attain various clinical benefits. These include: limiting the extent of injury by
decreasing tissue metabolism thereby reducing secondary cell death,1 providing analgesia,2 and facilitating rehabilitation.3
In spite of this, controversy and confusion exist within clinical practice and published literature over the therapeutic benefits,
and the most effective icing protocol. Textbooks,4 clinical guidelines, and surveys of practice5 show wide discrepancies in its
application. Many clinical research models may have employed an inadequate treatment dosage;6 and there may be an
erroneous supposition that one universal icing protocol will be equally effective, regardless of the pathological condition
or body tissue affected.
It is increasingly apparent that the clinical effectiveness of an electro-physical agent relates to the relative adequacy of
parameter choice and dosage.7 Current best evidence also suggests that to optimise the clinical effectiveness of cold
therapy, we must achieve a critical level of tissue cooling. Furthermore, the magnitude and depth of this critical cooling
level may change depending on the desired clinical effect (eg. decreasing secondary injury vs analgesia), or stage of injury
(immediate stages vs later stages of rehabilitation).3

It is common practice to apply ice to decrease pain. This is often seen in pitch side management of an acute sports
injury, whereby (in the absence of any serious structural damage), the primary objective is to reduce pain, and return the
athlete to the field of play as quickly as possible. In this situation, cold induced analgesia seems to work via a number
of mechanisms including: decreased receptor sensitivity,8,9 decreased receptor firing rate,9,10 decreased nerve conduction
velocity (NCV), reduced muscle spasm or as a counter irritant to pain.11 Evidence shows that cooling skin temperature to
between 10C and 13C results in localised analgesia,12,13 and a 10%-33% reduction in NCV.13,14 This is currently regarded
as the threshold for optimally inducing analgesia in the clinical setting.
Vant Hoffs law states that for every 10C reduction in tissue temperature, the rate of chemical reaction will decrease
between 2 and 3-fold. This forms some of the rationale for applying ice in the immediate stages after an acute injury.
Reducing cellular metabolism will decrease the risk of secondary ischaemic and enzymatic injury, thereby limiting the
overall extent of tissue damage.3,15 In this regard it is generally assumed that better clinical outcomes result from greater
and faster cooling of tissue temperature;16 however, it is often difficult to provide an exact magnitude or threshold.
The current best evidence seems limited to animal models, which advocate that metabolism is optimally reduced after
injury, at tissue temperatures between 5C and 15C.15,17,18 Furthermore, to maximise this effect, the relevant temperature
reductions must occur within the injured tissue, (eg. the muscle layer at and around the point of injury), and not simply the
over lying skin.
Objectives
The objective of Chapter 2 was to review recent literature to determine the rate and magnitude of tissue temperature
reduction associated with popular ice dosage. Results were compared and discussed in terms of the critical temperature
reductions, currently deemed necessary for achieving optimal analgesia, and metabolic decrease after injury.
Methods
Literature search
We searched the titles generated from the main search strategy outlined in Appendix Table 1-4. Relevant studies were
extracted, with exclusions made based on titles, abstracts or full text versions.
Inclusion criteria
No restrictions were made on study design. Participants could have been injured (post surgery/acute soft tissue injury) or
healthy and of any age or gender. Any type of ice/compression intervention was included, providing adequate details
of the mode and duration were provided. Outcomes must have included at least one of the following: skin temperature,
intramuscular temperature, joint temperature, measured before and after ice pack application. No restrictions were made
on the measuring device, however, for subcutaneous or muscle temperatures, the depth of measurement must have been
provided or calculable from the data.
Risk of bias
All included studies were assessed in terms of design, and methodological quality based on Cochrane risk of bias tool19
Results
Excluded studies (with reasons) are listed in Appendix Table 5.
Skin temperature
9 studies16,20-27 reported skin temperature reductions after ice based on applications directly onto the skin or through a
damp barrier; of which two21,24 compared to groups with a dry barrier or bandage between the cooling interface and
the skin. Two others compared ice applied over different types of dry, thick bandaging.28,29 Full details of the methods,
interventions and risk of bias are summarised in Appendix Table 6. Only two studies used adequate randomisation,24,29
three performed allocation concealment,22,24,29 and none of the outcomes were based on blinded assessment. Only one
of the studies29 used a group of injured subjects.

Figure 1
-Circles represent values from participants treated with
other cooling modes (frozen peas, cryocuff, water and
alcohol, gel packs)16,22,24,25,26
-Lines represent values from participants treated with
crushed ice16,20,21,23,25,26
-The arrows overlap the area showing the critical SKIN
temperature reduction of <10C.
-All studies applied either directly onto the skin, or via a
damp barrier22,23,25,26

Figure 1 summarises the mean skin temperature reductions associated with different modes of ice application over
time. Studies show that crushed ice reduces skin temperature to below 10C after 5,23 10,16,20 15,16,23 or 20 minutes of
application.21,25,26 In contrast, 13 sub-groups of participants, from six different studies16,22,24-27 using other modes of cooling
(frozen peas, gel pack, cryo-cuff, and cold water immersion) produced skin temperature reductions that were more likely
to be outside the defined critical range of <13C, after 5-20 minutes of application. One study30 also found evidence
that lower skin temperatures are achieved with heavier crushed ice packs; a 0.8kg and a 0.3kg pack resulted in final skin
temperatures of 5.1C and 7.7C respectively.
Numerous cooling modalities remain popular amongst clinicians; however it is clear that these should not be used
interchangeably. Each cooling modality has a different thermal property, and therefore a different cooling potential.
Although gel packs tend to have very low pre-application temperatures, (sometimes between -10C and -14C), solid
crushed ice at 0C has much more potential to extract heat energy from the skin based on its ability to undergo phase
change.16 The consensus from the current literature is that crushed ice consistently results in the fastest reduction in skin
temperature with, clinically important reductions (<10C) after as little as 5 minutes.
There is a trend in Figure 1 that for the majority of cooling modes, the fastest reduction in skin temperature occurs within
the first 5 minutes. More modest reductions seem to occur thereafter, and in many instances, skin temperature seemed to
plateau after approximately 10-15 minutes of cold application. Continuing clinical applications beyond this time period,
may offer little further reduction in skin temperature. This is important to consider if our goal is to simply induce short term
analgesia. However, removal of the ice pack after just 10 or 15 minutes might be associated with a faster re-warming rate
and therefore may not suitable in instances when pain reduction is required over a longer period of time.
Muscle temperature
4 studies31-34 were excluded as we were unable to determine the depth at which temperature was recorded. This left 11
studies16,35-44 measuring intra-muscular temperature reductions after ice application. Methods, intervention and risk of bias
details are summarised in Appendix Table 7. All used healthy subjects, and the majority focused on 1cm muscle depths. In
most cases, accuracy was controlled by inserting thermocouple needles to a depth of subcutaneous fat (over the treated
area) + 1 cm. Four studies recorded muscle temperatures down to approximately 2-3 cm16,38,42-44 by similar means.
Four of the studies used a randomised design;35-37,39 the remainder were observational. No study used blinded outcome
assessment. Sample size was small ranging from 6-47.

Figure 2
-Circles represent values from participants with 21-40mm skin-fold at the point of ice application
-Lines represent values from participants with <20 mm skin-fold at the point of ice application
-Arrows overlap the area representing potentially optimal INTRA-MUSCULAR temperature reduction of 5-15C.
-All data is based on crushed ice, applied directly onto the skin surface 16,35-44
-Three studies applied additional external compression16,41,42

Figure 3
Circles represent < 10 mm skin fold at the point of ice application38
Lines represent 11-20mm skin fold at the point of ice application16,38,43
Squares represent 21-30 mm skin fold at the point of ice application38,42,44

Joint temperature
Eight studies measured joint temperature reductions after either knee or shoulder surgery.45-52 In all cases, post operative
dressings and/or bandages were used between the cooling medium and the skin. Methods, intervention and risk of bias
details are summarised in Appendix Table 8. Sample size ranged from 12 to 30. All but one study51 used a control group.
Four 45-47,52 used a randomised controlled design, only one of which did not use adequate allocation concealment.47
Figure 4 shows that in each study, surgery increased joint temperatures in untreated controls. However, ice either reduced
the magnitude of this increase, or decreased joint temperature post surgery. Osbahrs results52 are not included in Figure 4
as there was no data on the post operative temperatures; values extracted from graphs show that ice caused intra-articular
temperatures to remain around 1C lower than untreated controls for the first day post operatively. Of note this was the
only study to use randomisation, allocation concealment and blinded outcome assessment.

Figure 4

Chapter 3
Can PRICE decrease the inflammatory response after acute soft tissue injury?
What is known in this area: PRICE is one of the simplest and oldest approaches for treating acute soft tissue
injuries. The rationale is that components of PRICE have an anti inflammatory effect after injury, but few clinicians
may look beyond the cardinal signs when providing justification for intervention. There have been a number of
recent advances in understanding the physiological and biochemical events associated with acute inflammation
after soft tissue injury.
Aim: To update the pathophysiological rationale for using PRICE in the management of soft tissue injury.
Clinical question: Can PRICE decrease the inflammatory response after acute soft tissue injury?
Objective: To review the rationale for PRICE intervention in the acute phases of soft tissue injury based on
physiological, cellular and molecular models of inflammation.
What this review adds: There is limited evidence from human studies that ice may reduce metabolism or aspects of
the inflammatory response based on tissue perfusion and microdialysis. Other human models based on EIMD had
a high risk of bias, and there were conflicting results; furthermore few have focused on biomarkers of inflammation
per se. There is evidence from animal models that ice seems to have a consistent effect on key cellular and
physiological events associated with inflammation after injury. This includes cell metabolism, white blood cell activity
within the vasculature, and potentially apoptosis. The relative benefits of these effects have yet to be fully elucidated
and it is difficult to contextualize within a human model. The effects on micro-circulation after injury are contrasting,
potentially due to heterogeneity of injury, intervention and outcome.
Limitations and future study recommendations: The evidence from human studies is of low methodological
quality and the inflammatory models used may have limited application to closed soft tissue injury. The remainder of
the evidence is derived solely from animal models, some of which is based on excised tissues samples. The models
are also limited to lab induced muscle injuries, and in many cases anaesthesia may have confounded the outcomes.
Future research should use human subjects and focus on more direct examination of inflammatory related markers,
and cellular oxidative stress. Considerations must be given to the type of soft tissue affected (eg. muscle vs ligament)
and the injuring force (crush vs tensile).

Background
PRICE is one of the simplest and oldest approaches for treating soft tissue injuries such as sprains, contusions, and
dislocations. The immediate phase after soft tissue injury is characterised by an acute inflammatory response. This often
presents clinically with cardinal signs such as heat, redness, pain and swelling. Currently few clinicians may look beyond
the cardinal signs when providing justification for intervention; and it is commonly accepted that components of PRICE,
particularly ice and compression have an anti inflammatory effect after soft tissue injury. Applying a cold compress to
hot, red and swollen tissue may seem pragmatic; however there is not always an obvious link between inflammation
visible under the microscope and that clinically apparent and characterised by the original cardinal signs.1 Paradoxically,
recent trends in sports medicine involve delivering growth factors into healing muscle tissue (eg. via platelet rich plasma or
autologous blood injections)2 which seems to lean more towards a pro-inflammatory treatment approach.
Objectives
The objective of this chapter was to review the rationale for PRICE intervention in the acute phases of soft tissue injury
based on physiological, cellular and molecular models of inflammation.
Methods
Literature search
We searched the titles generated from the main search strategy outlined in Appendix Tables 1-4. Relevant studies were
extracted, with exclusions made based on titles, abstracts or full text versions.

Inclusion criteria
No restrictions were made on study design; and both animal and human subjects with acute soft tissue injury were
considered. Acute injuries induced in a laboratory situation were included. Studies using exercise induced muscle damage
(EIMD) were included provided that it was induced with a well defined resistance (concentric or eccentric) or plyometric
exercise protocol. No restrictions were placed on the type of PRICE intervention. Outcomes could include any physiological,
cellular or molecular measurement associated with inflammation; recorded before injury, and up to one week post injury.
We were particularly interested in the following data: injury site, type and severity of injury, intervention technique and
dosage, and influence of anaesthesia (in the case of induced injury). If applicable, outcomes were extracted pre-injury and
post-injury. Qualitative comparisons were made and results were grouped and discussed by outcome.
Based on the difficulties associated with measuring haemodynamics in acutely injured subjects, we felt it was appropriate to
also consider evidence on the effect of cooling/compression on local blood flow in human participants, with no restrictions
made on injury type, or depth of measurement.
Risk of bias
All included studies were assessed in terms of design, and methodological quality based on Cochrane risk of bias tool3
(sequence generation, allocation concealment, assessor blinding, incomplete outcome data).
Results
Excluded studies (with reasons) are listed in Appendix Table 5.
Human
Overview of study methods
We found 20 relevant human models based on EIMD;4-23 Figure 1 provides a summary of risk of bias across studies. All
used randomisation, half of which were cross over designs. There were a number of methodological limitations; only one14
adequately described the methods of sequence generation and allocation concealment, and two7,14 used blinded outcome
assessors. In the majority of cases, comparisons groups were untreated controls; one was placebo ultrasound.8
Interventions focused on cooling,4,8,10,11,13-18 compression6,7,19-23 or protection/rest.5,9,12. Cold intervention largely involved
CWI, with two studies8,11 using 15 minutes of ice massage. CWI times ranged from 10-15 min, except Sellwood et al.14
who used a 1 minute CWI repeated 3 times. Water temperature ranged from 15C10 to 5C.14 Compression interventions
were clothing garments (tights or sleeves), put on immediately after EIMD and worn continuously for up to 4 days. Their
associated compressive forces were reported between 10 and 30mmHg. One study used a whole body compression suit.23
The remaining relevant studies undertaken on humans used post surgical24 or inflammatory arthritic models.25 Figures 2-5
summarise the key characteristics of relevant studies using injured human models.
PRICE and inflammation post EIMD
The EIMD studies mostly considered biomarkers of muscle damage after exercise, and all undertook daily recording of
serum Creatine Kinase (CK), for up to 5 days after exercise. 5 studies,11,14,15,17,18 found CK levels were similar across cooling
and control (no cooling) groups. In contrast others reported significantly lower CK levels within cold treated groups, at
24,13,16 48,4,13 724,8,16 and 96 hours10 post exercise. In two studies, compression sleeves were associated with lower levels of
CK at 727 and 120 hours;6 and a full body compression garment produced lower levels at 24.23 Graduated compression
tights19-22 did not affect CK levels after lower limb EIMD. In two cases5,9 rigid immobilisation significantly reduced CK levels
between 2 and 5 days after EIMD to the upper limb; whereas partial immobilisation in a sling had little effect.12
Others also recorded biomarkers of inflammation, these were: myoglobin,9,11,16,19 interleukin-6 (IL-6),16 C-Reactive Protein
(CRP)21 and lactate dehydrogenase (LDH);6,10,16,20,23 there were no significant differences between treatment groups and
controls for any of these outcomes.
PRICE and inflammation post surgery / arthritis
Using a case control design, Stalman et al24 used microdialysis to determine local metabolic and inflammatory responses
in knee synovium after ACL reconstruction, or minor arthroscopic surgery. The ACL group only was treated with intraarticular morphine and ice and compression post surgery; they had significantly lower levels of PGE2, and lower levels of
synovial lactate levels compared to the arthroscopic group. This provides promising evidence that the magnitude of the
metabolic and inflammatory responses can be reduced after surgery, however we cannot yet ascertain whether this relates
to the effects of ice and compression, morphine or both.

Animal models
Overview of study methods
There were 16 animal studies26-41 all of which used randomised controlled or controlled methods. Six26,28, 29,30,39,40 included
a blinded assessment of outcome. In all cases animals were subjected to a standardised injury, and then divided into
either a treatment group, or an untreated control group. Two studies injected cytokine (hrTNF )34 or formalin35 to induce
inflammation, one used EIMD41 and all others used blunt trauma to induce a muscle contusion. In all but one study,41
cooling was the primary intervention. Outcomes focused on secondary cell death, white blood cell behaviour, apoptosis,
blood flow and oedema formation. These were measured largely using immumohistological analysis of excised tissue,
and/or intravital microscopy (with and without laser Doppler imaging). Details of relevant animal models are summarised
in Figure 6.
In most of the animal models, cooling was initiated less than 15 minutes after injury. The duration of the interventions
ranged from 20 minutes up to 6 hours of continuous cooling. In two cases cooling was undertaken directly on exposed
animal muscle,38,39 with others applying over intact skin (both shaven and unshaven). Three studies26,28,33 used concomitant
compression when cooling and one study used cyclic compression alone.41 In all but three studies,29,30 the animals were
under anaesthesia for the duration of the intervention.
Figure 7 provides an overview of all studies assessing the effect of PRICE on inflammation by: model, injury type and
cooling dose
Secondary cell injury
Perhaps the most commonly cited rationale for applying ice after acute soft tissue injury relates to the secondary injury
model.42 This is based on the premise that after an initial trauma (e.g. muscle strain or contusion), the patho-physiological
events associated with acute inflammation can induce secondary damage to cells around the injury site. Of particular
concern is that this can involve collateral damage to healthy cells not injured in the initial trauma. This phenomenon is
known as secondary cell injury, and may be caused by both enzymatic and ischaemic mechanisms.43 One of the most
important cellular effects associated with icing is its potential to reduce the metabolic rate of tissues at, and surrounding
the injury site. This reduction in metabolic demand may allow the cells to better tolerate the ischaemic environment in the
immediate phases after injury, and thus minimising the potential for secondary cell injury or death.
Evidence to support secondary injury theory is based largely on studies of limb preservation. Osterman et al.44 and Sapega
et al.45 both used, phosphorous 31 nuclear magnetic resonance imaging to monitor cellular metabolism in ischaemic
(amputated) cat limbs, stored at a range of temperatures between 22C to 1C. Overall, they found that cells survived better
at lower muscle temperatures. This was exemplified by lower levels of ATP and PC depletion, and lower levels of acidosis,
during the period of ischaemia. Of note, these effects appeared to be reversed at more extreme muscle temperatures
reductions below 5C. This was attributed to extreme temperature reductions causing inhibition of the calcium pump of the
muscles sarcoplasmic reticulum.45
The current search found one related study in this area; Merrick and colleagues33 tried to quantify the effect of icing on
mitochondrial function after injury. Specifically they measured the activity of the mitochondrial enzyme, cytochrome c
oxidase, after experimental crush injury; comparing outcomes in cold treated and untreated muscle tissue. Fitting with
the secondary injury model, five hours of continuous cooling inhibited the loss of mitochondrial oxidative function after
injury when compared to the untreated controls. Although the model used by Merrick et al.33 does not directly determine
the effects of ice on the inflammatory process or muscle injury per se, it is the first study to have taken a novel approach to
indirectly assess the effects of secondary generated free radicals, and their possible interference with enzymes controlling
oxidative phosphorylation (cytochrome c oxidase) and thus ATP production after injury.
White Blood Cells
When muscle or joint injury occurs, phagocytic white cells, such as neutrophils, monocytes, eosinophils and macrophages
become activated and dominate the inflammatory response in the early stages. Although these cells have a critical role in
healing through their removal of necrotic debris and release of cytokines;46 they can also have a negative effect on soft tissue
healing after injury.46,47 For example, white cell activation results in a series of reactions termed the respiratory burst.48
These reactions are a source of reactive oxygen species (ROS) such as superoxide (O2.-), hydrogen peroxide (H2O2)
and hydroxyl (OH.); and hypochlorous acid (HOC1) which is a powerful antibacterial agent. In certain circumstances the
production of ROS and antibacterial agents are important immune defense mechanisms, however they can also be a
potentially dangerous mechanism if inappropriately activated. For example, overproduction of ROS may cause unwanted
collateral damage to adjacent tissues and surrounding molecules.49 This may be particularly likely in the event of a closed
soft tissue injury such as an ankle sprain, which is not associated with bacteria or infection. Indeed, there is evidence that
blocking the respiratory burst, using anti-CD11b antibody (M1/70), produces a three-fold reduction in myofibre damage
in an animal model at 24 h post-injury.50

We found three animal models; studying the effect that ice has on WBC behaviour after soft tissue injury. A popular
approach has been to use fluorescent intravital microscopy34,38-40 to observe the effect that ice has on leukocyte activity
within the microvasculature. These studies found a clear trend that icing significantly lowered the percentage of both
adherent and rolling neutrophils after injury, in comparison to injured untreated tissue. This finding was consistent over the
first 24 hours after injury.34,38-40 Cyclic compression initiated 30 minutes after EIMD in rat a model, attenuated leukocyte
infiltration.41
Other animal models26,28,35,39,40 undertook histological analysis on excised tissue after soft tissue injury. In each case,
various staining techniques were used to identify leukocyte sub-types at the injury site. Again each study made comparisons
between ice treated, and untreated injured tissue samples. Using an injured ligament model and assessor blinding, Farry
et al.26 found that ice treated groups had lower levels of WBCs (polymorphs, lymphocytes and plasma cells) at 48 hours,
in comparison to injured contra-lateral untreated limbs. Hurme et al.,28 who also used blinded outcome analysis; found
that at various time points post injury, the ice treated animal tissue had lower levels of erythrocytes (1 hour), neutrophils (6
hours) and macrophages (at 24 hours) in comparison to the untreated control limbs. Although Schaser et al.39 also found
cooling decreased neutrophilic granulocyte muscle infiltration, in comparison to control muscle, there were higher levels
of macrophages. In a follow up study40 using longer periods of cooling (5 hours), tissue analysis at 24 hours post trauma,
also found lower levels of neutrophilic granulocytes in the cold treated muscle.
In a related model, Kenjo et al.35 measured Fos protein expression in neural tissue, after formalin induced inflammation
in rat hind paws. Quantification of Fos labelled cells is thought to be a valuable marker of neuronal response to noxious
stimuli. Cold water immersion (CWI) of the paw immediately after formalin injection significantly reduced the number of
Fos labelled cells; in addition, the peak time for their expression was delayed when compared to an untreated control.
This provides further evidence that cooling influences the bodys response to inflammatory pain, and potentially highlights
a delay in the inflammatory reaction.
Although the examination of adherent and rolling neutrophils following injury and ice may, in some instances be beneficial,
these models must be developed if we are to further our understanding in this area. It may be more relevant for future
research to quantify the amount of direct neutrophil activation that occurs following injury. This approach may allow for
estimation as to how much secondary cell and surrounding tissue damage and inflammation will likely occur. A popular
marker that is commonly used to determine neutrophil activation is myeloperoxidase. This is produced by an increase in
ROS activity and it has been successfully used in studies looking at free radical production and immune response after
stretch injury in animal skeletal muscle.51
Apoptosis
Apoptosis is a programmed cell death. It is characterized by a cascade of biochemical events cumulating in altered cell
morphology and eventual cell death. Although apoptosis is the normal means by which cells die at the end of their life
span, its incidence may be affected by soft tissue injury. Higher numbers of apoptotic cells have been recorded around
the edges of rotator cuff tears when compared to un-injured control muscles.52 The reasons for this increase have not
yet been fully elucidated; however, the accumulation of reactive oxygen species in injured tissues (oxidative stress) could
again play a significant role.53 Cell survival requires multiple factors, including appropriate proportions of molecular
oxygen and various antioxidants. Although most oxidative insults can be overcome by the cells natural defenses, sustained
perturbation of this balance may result in apoptotic cell death.
There is limited evidence from animal models that ice can reduce the incidence of apoptosis after injury. Westermann34
found that after chemically induced inflammation, the number of apoptotic muscle cells (quantified by the number of cells
with nuclear condensation and fragmentation) was significantly higher in untreated controls, when compared to the ice
group. This is an interesting finding as reduced levels of apoptosis may again represent a protective effect of ice after soft
tissue injury. We can only postulate as to the reasons for this finding; however this may be further evidence that ice can
reduce inflammation and decrease secondary free radical production (from the respiratory burst), thereby causing less
interference with important proteins and other cell metabolites that control apoptosis.
Blood flow and Oedema
Acute soft tissue injury incurs a multitude of changes to local vasculature and microvasculature. These include: increased
vessel diameter;29,40 increased cell permeability and macromolecular leakage into the injured tissue;40 and decreased
tissue perfusion.34,39,40 One of the proposed mechanisms for cooling in the immediate stages after injury is to reverse these
effects, and it is well accepted that cooling has a vasoconstrictive effect on the vasculature. The original PRICE guidelines
found consistent evidence that various modes of cooling decreased blood flow at the ankle joint and knee joint, based on
impedance plethysmography or triple phase technetium bone scanning.54-57

Healthy human
Many recent studies have used laser Doppler fluxmetry to record microcirculation within the extremities (eg. fingers, hands).
There is consistent evidence that cooling almost immediately decreases superficial blood flow and tissue perfusion.58-63
This response is thought to relate to reflex sympathetic activity which increases the affinity of alpha adrenoceptors in the
vascular walls for norepinephrine, causing vasoconstriction.64 Interestingly, all cases also noted a paradoxical increase
in blood flow, after around 3-8 minutes of cooling. This pattern is consistent with Lewis original theory65 that prolonged
exposure of cooling leads to a secondary vasodilator effect. The precise mechanism for this is not known, and it may
relate to either central or peripheral processes. Lewis65 suggested an axon reflex whereby the decreasing local tissue
temperatures, interrupts sympathetic nerve conduction, with resultant vasodilatation.
It is important to consider that these patterns are based on isolated immersion of the extremities (e.g. single finger),
with outcomes limited to blood flow at depths of around 0.6 mm. This represents skin or superficial tissue blood flow,
which is generally not the central site of injury. Other recent studies have focused on the effect of ice on healthy tendon
haemodynamics at depths of 2-8mm. Using real time laser doppler spectrophotometry, Knobloch and colleagues66-69 have
found consistent evidence that various combinations of ice, and compression induce an immediate and significant reduction
in capillary blood flow and oxygen saturation, with facilitated venous capillary outflow. Using a similar spectrometry
technique, Yanagisawa70 found a cold induced decrease in haemoglobin/myoglobin concentration within muscle tissue,
again suggesting a decrease in local circulation. Of further interest was that none of these studies66-70 found evidence
of reactive vasodilatation during cooling. Similar patterns have been observed based on impedance or strain gauge
plethysmography57,71 with cold induced decreases in local blood flow in the ankle joint, and calf muscles, with no reports
of reactive vasodilatation. Only one study72 concluded that cooling did not result in any change to local tissue blood flow.
The reported outcome was based on strain gauge (venous occlusion) plethysmography, and cooling was a 20 minute CWI
in 13C. In this regard, cooling was undertaken in a gravity dependant position which may have had a concomitant effect
on blood flow.
Few have considered the effect of ice on the vasculature after significant soft tissue injury in humans; and models are
restricted to EIMD. Based on MRI, Yanagaisawa and colleagues73 found that short bouts of cooling prevented the usual
accumulation of post exercise oedema in ankle dorsiflexor muscles.
Microcirculation (injured animal)
A number of animal studies have considered the effect of cooling on microcirculation after acute crushing injury, using
intravital microscopy techniques. The results were often inconsistent: some studies found that ice application did not
significantly change capillary diameter,39,40 arteriole diameter;29,31,39 or capillary velocity after injury.39,40 In contrast, others
found that ice either significantly increased39 or decreased34 arteriole diameter after injury.
There may be clearer patterns associated with venular diameters. Three studies38-40 reported smaller venular diameters in
ice treated groups in comparison to injured (untreated) controls when measured at both the initial stages38,39 and at 24
hours40 post injury. In two of these studies,38,40 the differences were significant, and in one case, venular diameter in the
cold group had returned to pre-injury levels.40 Although this trend is supported with evidence that, iced tissue also had
higher levels of venular blood flow velocity in comparison to controls in the immediate stages post injury;34,38,39 this trend
was reversed at 24 hours post injury.
There is conflicting evidence on the effect that ice has on tissue perfusion post injury. Using fluorescent microscopic
assessment of the functional capillary density (length of erythrocyte-perfused capillaries per observation area), three
studies; 34,39,40 found that ice application significantly increases tissue perfusion after injury, in comparison to untreated
injured controls. Again, in two cases,39,40 perfusion was restored to pre-injury levels. In contrast, based on laser Doppler
imaging after injury, Curl31 found that cooling had little effect on microvascular perfusion.
Using a related outcome measurement, Schaser and colleagues39,40 monitored intramuscular pressures in rat limbs after
soft tissue injury, randomising limbs to receive either cold saline, or no intervention. Lower intramuscular pressures [17.7
mmHg (SD: 4.7)] were recorded at 1.5 hours post injury in the cooling group (treated with 20 minutes of saline cooling),
when compared to untreated controls [19.2 mm/Hg (SD: 3.1)]. Their follow up study, also found that longer periods
muscle cooling (5 hours) was associated with lower intramuscular pressures 18 (95% CI: 5.5 mmHg) in comparison to
untreated controls [26 (95% CI: 1.9 mmHg)], at 24 hours post injury.

References
1. Scott A, Khan KM, Roberts CR, Cook JL, Duronio V. What do we mean by the term inflammation? A contemporary basic science
update for sports medicine. British Journal of Sports Medicine 2004;38(3):372-80.
2. Creaney L, Hamilton B. Growth factor delivery methods in the management of sports injuries: the state of play. British Journal of
Sports Medicine 2008;42(5):314-20.
3. Higgins JPT, Altman DG (editors). Chapter 8: Assessing risk of bias in included studies. Section 8.5. In: Higgins JPT, Green S
(editors), Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.2 [updated September 2009]. Available from
www.cochrane-handbook.org.
4. Eston R, Peters D. Effects of cold water immersion on the symptoms of exercise-induced muscle damage. Journal of Sports Sciences
1999;17:231-238.
5. Sayers SP, Clarkson PM, Lee J. Activity and immobilisation after eccentric exercise: II. Serum CK. Medicine and Science in Sports and
Exercise. 2000 Sep;32(9):1593-7
6. Kraemer WJ, Bush JA, Wickham RB, Denegar CR, Gmez AL, Gotshalk LA, Duncan ND, Volek JS, Putukian M, Sebastianelli WJ.
Influence of compression therapy on symptoms following soft tissue injury from maximal eccentric exercise. Journal of Orthopaedic
and Sports Physical Therapy. 2001a Jun;31(6):282-90.
7. Kraemer WJ, Bush JA, Wickham RB, Denegar CR, Gmez AL, Gotshalk LA et al. Continuous compression as an effective therapeutic
intervention in treating eccentric-exercise-induced muscle soreness. Journal of Sports Rehabilitation. 2001b; 10: 11-23
8. Howatson G, Van Someren KA. Ice massage: effects on exercise-induced muscle damage. Journal of Sports Medicine and Physical
Fitness. 2003 Dec;43(4):500-5.
9. Sayers SP, Clarkson PM. Short-term immobilisation after eccentric exercise. Part II: creatine kinase and myoglobin. Medicine and
Science in Sports and Exercise 2003 May;35(5):762-8.
10. Yanagisawa O, Niitsu M, Yoshioka H, Goto K, Kudo H, Itai Y. The use of magnetic resonance imaging to evaluate the effects of
cooling on skeletal muscle after strenuous exercise. European Journal of Applied Physiology 2003;89:53-62
11. Howatson G, Gaze D, van Someren KA. The efficacy of ice massage in the treatment of exercise induced muscle damage.
Scandinavian Journal of Medicine and Science in Sports. 2005 Dec;15(6):416-22.
12. Zainuddin Z, Hope P, Newton M, Sacco P, Nosaka K. Effects of partial immobilisation after eccentric exercise damage on recovery
from muscle damage. Journal of Athletic Training. 2005 Jul-Sep;40(3):197-202.
13. Skurvydas A, Sipaviciene S, Krutulyte G, Gailiuniene A, Stasiulis A, Mamkus G, Stanislovaitis A.. Cooling leg muscles affects
dynamics of indirect indicators of skeletal muscle damage. Journal of Back and Musculoskeletal Rehabilitation 2006;19:141-151.
14. Sellwood KL, Brukner P, Williams D, Nicol A, Hinman R. Ice water immersion and delayed-onset muscle soreness: a randomised
controlled trial. British Journal of Sports Medicine 2007;41:392-397.
15. Goodall S, Howatson G. The effects of multiple cold water immersions on indices of muscle damage. Journal of Sports Science and
Medicine 2008;7:235-241.
16. Vaile J, Halson S, Gill N, Dawson B. Effect of hydrotherapy on the signs and symptoms of delayed onset muscle soreness. European
Journal of Applied Physiology. 2008 Mar;102(4):447-55.
17. Howatson G, Goodall S, van Someren KA. The influence of cold water immersions on adaptation following a single bout of
damaging exercise. European Journal of Applied Physiology. 2009 Mar;105(4):615-21. Epub 2008 Nov 26.
18. Jakeman JR, Macrae R, Eston R. A single 10-min bout of cold water immersion therapy after strenuous plyometric exercise has no
beneficial effect on recovery from the symptoms of exercise induced muscle damage. Ergonomics 2009;52:456-460.
19. French DN, Thompson KG, Garland SW, Barnes CA, Portas MD, Hood PE, Wilkes G. The effects of contrast bathing and compression
therapy on muscular performance. Medicine and Science in Sports and Exercise. 2008 Jul;40(7):1297-306.
20. Davies V, Thompson KG, Cooper SM. The effect of compression garments on recovery. Journal of Strength and Conditioning
Research. 2009 Sep;23(6):1786-94.
21. Duffield R, Cannon J, King M. The effects of compression garments on recovery of muscle performance following high intensity
sprint and plyometric exercise. Journal of Science and Medicine in Sport. 2010 Jan;13(1):136-40. Epub 2009 Jan 7.
22. Jakeman JR, Byrne C, Eston RG. Lower limb compression garment improves recovery from exercise-induced muscle damage in
young, active females. European Journal of Applied Physiology. 2010 Aug;109(6):1137-44. Epub 2010 Apr 8.
23. Kraemer WJ, Flanagan SD, Comstock BA, Fragala MS, Earp JE, Dunn-Lewis C, Ho JY, Thomas GA, Solomon-Hill G, Penwell ZR,
Powell MD, Wolf MR, Volek JS, Denegar CR, Maresh CM. Effects of a whole body compression garment on markers of recovery after
a heavy resistance workout in men and women. Journal of Strength and Conditioning Research. 2010 Mar;24(3):804-14.
24. Stlman A, Tsai JA, Wredmark T, Dungner E, Arner P, Fellnder-Tsai L. Local inflammatory and metabolic response in the knee
synovium after arthroscopy or arthroscopic anterior cruciate ligament reconstruction. Arthroscopy. 2008;24(5):579-84. Epub
2008 Feb 20.
25. Strunk J, Strube K, Klingenberger P, Mller-Ladner U, Lange U. Two and three dimensional Doppler sonographic evaluation of the
effect of local cryotherapy on synovial perfusion in wrist arthritis. Rheumatology (Oxford). 2006;45(5):637-40. Epub 2006 Mar 14.
26. Farry, P J. Prentice, N G. Hunter, A C. Wakelin, C A. Ice treatment of injured ligaments: an experimental model. New Zealand
Medical Journal 1980;91(651):12-4.
27. McMaster WC, Liddle S. Cryotherapy influence on posttraumatic limb edema. Clinical Orthopaedic and related Research
1980;150: 283-7.

28. Hurme T, Rantanen J, Kalimo H. Effects of early cryotherapy in experimental skeletal muscle injury. Scandinavian Journal of Medicine
and Science in Sports 1993;3:46-51.
29. Smith TL, Curl WW, Paterson Smith B, Holden MB, Wise T, Marr A, et al. New skeletal muscle model for the longitudinal study of
alterations in microcirculation following contusion and cryotherapy. Microsurgery 1993;14: 487-93.
30. Smith TL, Curl WW, George C, Rosencrance E. Effect of contusion and cryotherapy on microvascular perfusion in rat dorsal skeletal
muscle. Pathophysiology 1994;1: 229-233.
31. Curl WW, Smith BP, Marr A, Rosencrance E, Holden M, Smith TL. The effect of contusion and cryotherapy on skeletal muscle
microcirculation. Journal of Sports Medicine and Physical Fitness 1997;37(4):279-86.
32. Dolan MG, Thornton RM, Fish DR, Mendel FC. Effects of cold water immersion on edema formation after blunt injury to the hind
limbs of rats. Journal of Athletic Training 1997;32(3): 234-237.
33. Merrick MA, Rankin JM, Andres FA, Hinman CL. A preliminary examination of cryotherapy and secondary injury in skeletal muscle.
Medicine and Science in Sports and Exercise 1999;31:1516-21.
34. Westermann S, Vollmar B, Thorlacius H, Menger MD. Surface cooling inhibits tumor necrosis factor--induced microvascular
perfusion failure, leukocyte adhesion, and apoptosis in the striated muscle. Surgery 1999;126:881-9.
35. Deal DN, Tipton J, Rosencrance E, Curl WW, Smith TL. Ice reduces edema. A study of microvascular permeability in rats. Journal of
Bone and Joint Surgery (American) 2002;84-A(9):1573-8.
36. Kenjo T, Kikuchi S, Konno S. Cooling decreases fos-immunoreactivity in the rat after formalin injection. Clinics in Orthopaedic and
Related Research. 2002;(394):271-7.
37. Dolan MG; Mychaskiw AM; Mattacola CG; Mendel FC. Effect of cool-water immersion and high voltage electric stimulation for 3
continuous hours on acute edema in rats. Journal of Athletic Training 2003;38(4):325-329.
38. Lee H, Natsui H, Akimoto T, Yanagi K, Ohshima N, Kono I. Effects of cryotherapy after contusion using real-time intravital
microscopy. Medicine and Science in Sports and Exercise 2005;37(7):1093-8.
39. Schaser KD, Stover JF, Melcher I, Lauffer A, Haas NP, Bail HJ, et al. Local cooling restores microcirculatory hemodynamics after
closed soft-tissue trauma in rats. Journal of Trauma 2006;61:642-649.
40. Schaser KD, Disch AC, Stover JF, Lauffer A, Bail HJ, Mittlmeier T. Prolonged superficial local cryotherapy attenuates microcirculatory
impairment, regional inflammation, and muscle necrosis after closed soft tissue injury in rats. American Journal of Sports Medicine
2007;35(1):93-102.
41. Butterfield TA, Zhao Y, Agarwal S, Haq F, Best TM. Cyclic compressive loading facilitates recovery after eccentric exercise. Medicine
and Science in Sports and Exercise 2008 Jul;40(7):1289-96.
42. Knight KL. Cryotherapy in sports injury management. Champaign, IL: Human Kinetics;1995:3-98.
43. Merrick MA. Secondary injury after musculoskeletal trauma: A Review and Update. Journal of Athletic Training 2002;37(2):209-217
44. Osterman AL, Heppenstall RB, Sapega AA, Katz M, Chance B, Sokolow D. Muscle ischaemia and hypothermia: a bioenergetic study
using 31Phosphorous Nuclear Magnetic Resonance Spectroscopy. Journal of Trauma 1984;24(9):811-817.
45. Sapega AA, Heppenstall RB, Sokolow DP, Graham TJ, Maris JM, Ghosh AK, et al. The bioenergetics of preservation of limbs before
replantation. The rationale for intermediate hypothermia. Journal of Bone and Joint Surgery (American) 1988;70(10):1500-13.
46. Tidball JG. Inflammatory processes in muscle injury and repair. American Journal of Physiology. Regulatory, Integrative and
Comparative Physiology 2005;288(2):R345-53.
47. Toumi H, Best TM. The inflammatory response: friend or foe for muscle injury? British Journal of Sports Medicine 2003;37(4):284-6.
48. Blake DR, Allen RE, Lunec J. Free radicals in biological systems a review orientated to inflammatory processes. British Medical
Bulletin 1987;43(2):371-85.
49. Martinez-Cayuela M. Oxygen free radicals and human disease. Biochimie 1995;77(3):147-61.
50. Brickson S, Ji LL, Schell K, Olabisi R, St Pierre Schneider B, Best TM. M1/70 attenuates blood-borne neutrophil oxidants, activation
and myofiber damage following stretch injury. Journal of Applied Physiology 2003;95(3):969-76.
51. Brickson S, Hollander J, Corr DT, Ji LL, Best TM. Oxidant production and immune response after stretch injury in skeletal muscle.
Medicine and Science in Sports and Exercise 2001;339(12):2010-5.
52. Yuan J, Murrell GA, Wei AQ, Wang MX. Apoptosis in rotator cuff tendonopathy. Journal of Orthopaedic Research 2002;20(6):
1372-1379.
53. Yuan J, Murrell GA, Trickett A, Wang MX. Involvement of cytochrome c release and caspase-3 activation in the oxidative
stress-induced apoptosis in human tendon fibroblasts. Biochim Biophys Acta 2003;1641(1):35-41.
54. Ho SS, Coel MN, Kagawa R, Richardson AB. The effects of ice on blood flow and bone metabolism in knees. American Journal of
Sports Medicine. 1994;22(4):537-40.
55. Ho SS, Illgen RL, Meyer RW, Torok PJ, Cooper MD, Reider B. Comparison of various icing times in decreasing bone metabolism and
blood flow in the knee. American Journal of Sports Medicine. 1995;23(1):74-6.
56. Taber C, Contryman K, Fahrenbruch J, LaCount K, Cornwall MW. Measurement of reactive vasodilation during cold pack application
to nontraumatised ankles. Physical Therapy. 1992;72(4):294-9.
57. Weston M. Taber C. Casagranda L. Cornwall M. Changes in local blood volume during cold gel pack application to traumatized
ankles. Journal of Orthopaedic and Sports Physical Therapy 1994;19(4):197-199.
58. Sendowski I, Savourey G, Besnard Y, Bittel J. Cold induced vasodilatation and cardiovascular responses in humans during cold
water immersion of various upper limb areas. European Journal of Applied Physiol and Occupational Physiology. 1997;75(6):471-7.

59. Van den Brande P, De Coninck A, Lievens P. Skin microcirculation responses to severe local cooling. International Journal of
Microcirculation and Clin Exp. 1997;17(2):55-60.
60. Bergersen TK, Hisdal J, Walle L. Perfusion of the human finger during cold-induced vasodilatation. American Journal of Physiology.
1999;276(3 Pt 2):R731-7.
61. Daanen HA, Ducharme MB. Finger cold-induced vasodilation during mild hypothermia, hyperthermia and at thermoneutrality.
Aviation, Space and Environmental Medicine. 1999;70(12):1206-10.
62. Sendowski I, Savourey G, Launay JC, Besnard Y, Cottet-Emard JM, Pequignot JM, Bittel J. Sympathetic stimulation induced by hand
cooling alters cold-induced vasodilatation in humans. European Journal of Applied Physiology. 2000;81(4):303-9.
63. OBrien C, Montain SJ. Hypohydration effect on finger skin temperature and blood flow during cold-water finger immersion. Journal
of Applied Physiology. 2003;94(2):598-603. Epub 2002 Oct 11.
64. Shepherd JT, Rusch NJ, Vanhoutte PM. Effect of cold on the blood vessel wall. General Pharmacology 1983; 14:61-64.
65. Lewis T. Observations upon reactions of vessels in human skin to cold. Heart 1930;15:177-208.
66. Knobloch K, Kraemer R, Lichtenberg A, Jagodzinski M, Gosling T, Richter M, et al. Microcirculation of the ankle after Cryo/Cuff
application in healthy volunteers. International Journal of Sports Medicine 2006a;27(3):250-5.
67. Knobloch K, Grasemann R, Jagodzinski M, Richter M, Zeichen J, Krettek C. Changes of Achilles midportiontendon microcirculation
after repetitive simultaneous cryotherapy and compression using a Cryo/Cuff. American Journal of Sports Medicine.
2006b;34(12):1953-9.
68. Knobloch K, Grasemann R, Spies M, Vogt PM. Midportion Achilles tendon microcirculation after intermittent combined cryotherapy
and compression with cryotherapy alone: a randomized trial. American Journal of Sports Medicine 2008;36(11):2128-38.
69. Knobloch K, Grasemann R, Spies M, Vogt PM. Intermittent KoldBlue cryotherapy of 3x10 min changes mid-portion Achilles tendon
microcirculation. British Journal of Sports Medicine 2007;41(6):e4.
70. Yanagisawa O, Homma T, Okuwaki T, Shimao D, Takahashi H. Effects of cooling on human skin and skeletal muscle. European
Journal of Applied Physiology. 2007;100(6):737-45. Epub 2007 May 4.
71. Karunakara, R G. Lephart, S M. Pincivero, D M. Changes in forearm blood flow during single and intermittent cold application.
Journal of Orthopaedic and Sports Physical Therapy 1999; 29(3):177-80.
72. Fiscus KA, Kaminski TW, Powers ME. Changes in lower leg blood flow during warm, cold and contrast therapy. Archives of Physical
Medicine and Rehabilitation. 2005;86(7):1404-10.
73. Yanagisawa O, Kudo H, Takahashi N, Yoshioka H. Magnetic resonance imaging evaluation of cooling on blood flow and oedema in
skeletal muscles after exercise. European Journal of Applied Physiology. 2004;91(5-6):737-40. Epub 2004 Mar 13.

Figure 1
Risk of bias graph for Human EIMD inflammatory models: Review authors judgements about each risk of bias item presented as percentages across studies.

Adequate sequence generation?

Allocation concealment?
Blinding?

Incomplete outcome data addressed?

0%
Yes (low risk of bias)

Unclear

25%

50%

75%

100%

No (high risk of bias)

Figure 2
Summary of Human models (Ice and EIMD)
Study
(Factors lowering
risk of bias3)

Inclusion

EIMD

Interventions

Summary of results

N=15 healthy
females; mean
age 22 +/-2 y

5 sets of 8 maximal
elbow flexions
(eccentric and
concentric), 60 s
rest between
each set

-I (n=8): CWI at
15C for 15 min,
immediately post
exercise, and every
12 hours thereafter
for 3 days
-C (n=7): No
treatment

Plasma CK
I<C (48h*, 72h*)

N=9 healthy
resistance
trained males;
mean age 23.3
(SD: 3y)

3 sets of 10 biceps
curls at 70% 1RM

-I (n=9): 15 mins of
ice massage
-C (n=9): 5 min
sham ultrasound
Both interventions:
immediately, 24h
and 48h post
exercise

Plasma CK
I<C (24h, 48h,
72 h*)

N=28 healthy
untrained
participants;
mean age 23.8
(SD 1.8 y)

5 sets of 20 resisted
calf raises at 30%
of participants
maximal voluntary
contraction, 1 min
rest between sets

-Single I (n=9):
CWI in water at
5C for 15 mins,
Immediately after
exercise
-Double I (n=9):
CWI in water at
5C for 15 mins,
immediately after
exercise, and 24
hours after exercise
-C (n=10): No I

CK [change within
groups]
Post C > Pre C (48,
96 hrs*)
Post DI > Pre DI (48,
96 hrs*)
Post I>Pre I (96 hrs)

ICE INTERVENTIONS
Eston, 1999
RCT

Howatson, 2003
Randomised cross
over: 2 weeks wash
out; opposite arms
(Sequence generation)

Yanagisawa, 2003
RCT

LDH
Post C > Pre C (96
hrs*, 168hrs)
Post DI > Pre DI (96
hrs, 168hrs)
Post I>Pre I (96 hrs)
For both outcomes,
SI had the smallest
increase above
baseline

Howatson, 2005
Randomised cross
over : 2 weeks wash
out; contra-lateral
arms
(Sequence generation)

N=12 healthy
physically active
males; mean
age 24.8 (SD:
5.3y)

3 sets of 10
maximal eccentric
bicep contractions
on isokinetic
dynamometer, 3
min rest between
sets

-I (N=12): Ice
massage, 15 mins
-C (n=12): 5 min
sham ultrasound
All interventions:
immediately, 24,
and 48 hrs post
exercise

CK
I=C (24, 48)
I<C (72, 96)
Myoglobin
I=C (24, 48, 72, 96)

Study
(Factors lowering
risk of bias3)

Inclusion

EIMD

Interventions

Summary of results

N=20 healthy
physically active
males; aged
20.4 (+/-1.7y)

100
countermovement
jumps from a 0.75
height

-I (n=20): CWI at
15C (+/- 1C) for
15 mins x 2. 10
mins between each
immersion. Undertaken
immediately after
exercise and repeated
at 4h, 8 h, 24 h
-C (n=20): no I

Plasma CK
I<C (24h, 48h)

N=40 healthy
adults; mean
age 21 (SD:
4.3y)

5 sets of 10
repetitions
of eccentric
quadriceps
exercise at 120%
of 1 repetition
maximum, 1 mins
rest between sets

-I (n=20): CWI to ASIS

at 5C, 1 mins x 3,
initiated immediately
post exercise
-C (n=20): CWI to ASIS
in 24C water, 1 mins
x 3. Participants rested
out of the bath for 60,
initiated immediately
post exercise

Plasma CK
I=C (24h, 48h, 72h)

n=18 physically
active and
healthy males;
mean age: 24
(SD 5 y)

5 sets of 20
drop jumps on a
concrete based
floor. 10 s rest
between each jump
and 2 min rest
between each set.

-I (n=9): Seated CWI

15C to iliac crest for
12 mins. Undertaken
immediately post
exercise, and once
every 24 for 3 days
-C (n=9): Seated, no
immersion

Plasma CK
I=C (0-96h)

N=38 healthy
strength trained
males

7 sets of 10
eccentric repetitions
on a leg press
machine; 3 min rest
between sets)

-I (n=12): Whole body

CWI (excluding head
and neck) at 15C
-HWI (n=11): Whole
body HWI (excluding
head and neck) at 38C
-Contrast (n=15):
Whole body CWI
(excluding head and
neck) at 15C for 1
min, followed by whole
body HWI (excluding
head and neck) at 38C
for 1 min, repeated 7
times
-C (n=38):Seated with
minimal movement

ICE
INTERVENTIONS
Skurvydas, 2006
Randomised cross
over: 9-10 month
wash out

Sellwood, 2007
RCT
(Sequence
generation;
allocation
concealment;
blinded outcome
assessment; follow
up well described)
Goodall, 2008
RCT

Vaile, 2008
Randomised cross
over: 8 month wash
out
(Sequence
generation)

Interventions
were undertaken
immediately, 24,
48 and 72 hours
after exercise. All
interventions were 14
min durations

CK
I<C (24h*, 72h*)
HWI<C (48h)
LDH
I=C ( 0h, 24h, 48h,
72h)
Myoglobin
I=C (0h, 24h)
IL-6
I=C (0h,24h)

Study
(Factors lowering
risk of bias3)

Inclusion

EIMD

Interventions

Summary of results

N=16 healthy
recreational
male athletes;
mean age 23.3
(SD 3y)

100 drop jumps

from a 0.6m height
(5 sets of 20 with 2
mins between each
set)

-I (n=16): CWI in 15C

to ASIS for 12 min
-C (n=16): Seated for
12 min
Both interventions:
undertaken immediately
post exercise and at 24
hour increments for the
following 3 days

Plasma CK
I=C (24h, 48h, 72h,
96h)

N=18 healthy
females; aged
19.9+/- 0.97 y

10 sets of 10 CMJ,
adopting a 90 deg
knee angle on each
landing

-I (n=9?): Seated CWI

at 10C, for 10 min.
Undertaken within 10
min of exercise.
-C (n=9?): ?No I

Plasma CK
I=C (1h, 24h,
48h,72h, 96 h)

ICE
INTERVENTIONS
Howatson, 2009
Randomised Cross
Over
(Follow up well
described)

Jakeman, 2009
RCT
(Sequence
generation)

EIMD: Exercise induced muscle damage

I: Ice
C: Control (unless otherwise stated control involved rest/sitting)
CWI: cold water immersion
HWI: hot water immersion
CMJ: Countermovement Jump
CK: Creatine Kinase
Deg: degree
s: S
Min: Mins
M: metre
Y: years
ASIS: anterior superior iliac spine
Pre: pre exercise values
Post: post exercise values
h: hours post exercise/loading (eg. 24h = 24 hours post exercise/loading)
*: p<0.05
: p<0.001

Figure 3
Summary of Human models (Compression and EIMD)

Study
(Factors lowering
risk of bias3)

Inclusion

EIMD

Interventions

Summary of results

Kraemer, 2001a

N=20 healthy
females; mean age:
COMP 21.3 (SD:
2.9y), C 21.1 (SD:
3.3y)

2 sets of 50 passive
arm curls with a
maximal concentric
and eccentric
contraction every
fourth rep

COMP (n=10):
Compressive sleeves
for 5days (10mmHg)
C (n=10): no
intervention

Serum CK
COMP=C (24h)
COMP<C (48h*,
72h*, 96h*, 120h*)
Serum LDH
COMP=C (24h,
48h, 72h, 96h,
120h)

N=15 healthy
males; mean age:
COMP 22.3 (SD:
2.9y), C 22.1 (SD:
3.3y)

2 sets of 50 arm
eccentric curls, 3
minutes between
sets

COMP (n=8)
Compression
sleeve for 3 days
(10mmHg)
C (n=7): no
intervention

Serum CK
COMP=C (24h,
48h)
COMP<C (72h*)

N=26 healthy
active males; aged
24.12 +/-3.2y

6 sets of 10 back
squats at 100%
body weight, each
set included an
eccentric squat at
1RM

Contrast (n=10):
seated CWI 60
secs, immediately
followed by seated
HWI, 180 secs x 3
COMP (n=10):12 h
wearing tights with
mean compression
of 12mmHg at the
calf, 10mmHg at
thigh
C (n=6): no
intervention

Serum CK
COMP=C (1h, 24h,
48hr)
Serum myoglobin
COMP=C (1h, 24h,
48h)

N=11 healthy
netballers/
basketballers;
7 female (aged
19.7+/-0.5y), 4
male (aged 26.3
+/-5.1y)

5 sets of 20 CMJs

COMP (n=11):
graduated
compression tights
(15mm/Hg) for 48h
C (n=11): Passive
recovery for 48h

Serum CK
COMP=C (24h,
48h)
Serum LDH
COMP=C (24h,
48h)

N=11 team sport

athletes; mean age
20.9 (SD 2.7y)

10 sets of 20 m
sprints plus 10
sets of 10 double
leg bounds; 1 min
between sets

COMP (n=11):
Compression
garments, during
exercise and for 24
h post exercise
C (n=11): no
compressive
garments

Serum CK
COMP=C (0h, 2h,
24h)
AST
COMP=C (0h, 2h)
COMP<C (24h)
trend, but not
significant
CRP
COMP=C (0h, 2h,
24h)

RCT

Kraemer, 2001b
RCT

French, 2008
RCT

Davies, 2009
Randomised
cross over: 7 days
between intervention

Duffield, 2010
Randomised
cross over: 7
days between
interventions

Study
(Factors lowering
risk of bias3)

Inclusion

EIMD

Interventions

Summary of results

Jakeman, 2010

N=17 healthy
female participants;
mean age21.4 (SD:
1.7y)

10 sets of 10 CMJs
from a 0.6metre
height, 10 secs
between jumps, 1
min between sets

COMP (n=8): Full

leg compression
stockings, 12 h post
exercise
C (n=9): Passive
recovery

Serum CK
COMP=C (1h, 24h,
48h, 72h, 96h)

N=20 healthy
highly resistance
trained participants;
11 males [mean
age 23 (SD: 2.9y)];
9 females [mean
age 23.1(SD: 2.2y)]

3 sets of 8-10 reps

of 8 different whole
body resistance
exercises; 2-2.5
min rest between
sets

COMP: Whole
body compressive
garment for 24h

Serum CK
COMP<C (24h*)
LDH
COMP=C (24h)

RCT

Kraemer, 2010
RCT

C: non compression
garments

AST: Aspartate transaminase

C-RP: C-Reactive protein
LDH: Lactate dehydrogenase
CMJ: countermovement jump

Figure 4
Summary of Human models (Protection / Rest and EIMD)
Study
(Factors lowering
risk of bias3)

Inclusion

EIMD

Interventions

Summary of results

Sayers, 2000

N=26 participants

50 maximal
eccentric
contractions of the
elbow flexors

IMM (n=9): Cast

(immediately post
exercise) at 90 deg
elbow flexion for 4d
EX (n=9): 2 sets of
25 biceps curls/d
oer 4d, with a 5lb
weight, 2 min rest
between sets
C (8): no
intervention

Plasma CK
IMM=C=EX (24 h)
IMM<C=EX (48h*,
72h* , 92h*)

N=25 healthy male

participants

2 sets of 25
maximal eccentric
contractions of
elbow; 5 minutes
between sets

IMM (n=12): Cast

(immediately post
exercise) at 90 deg
elbow flexion for 4d
C (n=13): no
intervention

Plasma CK
IMM=C (24h)
IMM<C (48h*, 72h*
, 96h*)
Plasma myoglobin
IMM=C=EX (0-5d)

N=10 healthy
participants; mean
age 23+/- 4.2y; 5
male, 5 female

10 sets of 6
isokinetic eccentric
contractions of
elbow flexors

IMM arm: Sling

(30 minutes post
exercise) at 90 deg
elbow flexion for 4d
C arm: No
intervention for 4d

Plasma CK
IMM=C (24h, 48h,
72h, 96h; 7d)

RCT

Sayers, 2003
RCT

Zainuddin, 2005
Randomised
cross over: 2
weeks between
interventions,
contralateral arm as
control

Figure 5
Summary of Human Models (other soft tissue injury)
Study
(Factors lowering
risk of bias3)

Inclusion

Interventions

Summary of results

Stalman, 2008

N=20 participants
undergoing knee
surgery

-Arthroscopy (n=10): I/A

morphine and I/C

Microdialysis (Synovial
membrane vs adipose
reference tissue)
Lactate levels post surgery
-Arth: Syn Mem> ref tissue*
-ACL: Syn Mem=ref tissue

Case control

-ACL reconstruction (n=10):

No I/C

Consumption of glucose levels

post surgery:
-Arth: Syn Mem> ref tissue*
-ACL: Syn Mem=ref tissue
Glycerol levels post surgery
- ACL: Syn Mem=ref tissue
Prostaglandin (PGE2) levels
post surgery
-ACL<Arth*
Strunk, 2006
Case series
(Blinded outcome
assessor)

N= 13 participants
with clinically active
RA at the wrist

I/C: ice and compression

Arth: arthroscopy
RA: rheumatoid arthritis
Syn Mem: synovial membrane
2 and 3 D: 2 and 3 dimensions

I: cold pack for 20 min (skin

temperature reduced to 5.5C)

Power Doppler
ultrasonography (2 and 3D
PDUS)
Synovium vascularity (based
on a semi-quantitative
grading system)
-Decreased in 54% of patients
after I

Intact skin
(lowest
temperature
recorded in
deep muscle
20C)
Intact
unshaven
skin

Crushed ice
(n=5) and
compression
20 mins x 2

Cold pack with

compression
and elevation;
5 mins every
14 x 4 (n=14 )

Ice cylinders;
20 mins every
6 hrs x 3 (n=6)

Bilateral
crush injury
radiocarpal
ligament
(NS)

Blunt trauma
left calf
(NS)

Blunt trauma
skin side of
DMC (a)

Farry, 1980

Hurme, 1993

Smith, 1993

Intact skin

CWI intact
shaven skin

CWI at 20C; 1
hr (n=5)
CWI at 30 C 1
hr (n=5)
CWI both
limbs; 1 hr x 3
(n=10)

[total dosage]

Rolling crush
injury forelimb
(NS)

Directness
of cooling

McMaster,
1980

Details of
intervention

Injury site
(severity):

Study

COOLING

Summary of Animal Models

Figure 6

Immediate
(likely)

Immediate

Immediate
(likely)

Immediate
(likely)

Time after
injury of ice
initiation

No treatment
(n=6)

No treatment
(n=14)

Compression
at 106Nm2 to
contralateral
limb (n=5)

No treatment
(n=10)

Control group

Y (lightly)

Yes

Anaesthetized
during
trauma (Y/N

Y (lightly)

Yes

Anaesthetized
during
cooling (Y/N)

Y (lightly)

Yes

Anaesthetized
during
outcome
assessment
(Y/N)

Laser Doppler
fluxmetry

Intravital
microscopy
with MC (Y)

IHA (Y)

IHA (Y)

Water
displacement
(N)

Outcomes
(Blinded
assessor Y/N)

Intact
unshaven
skin

Intact skin

CWI to intact
shaved limbs

Unshaven
intact skin

Ice pellets
directly against
the fur; 20
minutes each
day for 4 days)

Ice cylinders;
20 mins every
6 hrs for 2
days (n=16)

CWI in 12.8
-15.6C);
30 mins x 4
(n=16)

Ice pack with

elastic tape; 5
hrs (n=10)

Blunt trauma
on skin side of
DMC
(a)

Blunt trauma
cutaneous
maximus
muscle
(NS?)

Blunt trauma
to plantar
aspect of foot,
bilateral
(a)

Blunt trauma
right calf
(NS)

Curl, 1997

Dolan, 1997

Merrick, 1999

Directness
of cooling

Smith, 1994

[total dosage]

Details of
intervention

Injury site
(severity):

Study

Not explicitly
stated (likely
immediate)

5 minutes

5 minutes

6 days?

Time after
injury of ice
initiation

No treatment
(n=9); also
contralateral
limb of
treatment
group used for

Yes (likely)

Yes (n=8
Ketamine/
Xylazine; n=8
Isoflurane)

No treatment
(contralateral
limb) (n=16)

WI in 2225.5C;
30 mins x
4 (n=16,
contralateral
limb)

Anaesthetized
during
cooling (Y/N)

Anaesthetized
during
trauma (Y/N

No treatment

Control group

Outcomes
undertaken on
excised tissue

No

Anaesthetized
during
outcome
assessment
(Y/N)

Biochemical
assay (N)

Water
displacement
(N)

Laser fluxmetry
(N)

Intravital
microscopy
with MC

Laser Doppler
fluxmetry
/ perfusion
imager (Y)

Intravital
microscopy
with
MC

Outcomes
(Blinded
assessor Y/N)

Unshaven
intact skin

Unshaven
intact skin

CWI intact
shaven limbs

CWI at 20C; 30
mins duration

Cylinder of ice
to skin side of
chamber; 20
minutes (n=15)

CWI at 12.8 C; 3
hrs, followed by 1
hrs rest

Formulin
injection
Plantar aspect
R hind paw

Blunt trauma
cutaneous
maximus
muscle (b)

Blunt trauma
to plantar
aspect of foot,
bilateral
(a)

Blunt trauma
to cremaster
muscle
(b)

Kenjo, 2002

Deal, 2002

Dolan, 2003

Lee, 2005

Saline at 27C; 10
minutes (n=7)

Saline at 3C; 10
minutes (n=7)

CWI at 12.8 C) for

1 hr; 2 hrs HVES;
1 hrs rest
Cooling
directly onto
exposed
muscle
surface

Through
microvasular
chamber

Ice cold saline

solution; 1 hr
duration; surface
temperature
decreased to 10
(2C) (n=9)

hrTNF : 2000
units (in 0.1mL
phosphatebuffered
solution) to
striated muscle
in back

Directness
of cooling

Westermann,
1999

[total dosage]

Details of
intervention

Injury site
(severity):

Study

5 minutes

5 mins

15 mins

Immediate

Immediate

Time after
injury of ice
initiation

HVES for 3 hrs

Saline at 37C;
10 minutes
(n=7)

Anaesthetized
during
trauma (Y/N

No treatment
(n=5)

No treatment?

No treatment
(n=8)

Control group

Likely (cooling
initiated 15
mins post
injury)

Anaesthetized
during
cooling (Y/N)

Outcomes
undertaken on
excised tissue

Anaesthetized
during
outcome
assessment
(Y/N)

Real time laser

Intravital
microscopy (N)

Water
displacement
(N)

Intravital
fluorescent
microscopy
with MC (N)

Limb volume
(N)

Fos immunereactivity
expression
(spinal cord)
(N)

Intravital
microscopy
with MC
(N)

Outcomes
(Blinded
assessor Y/N)

30 minutes
of cyclical
compression
(0.5Hz), one
treatment per day
for 4 days (n=6);

EIMD, tibialis
anterior: 7 sets
of 10 cyclic
lengthening
contractions,
2 mins rest
between sets

Butterfield,
2008
30 min

Anaesthetized
during trauma
(Y/N)
Y

Contralateral
limb EIMD, no
intervention

Y( 6 hrs)

Anaesthetized
during
outcome
assessment
(Y/N)

Anaesthetized
during outcome
assessment (Y/N)

Anaesthetized
during
cooling (Y/N)

Anaesthetized
during compression
(Y/N)

Yes (Likely)

Anaesthetized
during
trauma (Y/N

Control group

No treatment
(n=7)

No treatment
(n=7)

Control
group

IHA
(Y)

Outcomes
(Blinded assessor
Y/N)

IHA
(Y)

Intravital
microscopy

IHA
(Y)

IIntravital
microscopy

Outcomes
(Blinded
assessor Y/N)

a: injury severity enough to cause extravastation of RBCs, but not haemorrhage sufficient to obscure the vascular bed / bleeding suppressed to minimum / tissue trauma without rupturing major vessels
b: Rats with haematoma, local inflammation, ischaemia, necrosis, inconsistent microvascular flow, or an increased number of leukocytes were excluded.
c: Described as severe, but without incapacitating compartment syndrome
NS: Not stated
HVES: High voltage electrical stimulation
EDL: extensor digitorum longus
CWI: cold water immersion
WI: water immersion
Yes (Likely): although not stated specifically, it was likely based on the experimental set up that eg. Animals were anaesthetized during cooling.
IHA: Immunohistological analysis
MC: Microvascular chamber

Details of
intervention

Injury site
(severity):

Study

Time after injury of

treatment initiation

Immediate

Shaven intact
skin (muscle
surface
temperature
cooled to
10C)

Saline at 8C; 6 hrs

(n=7)

Blunt trauma
left tibial
muscle
compartment
(c)

Schaser, 2007

COMPRESSION

Immediate

Direct to
surgically
exposed
muscle
(muscle
surface
temperature
cooled to
10C)

Saline at 8C; 20
minutes (surface
temperature
reductions to 10 C)
(n=7)

Time after
injury of ice
initiation

Blunt trauma to
EDL muscle
(c)

Directness of
cooling

Schaser, 2006

[total dosage]

Details of
intervention

Injury site
(severity):

Study

Figure 7
Overview of studies assessing the effect of PRICE on inflammation by: model,
injury type and cooling dose

Chapter 4
What effect does mechanical loading have on inflammation and soft tissue
healing after acute injury?
What is known in this area: Acute soft tissue injury should be managed initially with protection/rest, followed by
progressive mobilisation and stress induction on the healing tissue. The basic science mechanisms underpinning
this approach have not been fully elucidated, and there are few definitive guidelines on optimal timing or dosage
of loading during recovery.
Aim: To update the pathophysiological rationale for using PRICE in the management of soft tissue injury.
Clinical question: What effect does mechanical loading have on inflammation and soft tissue healing after acute
injury?
Objective: To consider the basic scientific rationale for using protection, rest and progressive loading and consider
whether the nature, timing and dosage of tissue loading effects molecular, histological and mechanical outcome
after acute soft tissue injury.
What this review adds: There is further evidence from animal models that short periods of protection/rest are
required after soft tissue injury, and that aggressive ambulation or exercise should be avoided in the acute stages.
Longer periods of protection/rest are deleterious with adverse changes to tissue biomechanics and morphology
within 2-3 weeks of stress shielding. Paradoxically, progressive mechanical loading is more likely to restore the
strength and morphological characteristics of collagenous tissue after injury. The exact mechanism has yet to be fully
elucidated but there are consistent findings that mechanical loading upregulates gene expression for key proteins
associated with soft tissue healing.
Limitations and future study recommendations:
The clinical significance of many of these results is limited. Clearly, animal models are fundamentally different to
humans in terms of metabolism, anatomy and movement pattern. We must also consider that the protection/rest
protocols may not have been fully adhered. For example in an animal model, cast immobilisation was associated
with weight-bearing, and even hind limb suspension would not fully inhibit muscle activity and joint motion.10
Further limitations arise when we consider that the majority of injuries studied are based on surgical resection,
whereas soft tissue tears in humans are usually associated with frayed ends across the wound site. Almost all of the
studies have used a tendon model, and we can also not assume that other soft tissue will respond in the same way.
This review has provided further rationale that mechanical load should be induced on healing tissue; and that
there is an optimal window for its initiation. Further research is required on optimal intensities and durations for
mechanical loading in human subjects. The idea of tissue memory in tendon injuries is an interesting concept,16 and
is supported somewhat by evidence that mechanical loading can up-regulate or down-regulate gene expression for
key proteins during healing. This concept requires further human research, particularly with ligament and muscle
injury models.

Background
The protection and rest components of PRICE are commonly recommended after an acute soft tissue injury. The rationale
is to prevent further bleeding, excessive distension or re-rupture at the injury site.1 Protection and rest of an injured
ligament, usually involves casting or bracing of the entire joint plus or minus weight-bearing; whereas for muscle strains
or contusions, adhesive bandaging or crutch walking may be more practical.2
Deciding how long protection and rest should be enforced is controversial. Excessive immobilisation seems to result in poor
mechanical and functional outcome.1 Recent reviews1-5 suggest that 2-5 days of immobilisation is optimal, with similar
consensus reached in the original ACPSM guidelines on PRICE. There is further evidence1-4 that immobilisation should
be followed by progressive mobilisation and stress induction on the healing tissue. This is often defined as functional
treatment, and in clinical terms, involves early mobilisation and weight-bearing. Anecdotal evidence suggests that most
clinicians make the transition from protection/rest to weight-bearing carefully and within the limits of pain. However,
evidence based guidance on the optimal timing, dosage and nature of initiating loading after injury is limited, and it is not
clear if this should vary according to the type of soft tissue affected (eg. muscle, tendon, ligament).
Gaining a balance between protection/rest, and reloading after injury is important for optimal restoration of the
biomechanical and functional properties of soft tissue. The complexity of tissues response to the presence or absence of
mechanical stimuli is complex however, particularly at a physiological, cellular and molecular level.

Objective
The objective of this chapter was to consider the basic scientific rationale for using protection, rest and progressive loading,
and consider whether the nature, timing and dosage of tissue loading effects molecular, histological and mechanical
outcome after acute soft tissue injury.
Methods
Literature search
We searched the titles generated from the main search strategy outlined in Appendix tables 1-4. Relevant studies were
extracted, with exclusions made based on titles, abstracts or full text versions.
Inclusion criteria
No restrictions were made on study design. We included any animal or human participants with an acute muscle, ligament
or tendon injury. Injury could have been either induced (eg. surgically) or naturally incurred. Injury models using surgical
repair were not considered. Interventions were any form of intervention aiming to protect /rest, or modify activity, initiated
in the acute phases post injury (<72 hours). Studies involving protection/rest for longer than 8 weeks post injury were not
considered. Comparison could have been to different types of protection/rest, normal ambulatory activity or additional
exercise. Exercise could have been either weight-bearing or non weight-bearing. Outcomes could have involved any
type of histological or mechanical measures related to soft tissue healing. We also included outcomes relating to gene
expression for key proteins associated with inflammatory or proliferative response after injury.
Data extraction
We extracted details on: type of soft tissue injury, timing and dosage of intervention and comparison groups, cointerventions, outcomes and statistical significance. All included studies were assessed in terms of design, and assessor
blinding based on Cochrane risk of bias tool6
Results
Excluded studies (with reasons) are listed in Appendix Table 5. 12 studies were eligible for inclusion. Table 1 summarises
the key subject characteristics and extracted data. All studies7-18 focused on animal subjects, three10,11,14 used an injured
ligament model, with the remainder focusing on tendons. Injuries were induced surgically aside from one model,13 which
used a chemical irritant. In all studies, the tissues subjected to injury were part of a weight-bearing joint. Protection/rest was
undertaken using a range of protocols: non weight-bearing using suspension,16 cast8 or pin10 immobilisation, tenotomy,7
or botulinum toxin injection.15,18 In all cases, the comparison groups undertook unrestricted ambulation after injury, which
in some cases included additional exercise.9,12,13
Outcomes focused primarily on mechanical and histological tissue properties, with a number including genetic expression
for protein relevant to various inflammatory or proliferative healing events. All outcomes involved euthanizing animals for
analysis of excised tissue samples. Studies used a range of follow up times and sub-groups of animals were euthanized at
various time points during the study. Histological analysis was based on subjective assessment, and in one case10 this was
facilitated with scanning electron microscopy (SEM). Three studies16-18 reported blinded assessment of outcome.
Mechanical properties
Protection/rest vs ambulation
Halikis et al.8 measured the work of flexion (WOF) (ie. the tissues resistance to passive flexion), in healing chickens foot
tendons. WOF increased in all groups during the first week after injury; however this was significantly greater in tendons
subjected to immediate ambulation compared to cast immobilisation. This pattern was reversed after the first week and
WOF continued to increase in the protected group only. At three weeks post injury, WOF had returned to normal in the
ambulation groups, and values in the immobilised group remained 63% above baseline. In a related model, Kubota et
al.7 also found that longer periods of protection/rest had a negative effect on mechanical strength after injury. Chicken
tendons treated with four weeks of protection/rest (using both tenotomy and cast immobilisation) had significantly lower
tensile strength, compared to those subjected to full ambulation after injury.
As a secondary objective, this study also compared various combinations of tenotomy (to prevent tension) and casting (to
prevent motion), concluding that both forces must be induced to fully benefit from ambulation. Anderson et al.16 used a
rat model comparing complete unloading with tail suspension, to full time cage activity, after Achilles tendon injury. Of
note, all groups undertook full time activity for the first two days post injury, prior to intervention. 12 days post injury; the
loaded tendons had the best mechanical properties, with significantly higher amounts of tendon stiffness, cross sectional
area, and gap distance, ultimate stress, with three times the tensile strength of the unloaded group. Eliasson et al.18 also
used a rat model with injured Achilles tendons.

All animals undertook cage activity for three weeks post injury, however in one group; tendons were unloaded by injecting
the calf muscle with botulinum toxin. Although there were no significant differences between groups in the early stages post
injury, from day 8 onwards, the protection/rest group had poorer mechanical properties in terms of the length, stiffness
and displacement at rupture.
There was evidence from two rat ligament models,10,11 that prolonged immobilisation after injury is less effective than
ambulation in terms of mechanical outcome. Thorton et al.11 found that immobilisation in full knee flexion was associated
with significantly higher chance of scar failures at week 6 and 14 post injury; during both static and cyclical tensile loading.
Ambulation was also more likely to restore other visco-elastic properties such as creep and laxity. Using a similar model
and follow up (albeit with a more severe ligament injury), Provenzano et al.10 also noted similar patterns after shorter
periods of unloading. Significant decreases in maximal force, ultimate stress and elastic modulus were evident after 3
weeks of hind limb unloading tissue, compared to a loaded, ambulatory group. Similar findings were reported 7 weeks
post injury.
Ambulation plus exercise
Two rat studies9,12 compared the effects of undertaking a structured exercise program after injury, in addition to regular
cage ambulation. Murrel et al.9 found that, ten bouts of daily swimming offered no additional effect over free cage
activity alone, in terms of the mechanical properties of healing tendons. See et al.12 undertook a follow up study but also
included a running intervention group. Other notable differences were that only half of the Achilles tendon was transected,
and rather than exercising immediately post injury, a 5 day period of rest period was enforced in each group. Again,
swimming offered no additional effect over regular cage activity; however the addition of a daily running program resulted
in some mechanical superiority and ultimate tensile strength was significantly higher in this group at 1 month post injury.
Andersson and colleagues16 also considered whether there is an optimal dose of exercise after tendon injury; comparing
regular cage ambulation, tail suspension (full time), or tail suspension interspersed with short bouts of daily exercise.
Full time tail suspension resulted in the worst mechanical outcome, whereas, regular cage activity was associated with
significantly higher tensile strength. Notably, groups undertaking one daily bout of 15 minute treadmill exercise had tissue
strength that was approximately half that of the full time cage activity group at 12 days post injury. Of further interest,
increasing the running dose to, two bouts of 15 minutes, or 30-60 minutes of continuous running offered little additional
change to mechanical properties.
In a novel method of injury induction, Godbout et al.13 injected rat TA with collagenase. All animals were permitted
free cage activity; however two sub-groups undertook additional exercise on a running wheel. This was initiated either
immediately after injury, or after a seven day delay. Mechanical testing at one month post injury found that the immediate
exercise group had significantly lower levels of stiffness and force at rupture point. Interestingly, the inclusion of delayed
exercise (in addition to free cage activity) produced the best mechanical outcome at 1 month.
Histology
Protection/rest vs ambulation
Three studies found evidence of superior tendon healing in ambulated animals, based on tissue histology,7,17 and scanning
electron microscopy (SEM).10 In general, the ambulated groups exemplified more regenerative activity and superior tissue
morphology in comparison to protection/rest groups. Specifically ambulation resulted in better ECM continuity,10 better
alignment collagen bundles in relation to the longitudinal axis of the ligament,7,10,17 and higher numbers of newly formed
capillaries, and spool shaped fibroblasts.17 Many of these observations were made at around 1 week post injury,17 with
differences between groups becoming more evident at 3-4 weeks.10,17
Delayed vs immediate exercise
Godbout et al13 used immunohistochemistry to assess inflammatory and proliferative cell accumulation at various time
points after injury. Immediate exercise seemed to exacerbate the inflammatory response, exemplified by significantly higher
levels of neutrophils, ED1+ and ED2+ macrophages, at day 3 and 7 post injury. There were no differences in proliferative
cell accumulation at day 7 between groups.
Gene expression
Protection/rest vs ambulation
Three studies10,17,18 measured the effect that protection/rest or ambulation had on mRNA expression after injury. In
general, they focused on the expression levels of key proteins associated with inflammation, growth, tendon specificity
and extracellular matrix (ECM) deposition. In the acute stages, ambulation was associated with a lower expression of TNF
alpha, transforming growth factor-beta 1, and procollagens I and III, and a significantly lower expression for IL1 beta.18
By day 21, procollagen I, cartilage oligomeric matrix protein (COMP), tenascin-c, tenomodulin and scleraxis were more
expressed in this group, two reaching significance (COMP and tenomodulin). Bring and colleagues17 found similar patterns
from at day 17 post injury, with ambulated groups displaying significant upregulation for collagen I and III,

versican, decorin and biglycan expression; many of these values were approximately 14 times higher than the protection/
rest group. Ambulation was also associated with higher levels of expression for neuropeptide receptors [Substance P (NK1)
- and calcitonin gene-related peptide (CRLR and RAMP-1)] between day 8 and 17 post injury.
Using an injured ligament model, Martinez et al.14 compared unloading with tail suspension, to normal ambulation. After
3 weeks, there were trends towards increased gene expression for collagen Type I and III in the loaded group, and at 7
weeks, collagen type I and V were significantly down regulated in unloaded tissue, and collagen type III was significantly
upgraded in loaded group. There were further significant differences between groups in the expression of other key ECM
constituents; in the unloaded tissue, decorin, tissue inhibitor of matrix metalloproteinase 1 (TIMP-1) and lysyl oxidase were
all significantly down regulated at 3 weeks, with lower expression of matrix metalloproteinase 2 (MMP-2), and higher
expression of TIMP-1 at 7 weeks. Unloaded ligaments also had increased cellularity measured via DNA content at weeks
3 and 7.14
An explorative study by Eliasson et al.18 recorded gene expression of bone morphogenetic protein (BMP) signalling system.
8 genes were studied, however, the only significant finding, was a lower expression of the antagonist follistatin within the
ambulated group. This was observed at multiple time points up to three weeks post injury.

References
1. Kannus P, Parkkari J, Jarvinen TLN, Jarvinen TAH, Jarvinen M. Basic science and clinical studies coincide: active treatment approach
is needed after a sports injury. Scandinavian Journal of Medicine and Science in Sports 2003:13:150-154.
2. Jarvinen TAH Jarvinen TLN, Kaariainen M, Aarimaa V, Vaittinen S, Kalimo H, Jarvinen M. Muscle injuries: optimising recovery. Best
Practice and Research Clinical Rheumatology 2007:21 (2): 317-331.
3. Jarvinen M, Lehto MUK. The effect of early mobilisation and immobilisation on the healing process following muscle injuries. Sports
Medicine 1993;15:78-89
4. Buckwalter JA, Grodzindkey AJ. Loading of healing bone, fibrous tissue, and muscle: implications for orthopaedic practice. Journal
of American Academy of Orthopaedic Surgery 1999; 7:291-299.
5. Jrvinen TAH, Jrvinen TLN, Kriinen M, Kalimo H, Jrvinen M. Muscle Injuries: Biology and Treatment. American Journal of
Sports Medicine 2005 33: 745
6. Higgins JPT, Altman DG (editors). Chapter 8: Assessing risk of bias in included studies. Section 8.5. In: Higgins JPT, Green S
(editors), Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.2 [updated September 2009]. Available
from www.cochrane-handbook.org.
7. Kubota H, Manske PR, Aoki M, Pruitt DL, Larson BJ. Effect of motion and tension on injured flexor tendons in chickens. Journal of
Hand Surgery (American). 1996 May;21(3):456-63.
8. Halikis MN, Manske PR, Kubota H, Aoki M. Effect of immobilization, immediate mobilization, and delayed mobilization on the
resistance to digital flexion using a tendon injury model. Journal of Hand Surgery (American). 1997 May;22(3):464-72.
9. Murrell GA, Jang D, Deng XH, Hannafin JA, Warren RF. Effects of exercise on achilles tendon healing in a rat model. Foot Ankle
International. 1998 Sep;19(9):598-603.
10. Provenzano PP, Martinez DA, Grindeland RE, Dwyer KW , Turner J, Vailas AC, Vanderby R Jr. Hindlimb unloading alters ligaments
healing. Journal of Applied Physiology. 2003 Jan;94(1):314-24. Epub 2002 Sep 20.
11. Thornton GM, Shrive NG, Frank CB. Healing ligaments have decreased cyclic modulus compared to normal ligaments and
immobilisation further compromises healing ligament response to cyclic loading. Journal of Orthopaedic Research.
2003 Jul;21(4):716-22.
12. See EK, Ng GY, Ng CO, Fung DT. Running exercises improve the strength of partially ruptured Achilles tendon. British Journal of
Sports Medicine. 2004 Oct;38(5):597-600.
13. Godbout C, Ang O, Frenette J. Early voluntary exercise does not promote healing in a rat model of Achilles tendon injury. Journal
of Applied Physiology. 2006 Dec;101(6):1720-6. Epub 2006 Aug 17.
14. Martinez DA, Vailas AC, Vanderby R Jr, Grindeland RE. Temporal extracellular matrix adaptations in ligament during wound healing
and hindlimb unloading. American Journal of Physiology, Regulatory, Integrative and Comparative Physiol. 2007
Oct;293(4):R1552-60. Epub 2007 Aug 15.
15. Eliasson P, Fahlgren A, Aspenberg P. Mechanical load and BMP signalling during tendon repair. Clinics in Orthopaedics and Related
Research 2008; 466: 1592-1597.
16. Andersson T, Elliasson P, Aspenberg P. Tissue memory in healing tendons: short loading episodes stimulate healing. Journal of
Applied Physiology 2009;107: 417-421.
17. Bring DK, Reno C, Renstrom P, Salo P, Hart DA, Ackermann PW. Joint immobilization reduces the expression of sensory neuropeptide
receptors and impairs healing after tendon rupture in a rat model. Journal of Orthopaedic Research. 2009 Feb;27(2):274-80.
18. Elliasson P, Andersson T, Aspenberg P. Rat Achilles tendon healing: mechanical loading and gene expression. Journal of Applied
Physiology 2009;107: 399-407.

Subject (injury)

N=53 chickens
(transverse partial
laceration across
50% of FPT)

N=84 Chickens
(surgical trauma to
tendon)

N=20 Rats
(surgical
transection of TA)

N=60 Rats
(MCL surgically
ruptured)

Study
Study type
(assessor
blinding
Y/N)

Kubota,
1996
(N)

Halikis,
1997
(N)

Murrell,
1998
(N)

Provenzano,
2003
(N)

Table 1

-PR (hind limb unloading)

-AMB (free cage activity)

-AMB (free cage activity)

-SWIMMING EX (free cage activity PLUS
swimming exercise = 15 minutes / day
for 15 days)

-PR (cast immobilisation)

-Immediate AMB (light dressing)
-AMB after 3 days (cast for 3 days
post injury)
-AMB after 5 days (cast for 5 days
post injury)

Note: all groups were permitted

unrestricted cage activity. No tension
was achieved by proximal tenotomy,
and no motion was achieved by cast
immobilisation.

- PR (No motion or tension)

- AMB (Motion and tension)
- AMB (Motion only)
- AMB(Tension only)

Intervention

Week: 3, 7

Day: 15

Day: 3, 7,
14, 21

Day: 14,30

Time of
euthanasia
(follow)

Histology (SEM)
-AM group had more typical scar morphology at 3 and 7 weeks.
-PR group had pockets of cell clusters with tissue voids at 3 weeks. At week 7 there was
poor orientation of collagen bundles in relation to the longitudinal axis of the ligament.

Mechanical
-Elastic modulus: AMB > PR (week 3* and 7*)
-Maximum force: AMB > PR (week 3* and 7)
-Ultimate stress: AMB > PR (week 3* and 7*)
-Strain at failure: no difference

Mechanical
No differences: displacement, stiffness, energy, modulus, maximum stress (day 15)

Note: At 21 days WOF had returned to normal (baseline un-injured values) in all AMB
groups, whereas the PR group were 63% greater than baseline levels.

Mechanical
-WOF (Energy required to passively flex digit at 2.4cm/min): PR<Immediate AMB (day 3,
7); Immediate AMB < AMB after 3 days < AMB after 5 days < PR (day 14, 21)

Histology
-Collagen fibre formation: AMB (all types) > PR (day 14 and 30)

Mechanical
-Breaking strength: no differences (day 14); AMB (all types) > PR (day 30*); AMB (motion
and tension)>AMB (motion only or tension only) (day 30*).

Summary of results

Subject (injury)

N=53 Rabbits
(MCL surgically
gapped at mid
substance; bilateral
or unilateral)

N=30 rats
(surgical
transection of right
medial portion
ofTA)

N=180 Rats
(collagenase
injection into TA)

N=60 rats
(bilateral MCL
resection)

Study
Study type
(assessor
blinding
Y/N)

Thornton,
2003
(N)

See 2004
(N) -only
assessment
of the
lower limb
functional
outcome was
blinded)

Godbout,
2006
(N)

Martinez,
2007
(N)

-PR (hind limb unloading)

-AMB (unrestricted cage activity)

-AMB (free cage activity)

-IMMED EX (free cage activity PLUS
immediate exercise on running wheel at
approximately 1 km per day over the first
week)
- DELAY EX (free cage activity PLUS delayed
exercise initiated at 7 days post injury

Note: Running/swimming EX undertaken

for 15-19 minutes per day, from day 5-30
post injury.

Week: 3, 7

Day: 3, 7, 28

Day: 30

Week: 3, 6, 14

-PR (pin immobilised in full flexion)

-AMB (free cage activity)

-AMB (free cage activity)

-SWIMMING EX (free cage activity
PLUS running)
-RUNNING EX (free cage activity
PLUS swimming)

Time of
euthanasia
(follow)

Intervention

Tissue cellularity (based on DNA content)

-AM<PR (week 3* and 7*)

mRNA expression:
-Collagen I and III: AM>PR (3 weeks)
-Decorin, TIMP1, lysyl oxidase: AM>PR (3 weeks*)
-Collagen I, III, V : AM>PR (week 7*)
-MMP2: AM>PR (week 7*)
-TIMP1: AM<PR (week 7*)

Mechanical
-Stiffness; force at rupture point: no difference (day 7); DELAY EX > AMB > IMMED EX
(day 28*)

Immunohistochemistry
-Neutrophil concentration: IMMED EX > AMB (day 3, 7)
-ED1+ macrophage concentration: IMMED EX > AMB (day 3*, 7*)
-ED2+ macrophage concentration: IMMED EX >AMB (day 3*)
-Proliferative cells: no difference (day 7)

Mechanical
-Tensile strength: EX running > AMB (day 30)
-Load relaxation, stiffness,: EX running = EX swimming = AMB (day 30)

Mechanical
-Number of failures during cyclic low load loading: AMB < PR (3, 6 and 14* weeks)
-Laxity: AMB > PR (week 3, 6 and 14*)

Summary of results

Subject (injury)

N=40 rats
(3mm transverse
segment from TA)

N=70 rats
(3 mm full thickness
segment surgically
removed from TA)

N=64 Rats
(TA rupture with a
blunt instrument)

Study
Study type
(assessor
blinding
Y/N)

Elliasson,
2008
(N)

Andersson,
2009
(Y)

Bring, 2009
(Y)

Day: 12

-PR (Full time unloading using tail

suspension)
-AMB (free cage activity)
-EX (tail suspension BUT with intermittent:
a) 15 min cage ambulation
b) 15 min treadmill
c) 30 min treadmill
d) 60 min treadmill
e) 15 min x 2 treadmill

-PR (plaster cast with free cage activity)

-AMB (free cage activity)
Day: 8, 14,
17, 28

Day: 3, 8, 14,
21

-PR (free cage activity however calf muscle

on injured TA injected with botulinum
toxin)
-AMB (free cage activity)

Note: all groups undertook free cage

activity for the first 2 days post injury

Time of
euthanasia
(follow)

Intervention

Histology
Generally higher regenerating activity in AMB. Better structural organisation in AMB,
particularly at day 28
-Amount of newly formed capillaries, longitudinally organised collagen: AMB > PR (day
14)
- Amount of longitudinally organised collagen, fibroblasts number, and number of spool
shaped fibroblasts: AMB<PR (day 28)
-Amount of mature collagen: PR>AMB (day 28) (notes this was more disorganised)

mRNA expression
-Collagen I and III, versican, decorin and biglycan: AMB=PR (day 8); AMB > control
gene; (day 17*); PR=control gene (day 17)
-Sensory neuropeptide receptors [SP-(NK1), CRLR, RAMP-1]: AMB = PR (day 8); AMB>PR
(day 17*)

Mechanical
-Ultimate stress: AMB=EX; AMB>PR (day 12*)
-Peak force: AMB>EX>PR (day 12*)
-Stiffness: AMB>EX >PR (day 12*)
-Cross sectional area: AMB > EX=PR (day 12*)
-Gap distance (between tendon stumps): AMB > EX=PR (day 12*)
-Ultimate stress, Elastic modulus: No difference (day 12)

-mRNA expression for BMP proteins

Follistatin: AMB<PR (day 3*, 8*, 14*, 21*)
OP-1, GDF-5, GDF-6, GDF-7, BMPR-1b, BMPR-2: No differences
Noggin was not detected

Summary of results

N=110 rats
(Surgically
transacted TA)

Eliasson,
2009
(Y)

- PR (free cage activity however calf

muscle on injured TA injected with
botulinum toxin )
-AMB (free cage activity)

Intervention

Day: 3, 8, 14,
21

Time of
euthanasia
(follow)

- Procollagens I and III; LOX, scleraxis: AMB < PR (day 8);

- Procollagen I: AMB > PR (day 14*, 21)
- Tenomodulin, COMP: AMB > PR (day 14*, 21*)
- Tenascin-C, scleraxis: AMB>PR (day 21)

mRNA expression:
-TNF-alpha, TGF-beta 1, procollagens I and III: AMB < PR (day 3)
-IL 1 beta: AMB < PR (day 3*)

Mechanical
-Length, stiffness, displacement at rupture and energy: AMB > PR (day 3, 8*,14* and 21*)
-Peak stress: AMB >PR (day 3, 8, 14, 21)

Summary of results

: Intervention continued up to these time points

Botox: botulinen toxin injection to calf
ED1+ and ED2+: represent subpopulation of macrophage
BMP: Bone morphogenetic proteins (BMP), two receptors, and two antagonists
FPT: Flexor Profundus Tendon
TA: Achilles Tendon
MCL: Medial collateral ligament of knee
NOTE: in all cases animals were euthanized prior to testing
TIMP1: tissue inhibitor of matrix metalloproteinase 1
MMP-2: matrix metalloproteinase 2
SEM: Scanning electron microscope
WOF: work of flexion
LOX: lysyl oxidase
COMP: cartilage oligomeric matrix protein
In the outcome column, the greater than (>) and less than (<) symbols refer to the absolute differences between groups. NOTE that the effect is always in favour of the intervention listed first. Ie. A>B for
outcomes in which higher values are deemed to be more positive, and: A< B; for outcomes where lower values are more positive. * indicates a significant difference between groups at that time point,
except in highlighted cases where gene expression is compared to control or housekeeping gene.

Subject (injury)

Study
Study type
(assessor
blinding
Y/N)

Chapter 5
Do the physiological effects of local tissue cooling affect function,
sporting performance and injury risk?
What is known in this area: Lowering local soft tissue temperature may affect a number of physiological systems.
This may influence: muscle strength, flexibility, range of movement, nerve conduction, and sensori-motor function.
There is little evidenced based consensus on the magnitude or clinical relevance of this, particularly in clinical
scenarios when ice is applied prior to therapeutic exercise or athletic competition.
Aim: To update the pathophysiological rationale for using PRICE in the management of soft tissue injury.
Clinical question: Do the physiological effects of local tissue cooling affect function, sporting performance and
injury risk?
Objective: To review recent literature and summarise the physiological effects of local tissue cooling and determine
its clinical relevance. We were particularly interested in environments when ice may be applied immediately prior to
therapeutic exercise or athletic competition.
What this review adds: There is evidence to suggest that in healthy human models, cooling can have a depressive
effect on a number of physiological systems including: decreased NCV, isometric strength, and rate of force
production. Other aspects of physical performance were influenced to varying degrees: proprioception and functional
performance were either unaffected or slightly adversely affected after icing, and there is conflicting evidence on the
effect of cooling on the visco-elastic properties of soft tissue in terms of stiffness and range of movement. The clinical
relevance of many of the observed physiological changes is questionable. Short periods of ice application seem
to have minimal impact on basic functional activity such as light exercise, walking or therapeutic rehabilitation.
Longer periods of intense cooling negatively influence higher level functional activities (eg. athletic performance),
and potentially increase injury risk. The relevance of these changes and the magnitude of risk depend on the level
of sporting activity, the time period between completing ice application and returning to activity, and the type of soft
tissue cooled. Few relevant studies have been undertaken on injured subjects; applying ice to an acutely injured
ankle has a further negative effect on postural control. Of particular interest were trends that isolated joint cooling
or short periods of muscle cooling, can have an excitatory effect on muscle activation. This was observed in both
healthy and injured (inhibited) models, and seems to align well with the concept of cryokinetics.
Limitations and future study recommendations:
Relevant studies used a wide range of icing protocols, and therefore variable levels of tissue cooling. Some of
the outcomes or measuring techniques were of questionable validity and relevance, and none of the studies used
blinded outcome assessment. The majority of studies used a randomised cross over design; and in some cases the
wash out time between interventions was either minimal or unclear. Few studies have been undertaken on injured
subjects. Future research is needed to determine the potential for using cooling as an adjunct to therapeutic exercise
(cryokinetics); this should be based on acutely injured muscle or joint models.

Background
The basic premise for applying ice is to induce physiological effects which enhance recovery after injury. The most important
of these, are usually considered to be cold induced analgesia, and reduced cellular metabolism. However, local cooling
also has potential to produce concomitant effects on many other physiological systems. This may involve changes to:
muscle strength, flexibility, range of movement, nerve conduction, and sensori-motor control. Although some of these
changes may be subtle, there is little evidenced based consensus on their magnitude or clinical relevance. The proposed
benefits of cryotherapy must be balanced with any potential adverse effects, before making recommendations for its use.
Ice is often used to treat acute injuries in sport. In situations involving minor trauma, athletes will usually return to a
competitive environment, shortly or immediately after the application of cold. Similarly, there is a growing trend of using
ice prior to undertaking therapeutic exercise (cryokinetics).

Objective
The objective of this chapter was to update the evidence base for the physiological effects of local cooling, and consider
their clinical relevance, in terms of function, sporting performance and injury risk.
Methods
We searched the titles generated from the main search strategy outlined in Appendix Table 1-4. Relevant studies were
extracted, with exclusions made based on titles, abstracts or full text versions.
Inclusion criteria
No restrictions were made on study design or participants (injured or healthy). Interventions were any cooling intervention,
with no limitations on mode or dosage of application. There were no restrictions made on study comparison; for observational
studies, pre and post treatment values were compared. Outcomes could involve any physiological outcome relating to
function. Primary outcomes were key correlates of sporting performance such as: muscle strength (and subcomponents
of strength); range of movement; sensori-motor control (proprioception, postural/neuromuscular control); and functional
performance eg. vertical jump, sprint speed, accuracy and precision of movement. Secondary outcomes of interest were
nerve conduction, tissue stiffness, muscle activation patterns, ground reaction force and submaximal strength. Studies
measuring strength or force production during evoked muscle contractions were not considered.
Data extraction
Data were extracted for: study type, participant demographics, cooling mode and dosage, and key outcomes (pre vs
post; or ice vs other intervention/control). We considered outcomes measured at all time points. As one of our primary
objectives was to determine the safety of icing prior physical activity, we were particularly interested in the maximum
potential change in outcome.
Risk of bias
All included studies were assessed in terms of design, and methodological quality based on Cochrane risk of bias tool1
(sequence generation, allocation concealment, assessor blinding, incomplete outcome data).
Results
Excluded studies (with reasons) are listed in Appendix Table 5. Characteristics of included studies2-42 are summarised
in Figure 1. Each included study assessed at least one relevant physiological or clinically based outcome, before
(baseline) and after (post test) cryotherapy. In most cases, a control (rest) or comparison (heating) condition was also
used, with the majority using a cross over design. The time between conditions varied, from a few hours,11 13 15 up to two
days,7,10,21,26,36 with the longest being 15,16,20 to 2 weeks.17 Seven used a randomised controlled methodology to compare
across conditions,4,8,19,35,38-41 of which two4,19 used allocation concealment. No study undertook blinded assessment of
outcome. All but three studies38-41 used healthy participants; the injured models used were acute sprain40 or induced joint
distension at the knee.38,39,41
The majority of studies applied cooling for between 20 and 30 minutes. Three used shorter periods of just 3 minutes,15,16,32
and the longest duration of cooling was 1 hour.34 In most studies, outcome re-testing was undertaken immediately after
cryotherapy removal only; five used multiple follow ups, retesting outcomes up to 15,7 20,40 30,21 6039 and 9041 minutes
post intervention. Figure 2 provides a summary of the cooling protocols used prior to recording primary outcomes.
Primary outcomes
Maximal strength
Four trials studied the effect of 10-30 minute cryotherapy treatments on isokinetic ankle strength, using comparison with
untreated control. In a randomised cross over study,2 30 minute CWI at 10C, did not affect PF peak torque; however it
caused a significant increase in muscular endurance, compared to control. An RCT by Hopkins and colleagues4 which
used allocation concealment also found that cold tended to enhance ankle muscle performance. Of note, cooling was
with a single ice pack for 30 minutes and restricted to the lateral ankle joint; this intervention significantly increased PF
peak torque, compared to control. In contrast, both Borgmeyer et al.5 and Hatzel et al.3 using ice massage and CWI
respectively, found little effect on ankle muscle strength. Hatzel et al.3 did not use any randomisation, and there was just
1 day rest, between cooling and control conditions. They recorded a wide spectrum of strength outcomes (concentric and
eccentric strength, in ankle plantar flexion, dorsiflexion, eversion and inversion but the only significant difference

difference between interventions was that CWI caused a significant reduction in concentric DF. In a randomised cross over
study, Borgmeyer et al.5 found trends that muscle cooling increased concentric (30 deg/sec) arm strength this change was
not significant.
Four studies considered the effects of icing on isometric strength6-9 however, contrasting results were reported. Two used
observational designs6,7 comparing strength before and after icing. Zhou et al.6 found reductions in knee extension
force when muscle temperatures were cooled below 34C; and force production was decreased further (25% lower than
baseline) when muscle temperatures reached 30C. Douris et al.7 demonstrated a significant reduction in grip strength (by
24%) immediately after a 5 minute forearm immersion. This effect diminished with time, and a 5.9 % reduction remained,
15 minutes after immersion. Two randomised studies compared the effects of cooling and heating, prior to isometric
testing at the biceps,8 and triceps surae9 muscles. Although Nosaka et al.8 noted a 10C difference in tissue temperature
(at 15-20mm below skin surface) between groups, this did not affect maximal isometric elbow flexion. In contrast, Kubo
and colleagues9 who seemed to use more intense cooling intervention for the entire lower leg (CWI at 5C), reported cold
induced decreases in isometric PF strength.
Range of movement
In two studies, cooling was applied immediately before35 or during37 lower limb stretching. 20 minutes of simultaneous
cooling (ice pack on hamstrings) and stretching, resulted in significantly larger increases in active SLR compared to heating
and stretching, or stretching alone.37 Burke et al.35 showed an opposite, but non-significant effect, and cooling was
associated with smaller increases hamstring flexibility, in comparison to heating or stretching alone.
Two small studies9,34 focused on how local temperature influences tissues passive resistance to movement. Both used
a force transducer to determine the resistance of the triceps surae muscle and tendon unit, during either passive knee
flexion34 or ankle PF.9 One of the studies34 used EMG to ensure that active muscle contraction did not confound the
outcomes. In both cases, the passive resistance increased after cooling, whereas heating or control interventions had
little effect. In contrast, a cross over study36 found that temperature did not affect knee joint laxity, based on anterior tibial
displacement during arthrometer testing (KT 1000).
Proprioception
Clinical assessment of proprioception involved measurement of: joint positional sense (JPS), or force perception. The
effects of cooling on JPS were mixed. Three observational studies noted no effect on single plane JPS at the knee22 or multiplane JPS at the shoulder joint.26,27 In contrast, two studies found that 15 minutes of cooling significantly decreased active
JPS at the ankle23 and knee25 using randomised cross over and observational designs respectively. One observational24
reported that cooling did not significantly affect participants ability to discriminate between different levels of weight
resistance during open chain knee extension.
Functional performance
A number of studies considered the effect of cooling immediately prior to undertaking various types of functional, or
sports specific tasks. One study42 focused on low level function; 30 minutes of ice pack application on the ankle joint, did
not affect lower extremity kinetics and kinematics in the closed chain during a semi-recumbant stepping exercise. Many
others have considered the effects of joint cooling prior to higher level functional activity, however the results are variable.
Miniello et al.31 concluded that CWI of the entire lower leg did not impair ankle stabilization following landing from a
jump; similarly, others found that ice and compression over the ankle, knee or both, did not alter the vertical ground
reaction forces associated with landing from a double30 or single leg14 jump. Others have found conflicting evidence
based on measures of lower limb power and speed.28,29,32,33 3 minute ice treatments32 had little effect on lower limb
functional performance; however increasing the application duration to 10 minutes, decreased performance on vertical
jump and shuttle sprint. Using similar methods, but with a longer application time of 20 minutes, others,28,29,33 also noted
cold induced decreases in athletic performance based on vertical jump height;28,29,33 shuttle sprint times28,33 and ground
reaction forces during landing.29 There was further evidence from an observational study that 20 minutes of shoulder joint
cooling, significantly decreased throwing accuracy.27

Secondary outcomes
Sub-maximal strength
Two small cross over studies measured how cooling affected the accuracy and variation of sub-maximal force production
[at a predetermined % of subjects maximum voluntary contraction (MVC)] at the wrist or hand. In each case, force
production was measured using a force gauge or hand held dynamometer, with visual feedback.10,11 In each case, 10-15
minute CWIs were compared to either heat11 or no treatment.10 Rubley et al.10 found no differences between groups in
the accuracy and variation when reproducing 10%, 25% and 40% of thumb-index pincer MVC. Similarly, Geurts et al.11
reported that force control at 25% and 50% of first dorsal interossei MVC was unaffected by local temperature. Again
limitations across both these studies include a lack of randomisation, and little, to no wash out time between treatment
interventions.
Muscle recruitment
The effect of cooling on local muscle recruitment has been studied using various outcome measures. Three studies4,12,17
focused on the effect of ankle joint cooling on soleus muscle recruitment based on H reflex recordings. There were trends
from each study that cooling resulted in increased muscle excitability (based on H-reflex facilitation) in one case a 30
minute treatment resulted in a 15% increase in Hmax/Mmax ratio.4 Similar findings were noted with knee joint cooling,
based on quadriceps central activation ratio (CAR) in healthy subjects; a 20 minute ice pack treatment significantly
increased CAR for up to 25 minutes after treatment, compared to control.18
A follow up study by Krause et al.13 used surface EMG to determine the amplitude and frequency of wrist extensor muscle
activity at 30%, 50% and 85%, of MVC, at different tissue temperatures. Skin temperature varied between 27C and 37C;
however this did not affect EMG amplitude. In contrast, EMG frequency decreased with tissue cooling at all levels of force
production. Based on normalised EMG readings, CWI significantly decreased peroneus longus activity during a single leg
jump landing;31 this returned to baseline after a 5 minute re-warming period. Others14 found little change in EMG muscle
activity in a number of muscles of the lower limb after 30 minutes of cooling on the knee joint.
Kinugasas group undertook two similar studies15,16 but they cooled muscle directly, and measured muscle firing patterns
using novel outcome. MRI images of the quadriceps muscles were taken before and after an exercise task. Using a
cross over design, exercise was undertaken either with or without short periods of pre-cooling of the vastus lateralis
(VL)15 or vastus medialis (VM).16 Analysis of T2 weighted MRI scans with imaging software showed that exercise resulted
in increased signal in all quadriceps muscle compartments, under control (no pre-cooling) conditions. Similar exercise
induced increases were found in the VL, VM and rectus femoris (RF) compartments when pre-cooling was used, but
the change in T2 uptake between pre and post exercise images was significantly higher in the vastus intermedialis (VI)
compartment. The authors concluded that this related to a cold induced activation of VI, resulting in increased metabolic
activity. Of note, the time between conditions ranged from 1 week,16 to just a few hours.15
Nerve conduction velocity (NCV)
Cooling skin temperatures to 10C, decreased NCV by 33% (to 22m/s), whereas there were no changes within an
untreated control group.19 Two smaller studies6,9 also found cooling has a similar effect on muscle fibre conduction
velocity; cooling increased the time between the onset of quadriceps muscle activity (based on EMG) and force production
(extension at the knee joint). The largest detriments were noted at muscle temperatures below 32C,6 interestingly one of
the studies found similar patterns with heating.9
Reflex pathways
In a different approach, two RCTs20,21 cooled the lateral ankle region, before monitoring the reflex response to sudden
ankle perturbation. Interestingly, both found that the magnitude and speed of peroneal muscles response to sudden joint
movement was not affected by ankle joint cooling.

Injured models
Only three studies used an injured model. Kernozek et al.40 measured postural sway on a force platform, before and after
a 20 minute CWI, using a small sample of participants post acute ankle sprain. Cooling significantly increased sway and
values remained higher than baseline for up to 20 minutes after treatment cessation. Three others38,39,41 used models
of acute injury based on laboratory induced knee joint effusion. Hopkins et al.41 measured H Reflex using percutaneous
simulation of the femoral nerve, and surface EMG of the VM. H reflex measurements were decreased immediately after
joint distension; 20 minutes of knee joint cooling significantly increased H reflex compared to control. Similarly, Rice et
al.38 showed that knee joint distension decreased quadriceps EMG amplitude, muscle fibre conduction velocity (MFCV),
and peak torque. Ice application to the knee joint subsequently increased MFCV and peak torque in comparison to
untreated controls, and there were further trends that cooling restored EMG amplitude (based on the root mean squares)
closer to normal values (90% of baseline), in comparison to controls, which remained at 75% of baseline. A larger study
by Hopkins et al.39 also compared ice to untreated control, but they measured quadriceps peak power, in addition to EMG
muscle activity and peak force. Again, peak torque, power and VL activity declined more in untreated knees compared to
the cryotherapy group. There were no differences between groups based on integrated EMG outcomes.

References
1. Higgins JPT, Altman DG (editors). Chapter 8: Assessing risk of bias in included studies. Section 8.5. In: Higgins JPT, Green S
(editors), Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.2 [updated September 2009]. Available from
www.cochrane-handbook.org.
2. Kimura IF, Thompson GT, Gullick DT. The effect of cryotherapy on eccentric plantar flexion peak torque and endurance. Journal of
Athletic Training 1997;32(2):124-126.
3. Hatzel BM, Kaminski TW. The effects of ice immersion on concentric and eccentric muscle performance in the ankle. Isokinetics and
Exercise Science 2000; 8: 103-107.
4. Hopkins JT, Stencil R. Ankle cryotherapy facilitates soleus function. Journal of Orthopaedics and Sports Physical Therapy
2002a;32(12):622-7.
5. Borgmeyer JA, Scott BA, Mayhew JL. The effects of ice massage on maximum isokinetic-torque production. Journal of Sports
Rehabilitation 2004; 13:1-8.
6. Zhou S, Carey MF, Snow RJ, Lawson DL, Morrison WE. Effects of muscle fatigue and temperature on electromechanical delay.
Electromyogr Clin Neurophysiol. 1998;38(2):67-73.
7. Douris P, McKenna R, Madigan K, Cesarski B, Costiera R, Lu M. Recovery of maximal isometric grip strength following cold
immersion. Journal of Strength and Conditioning Research. 2003;17(3):509-13.
8. Nosaka K, Sakamoto K, Newton M, Sacco P. Influence of pre-exercise muscle temperature on responses to eccentric exercise.
Journal of Athletic Training. 2004 Jun;39(2):132-137.
9. Kubo K, Kanehisa H, Fukunaga T. Effects of cold and hot water immersion on the mechanical properties of human muscle and
tendon in vivo. Clin Biomech (Bristol, Avon). 2005;20(3):291-300.
10. Rubley MD, Denegar CR, Buckley WE, Newell KM. Cryotherapy, sensation and isometric force variability. Journal of Athletic Training.
2003;38(2):113-119.
11. Geurts C, Sleivert GG, Cheung SS. Temperature effects on the contractile characteristics and sub-maximal voluntary isometric force
production of the first dorsal interosseus muscle. Eur Journal of Applied Physiology. 2004;91(1):41-5. Epub 2003 Sep 24.
12. Krause AB, Hopkins JT, Ingersoll CD, Cordova ML, Edwards JE. The relationships of ankle temperature during cooling and rewarming
to the human soleus H Reflex. Journal of Sports Rehabilitation 2000;9(3):253-292.
13. Krause KH, Magyarosy I, Gall H, Ernst E, Pongratz D, Schoeps P. Effects of heat and cold application on turns and amplitude in
surface EMG. Electromyogr Clin Neurophysiol. 2001;41(2):67-70.
14. Hart JM, Leonard JL, Ingersoll CD. Single leg landing strategy after knee joint cryotherapy. Journal of Sport Rehabilitation 2005;
14: 313-320.
15. Kinugasa R, Yoshida K, Watanabe T, Kuchiki K, Horii A. Skin cooling alters the activation patterns of difference heads of the
quadriceps. Canadian Journal of Applied Physiology 2005a;30(2): 127-139.
16. Kinugasa R, Yoshida K, Horii A. The effects of ice application over the vastus medialis on the activity quadriceps femoris assessed
by muscle function magnetic resonance imaging. Journal of Sports Medicine and Physical Fitness. 2005b;45(3):360-4.
17. Palmieri-Smith RM, Leonard-Frye JL, Garrison CJ, Weltman A, Ingersoll CD. Peripheral joint cooling increases spinal reflex excitability
and serum norepinephrine. International Journal of Neuroscience. 2007;117(2):229-42.
18. Pietrosimone BG, Ingersoll CD. Focal knee joint cooling increase the quadriceps central activation ratio. Journal of Sports Science.
2009a;27(8):873-9.
19. Algafly AA, George KP. The effect of cryotherapy on nerve conduction velocity, pain threshold and pain tolerance. British Journal of
Sports Medicine 2007;41(6):365-9;
20. Hopkins JT, Hunter I, McLoda T. Effects of ankle joint cooling on peroneal short latency response. Journal of Sports Science and
Medicine 2006a;5:333-339.

21. Berg CL, Hart JM, Palmieri-Smith R, Cross KM, Ingersoll CD. Cryotherapy does not affect peroneal reaction following sudden
inversion. Journal of Sports Rehabilitation. 2007;16(4):285-94.
22. Thieme HA, Ozmun JC, Ingersoll CD, Knight KL. Cooling does not affect knee proprioception. Journal of Athletic Training
1996;31(1):8-11.
23. Hopper D, Whittington D, Davies J. Does ice immersion influence ankle joint position sense. Physiotherapy Research International.
1997;2(4):223-36.
24. Tremblay F, Estephan L, Legendre M, Sulpher S. Influence of local cooling on proprioceptive acuity in the quadriceps muscle. Journal
of Athletic Training. 2001;36(2):119-123.
25. Uchio Y, Ochi M, Fujihara A, Adachi N, Iwasa J, Sakai Y. Cryotherapy influences joint laxity and position sense of the healthy knee
joint. Archives of Physical Medicine and Rehabilitation. 2003;84(1):131-5.
26. Dover G, Powers ME. Cryotherapy does not impair shoulder joint position sense. Archives of Physical Medicine and Rehabilitation.
2004;85(8):1241-6
27. Wassinger CA, Myers JB, Gatti JM, Conley KM, Lephart SM. Proprioception and throwing accuracy in the dominant shoulder after
cryotherapy. Journal of Athletic Training. 2007;42(1):84-9.
28. Cross KM, Wilson RW, Perrin DH. Functional performance following an ice immersion to the lower extremity. Journal of Athletic
Training. 1996;31(2):113-6.
29. Kinzey SJ, Cordova ML, Gallen KJ, Smith JC, Moore JB. The effects of cryotherapy on ground reaction forces produced during
functional tasks. Journal of Sport Rehabilitation 2000;9,3-14.
30. Jameson AG, Kinzey SJ, Hallam JS. Lower extremity joint cryotherapy does not affect vertical ground reaction forces during landing.
Journal of Sports Rehabilitation 2001;10:132-142.
31. Miniello S, Dover G, Powers M, Tillman M, Wikstrom E. Lower leg cold immersion does not impair dynamic stability in healthy
women. Journal of Sport Rehabilitation 2005;14: 234-247.
32. Fischer J, Van Lunen BL, Branch JD, Pirone JL. Functional performance following an ice bag application to the hamstrings. Journal
of Strength and Conditioning Research. 2009;23(1):44-50.
33. Richendollar ML, Darby LA, Brown TM. Ice bag application, active warm-up, and 3 measures of maximal functional performance.
Journal of Athletic Training 2006;41(4):364-70.
34. Muraoka T, Omuro K, Wakahara T, Muramatsu T, Kanehisa H, Fukunaga T, Kanosue K. Effects of muscle cooling on the stiffness of
the human gastrocnemius muscle in vivo. Cells Tissues Organs. 2008;187(2):152-60. Epub 2007 Oct 15.
35. Burke DG, Holt LE, Rasmussen R, MacKinnon NC, Vossen JF, Pelham TW. Effect of hot or cold water immersion and modified
proprioceptive neuromuscular facilitation flexibility exercise on hamstring length. Journal of Athletic Training 2001;39(1): 16-19.
36. Benoit TG, Martin DE, Perrin DH. Hot and cold whirlpool treatments and knee joint laxity. Journal of Athletic Training 1996; 31(3):
242-244.
37. Brodowicz GR, Welsh R, Wallis J. Comparison of stretching with ice, stretching with heat, or stretching alone on hamstring flexibility.
Journal of Athletic Training. 1996;31(4):324-7.
38. Rice D, McNair PJ, Dalbeth N. Effects of cryotherapy on arthrogenic muscle inhibition using an experimental model of knee swelling.
Arthritis Rheumatology. 2009;61(1):78-83.
39. Hopkins JT. Knee joint effusion and cryotherapy alter lower chain kinetics and muscle activity. Journal of Athletic Training. 2006b
Apr-Jun;41(2):177-84.
40. Kernozek TW, Greany JF, Anderson DR, Van Heel D, Youngdahl RL, Benesh BG, Durall CJ. The effect of immersion cryotherapy on
medial-lateral postural sway variability in individuals with a lateral ankle sprain. Physiotherapy Research International.
2008;13(2):107-18.
41. Hopkins J, Ingersoll CD, Edwards J, Klootwyk TE. Cryotherapy and transcutaneous electric neuromuscular stimulation decrease
arthrogenic muscle inhibition of the vastus medialis after knee joint effusion. Journal of Athletic Training. 2002b;37(1):25-31.
42. Hopkins JT, Adolph JT. Effects of joint cryotherapy on lower chain function. Clinical Kinesiology 2003;57:42-48.

Subject / inclusion
criteria

Pre test Intervention

N=22 healthy
11 male, 11 female
Mean age 23.8 (+/3.5 yrs)

N=20 healthy
Mean age: 19.6 (+/1.3 yrs)

N=30 healthy
16 male, 14 female
Mean age: 21 (+/3yrs)

N=11 healthy males

Mean age: 20.9 (+/1.1 yrs)

Kimura, 1997
(Randomised
cross over 7
to 14 days
between
conditions)

Hatzel, 2000

Hopkins,
2002a
(RCT;
allocation
concealment)

Borgmeyer,
2004
(Randomised
cross over 1
week between
conditions)

-Ice 10 min (ice massage to biceps)

-Control

-Ice 30 min (1.5L of crushed ice to lateral ankle,

final skin temperature approx. 16C)
-Control

-Ice 20 min (CWI to tibial plateau at 10C)

-Ice 30 min (CWI to mid thigh at 10C)

-Control

MUSCLE STRENGTH (CONCENTRIC / ECCENTRIC)

Author
(Study type,
Factors
lowering risk
of bias)

Figure 1

Note: trend towards increased torque in ice condition

Peak torque elbow flexion concentric (30 deg/sec; 1 rep every 2 minutes for 20 minutes)
I= C

Note: I increased peak torque by approximately 2% immediately after Rx, just under 5% 30
minutes post ice, and 2% 90 minutes post ice

Peak torque ankle PF

I>C (30, 60, 90 minutes)

Note: I = 6 Nm DECREASE (concentric DF at 60 deg/sec); and 5Nm DECREASE (concentric

DF at 120 deg/sec)

Ankle Peak torque

-Concentric/Eccentric - PF,INV,EV at 60 deg/sec and 120 deg/sec: no change
-Concentric DF (at 60 and 120 deg / sec): decrease post I*

Note: I = 380 Nm INCREASE (muscular endurance over 100 maximal reps)

Peak torque ankle PF eccentric (5 maximal repetitions at 30 deg/sec and 120 deg/sec)
I=C
Muscular endurance (total work during 100 maximal reps at 120 deg/sec)
I>C*

Outcomes / Results

Subject / inclusion
criteria

N=16 healthy
Mean age: 32 (+/-6.3
yrs), range: 20-42 yrs

N=20 healthy females

Mean age: 19.6 (+/2.8y)

N=8 healthy males

Mean age 26 (+/2yrs)

Douris, 2003
(Cross Over 4
days between
conditions)

Nosaka, 2004
(RCT)

Kubo, 2005
(Randomised
Cross Over
states
conditions were
on separate
days)

Rubley 2003
(Cross over 1
day between
conditions)

N=15 healthy
8 male, 7 female
Mean age 22 (+/3 yrs)

SUB-MAXIMAL STRENGTH

N=7 healthy
4 male, 3 female
Mean age: 20.3 (+/1.4 yrs)

Zhou, 1998)
(Observational)

MUSCLE STRENGTH (ISOMETRIC)

Author
(Study type,
Factors
lowering risk
of bias)

-Ice 15 min (CWI at 10C up to medial

epicondyle)
-Control

MVC PF
Post I < Pre I*

-Ice 30 min (CWI in 5C, to head of fibula)

-Heat 30 min (HWI in 42C to head of fibula)

Sub-maximal isometric force production of 10%, 25%, 40% of MVC

Accuracy (root mean square error of each rep) or variation (SD across 5 reps) of force
production was not affected by cooling.

Thumb and index finger pinch at 10%, 25% and 40% of MVC (following visual feedback).
Maintain for 30 secs (the first 5 secs discounted), 5 reps at each %.

Note: Ice decreased MVC by approx. 9Nm

Isometric force at elbow flexors

I=H

Note: Grip strength was decreased by 24.2% immediately after CWI, and remained 5.5%
lower 15 minutes after CWI cessation

Maximal isometric hand grip strength (hand dynamometer, 180 deg of shoulder flexion,
dominant arm, 3 second holds, verbal encouragement)
I = decrease* at all durations

Note: Decreasing muscle temperature from resting levels (37C) to 30C causes: a 125N
decrease in peak contraction force, a 0.5 m/s-1decrease in MFCV

Electromechanical delay (time lag between onset of EMG and muscle tension development)
Decrease post ice*
Peak contraction force
Decrease post ice*
MFCV
Decrease post ice*

Outcomes / Results

-Heat 15 min (microwave)

-Ice 15 min (biceps; ice cold water bag 10 cm x
15cm at 0C; towel between ice bag and skin;
biceps)

-Ice for 5, 10, 15 or 20 minutes (CWI elbow,

forearm and hand at 10C)

-Ice bag applied until thigh IM temperature

reached 30C

Pre test Intervention

N=10 healthy
Aged 23.9 (+/-2.3 yrs)

N=32 healthy
Aged 20-30 yrs
Amplitude and
frequency in surface
EMG from Before and
after cooling / heating

N=30 healthy
16 male, 14 female
Mean age: 21 (+/3yrs)

N=20 healthy
9 male, 11 female
Mean age 23.8 (+/3.6 yrs)

7 healthy males
Mean age: 24.9 (+/2.1 yrs)

Krause, 2000
(Observational)

Krause, 2001
(Cross over
time between
conditions
unclear)

Hopkins,
2002a
(RCT; allocation
concealment)

Hart, 2005
(Observational)

Kinugasa,
2005b
(Cross over - 1
week between
conditions)

-Ice 3 mins [ice pack over VM before initiating

quads exercise, and applied between each set
(60 secs), skin temperature: VM: 10.6 (2.4 C)]
-Control

-Ice 20 min (crushed ice to anterior knee with

elastic wrap)

-Ice 30 min (1.5L of crushed ice to lateral ankle,

final skin temperature approx. 16C)
-Control

-Heat 10 min (hot pack, skin temperature to

45C)
-Control

-Ice 20 min [5.6 L of crushed ice x 2 to medial

and lateral ankle] (skin temperature monitored)
-Ice 20 min ( ice pack, skin temperature
reduction to 28C)

-Ice 10 min [CWI in 8C, mean skin temp: 17.7

(1.3C)]
-Heat [Water perfusion patch and hot pack,
mean skin temp: 27.9 (2.3C)]

N=10
6 male, 4 female
Aged 23-41y

Geurts, 2004
(Cross over
no time
between
conditions?)

MUSCLE ACTIVATION PATTERNS

Pre test Intervention

Subject / inclusion
criteria

Author
(Study type,
Factors
lowering risk
of bias)

Quad activation patterns assessed by comparing mf MRI (T2 weighted) before and after knee
extension exercise (10 reps x 5 sets at 70% of 10 rep max, 1 min rest between sets)
% change in T2 value before and after exercise (using imaging software):
I=C in RF,VL,VM
I>C* in VI

Note: Approx. 15% increase in Hmax / Mmax ratio after 30 min cryotherapy treatment.
Forward jump with single leg landing
EMG activity gluteus medius, biceps femoris, vastus lateralis, medial gastrocnemius
Pre I = Post I (immediately, 15 and 30 min post treatment)

H Reflex (percutaneous stimulation of tibial nerve, surface EMG of soleus, normalised to peak
M-response)
-Change from baseline (H max / M Max ratio)
I>C (30*, 60* and 90* min)

Surface EMG at hand extensors during 20 second holds at 10%, 30%, 50 and 80% of MVC
(dynamometer pressure gauge)
EMG Amplitude
No differences at any level of muscle contractions
EMG Frequency
I<C at 30% or more of MVC

H reflex H Reflex (percutaneous stimulation of tibial nerve, surface EMG of soleus)

Inverse correlation between H reflex and skin temperature ie. Increased motoneurone-pool
recruitment during and after cooling

Sub-maximal force control at 25 and 50% of MVC (1st dorsal interosseous muscle)
No change in sub-maximal force control

Outcomes / Results

-Ice bag applied until thigh IM temperature

reached 30C

-Ice 30 min (CWI in 5C, to head of fibula)

-Heat 30 min (HWI in 42C to head of fibula)

N=7 healthy
4 male, 3 female
Mean age: 20.3 (+/1.4 yrs)

N=8 healthy males

Mean age 26 (+/-2yrs)

Zhou, 1998)
(Observational)

Kubo, 2005
(Randomised
cross over
- states
conditions
were on
separate days)

NERVE CONDUCTION VELOCITY

N=11 healthy
Mean age: 25 (+/5yrs)

Pietrosimone,
2009a
(Cross over
3 to 14
days between
conditions)

Note: Both ice and heat increased the time between muscle activation (on EMG) and force
development, by 25% and 19% respectively

Electromechanical delay [time lag between initiating calf muscle contraction (PF) on EMG and
force development]
Post I > Pre I*
Post H > Pre H*

Note: Decreasing muscle temperature from resting levels (37C) to 30C causes: a 125N decrease
in peak contraction force, a 0.5 m/s-1decrease in muscle fibre conduction velocity

Electromechanical delay (time lag between initiating quads contraction on EMG and force
development)
Decrease post ice*
Peak contraction force
Decrease post ice*
Muscle fibre conduction velocity
Decrease post ice*

Quadriceps CAR
-I>C (immed*, 10 and 25* mins post Rx)

H max, M max, H ratio

I>C (10*, 20* mins during Rx and 10*, 20* post Rx)
Plasma norepinepherine
I>C (10* during Rx)

-Ice 20 mins (ice bag anterior ankle)

- Control

N=22 healthy
Mean age: 25 (+/-14
yrs)

Palmieri-Smith,
2007
(Cross over- 2
weeks between
conditions)

-Ice 20 mins (knee joint)

-Control

Quad activation patterns assessed by comparing mf MRI (T2 weighted) before and after knee
extension exercise (10 reps x 5 sets at 60% of 10 rep max, 1 min rest between sets)
% change in T2 value before and after exercise (using imaging software
I=C VL
I>C in RF and Vm
I>C*in VI

-Ice 3 mins [(2 ice packs over VL before

initiating quads exercise, and applied
between each set (60 secs), skin temperature
; VL: 15.0 (1.8 C)
-Control

7 healthy males
Mean age: 25 (2.0 yrs)

Kinugasa,
2005a
(Randomised
cross over 2
hrs between
conditions)

Outcomes / Results

Pre test Intervention

Subject / inclusion
criteria

Author
(Study type,
Factors
lowering risk
of bias)

N=27 healthy subjects

14 males, 13 females
Mean age: 24 (+/2.7yrs)

Berg, 2007
(Randomised
cross over 24
hrs between
conditions)

N=37 healthy
16 male, 21 female
Mean age 23.4 (+/6.3yrs)

N=49 healthy
42 female, 7 male
Mean age: 19.4
(range: 17-28)

Thieme, 1996
(Randomised
cross over
states attended
two sessions)

Hopper, 1997
(Observational)

PROPRIOCEPTION

N=13 healthy
Mean age 23 (+/- 4
yrs)
N=27 healthy subjects
14 males, 13 females
Mean age: 24 (+/2.7yrs)

Hopkins,
2006a
(Randomised
cross over -1
week between
conditions)
Berg, 2007
(Randomised
cross over 24
hrs between
conditions)

-Ice 15 mins (CWI immersion at<5C,

5cm above the lateral maleolus, skin
temperature 15C)

-Ice 20 mins (2 ice packs (each 30.5 x

49cm, with 1160 of ice), 1 anteriorly,
covering an area 10 cm above and 10
cm below the patella, with the other
around the popliteal fossa),
-Control

-Ice 20 mins (ice bag with compression

over lateral ankle joint)
-Control

-Ice 30 min (1.5 L crushed ice with

elastic wrap; lateral ankle joint)
-Control 30 min (sham treatment: 1.5L
of dry clay with elastic wrap; lateral
ankle joint)
-Ice 20 mins (ice bag with compression
over lateral ankle joint)
-Control

-Ice (crushed ice applied until skin

temperature reduced to 10C foot)
-Control

N=23 healthy

Algafly, 2007
(RCT with
allocation
concealment)

REFLEX PATHWAYS

Pre test Intervention

Subject / inclusion
criteria

Author
(Study type,
Factors
lowering risk
of bias)

Note: Ice increased JPS error by 0.5 degrees

Active JPS error at ankle (40% and 80% of maximum inversion angle, at 42 degrees of PF)
Post I > Pre I*

Active JPS error at the knee (30, 60 and 90 degrees)

I=C

Peroneal reaction time with surface EMG.

I=C (immediately, 15 min and 30 min after removal of ice)

Sudden ankle perturbation to 25 degrees of inversion.

-Peroneal muscle activity (mV)
I=C (immediately, 15 min and 30 min after removal of ice)

Peroneal reaction time with surface EMG.

I=C (immediately, 15 min and 30 min after removal of ice)

Sudden ankle perturbation to 25 degrees of inversion.

-Peroneal muscle activity (mV)
I=C (immediately, 15 min and 30 min after removal of ice)

Ankle inversion perturbation (up to 28 degrees) while walking

Peroneal reaction time (Surface EMG peroneus longus)
I=C

Note: Cooling from baseline to 10C, decreased NCV by 33% (from approx. 35m/s to 22m/s)

NCV (foot)
Decreased* at skin temperatures reductions to 15C, continued to decreased further at 10C, no
changes in control

Outcomes / Results

Cross, 1996
(RCT)

N= 20 healthy
Mean age: 19.3 (+/-1.2
yrs)

-Ice 20 mins (CWI at 13C up to fibular head,

with water turbulence)
-Control

FUNCTIONAL / ATHLETIC PERFORMANCE

Wassinger,
N=22 healthy
2007
14 male, 8 female
(Observational) Mean age: 21.6 (+/2.4 yrs)

15 female:
Mean age: 20.7 (+/-1.4)

-Ice 20 mins (1.5kg ice cubes in 1.15 L bag,

secured with standardised elastic bandage to
centre of bag over the tip of Acromion)

-Ice 30 mins (1kg ice cubes in 20 x 25

cm bag, secured with standardised elastic
bandage, centre of bag over the tip of
Acromion; skin temperature reduced to
13.3C)
-Control

Dover, 2004
(Randomised
cross over - 48
hrs between
conditions)

N=30 healthy
15 male
Mean age: 23.7 (+/-5.5)

-Ice 15 mins (icing system, cooling pad at 4C,

skin temperature: 21.6C at knee)

Uchio, 2003
N=20 healthy
(Observational) 10 male, 10 female
Aged: 21-28 yrs

Pre test Intervention

-Ice 20 mins (Crushed ice with moist towel

barrier, to quadriceps muscle belly)

Subject / inclusion
criteria

Tremblay, 2001 N=20 healthy

(Observational) 14 male, 6 female
Mean age 22.1 (+/2.6yrs)

Author
(Study type,
Factors
lowering risk
of bias)

Note: Ice decreased vertical jump by approximately 2 cm, and decreased shuttle run by
approximately 0.16 seconds

Hop test
I=C
Vertical jump
I<C
Shuttle run
I<C

Active JPS error at shoulder (20 degrees FL, 90 degrees ABD)

Pre I=Post I

Active JPS error at shoulder (90% IR and ER)

I=C

Note: Active JPS error increased by 1.7 deg immediately after cooling (p=0.003), remained 0.9
deg worse than baseline for up to 15 minutes after cessation of ice

Active JPS error at knee (5 and 25 degrees)

Immed post I > Pre I*
15 mins post I > Pre I

Threshold for weight discrimination during open chain leg extension (kg)
Pre I = Post I

Outcomes / Results

Vertical GRF (peak, average, integrated and time to peak during 2 footed jump)
Pre=post (all groups)
I (ankle) = I (knee) = I (ankle/knee) = C

Forward jump with single leg landing

GRF; knee joint angle
Pre I = Post I (immediately, 15 and 30 min post treatment)

-Ice 20 mins (crushed ice pack to ankle)

-Ice 20 mins (crushed ice pack to knee)
-Ice 20 mins (crushed ice packs to ankle and
knee)
-Control (no ice)

-Ice 20 min (crushed ice to anterior knee with

elastic wrap)

-Ice 20 mins (CWI at 13-15C up to tibial

tuberosity)

N=10 healthy
Mean age: 22.4 (+/1.26 yrs)

N=20 healthy
9 male, 11 female
Mean age 23.8 (+/3.6 yrs)

N=17 healthy females

Mean age: 20.9 (+/1.1 yrs)

N=22 healthy
14 male, 8 female
Mean age: 21.6 (+/2.4 yrs)

Hart, 2005
(Observational)

Miniello, 2005
(Observational)

Wassinger,
2007
(Observational)

-Ice 20 mins (1.5kg ice cubes in 1.15 L bag,

secured with standardised elastic bandage to
centre of bag over the tip of Acromion)

-Functional throwing performance index (number of throws to hit a target and number of throws
in 30 secs)
Post I < Pre I*
-JPS
Post I=Pre I

Note: EMG returned to baseline 5 minutes after cooling

EMG activity (Preparatory and reactive) counter movement jump

-Tibialis anterior
Preparatory activity increased immediately post I*
-Peroneus longus
Preparatory and reactive activity decreased immediately post I*
Time to stabilisation
Pre I = Post I

Note: a 0.04% difference in vertical GRF was significant

One leg vertical jump (5 jumps x 5 sets) before, immediate after CWI
Peak vertical GRF
Post I > Pre I*
Average GRF
Post I =Pre I
Vertical impulse
Post I<Pre I*

Jameson, 2001
(Cross over 24hrs between
conditions)

-Ice 20 min (CWI at 10C up to patella)

N=15 healthy
7 male, 8 female
Mean age: 22.3 (+/2.1 yrs)

Outcomes / Results

Kinzey, 2000
(Observational)

Pre test Intervention

Subject / inclusion
criteria

Author
(Study type,
Factors
lowering risk
of bias)

-Ice 3 min (x2 ice bags (28 x 46cm) with

cubed ice, hamstring muscle belly, secured
with plastic wrap)
-Ice 10 min (x2 ice bags (28 x 46cm) with
cubed ice, hamstring muscle belly, secured
with plastic wrap)
-Control

N=42 healthy
25 female
Mean age 22 (+/0.5y)
17 male
Mean age 23 (+/0.5y)

N=24 healthy males

Mean age: 21.3 (+/3.3 yrs)

Fischer, 2009
(Cross over 1
day between
conditions)

Richendollar,
2009
(Randomised
cross over - 24
hrs between ice
conditions)

Muraoka, 2008
(Observational)

N=6 healthy males

Mean age: 27 (+/4yrs)

-Ice 60 min (CWI at 5-8C, lower leg, IM

temperature decrease by 5.8)

RANGE OF MOVEMENT / MECHANICAL PROPERTIES

-No ice followed by 20 minute rest

-No ice followed by 20 minute warm up
-Ice 20 min (1.4 kg of crushed ice in plastic
bag, secured with compression wrap over
anterior thigh) followed by 20 minute rest
-Ice 20 min (1.4 kg of crushed ice in plastic
bag, secured with compression wrap over
anterior thigh) followed by 20 minute warm
up

Pre test Intervention

Subject / inclusion
criteria

Author
(Study type,
Factors
lowering risk
of bias)

Note; Ice increased passive TA force by 19%, and increased passive tissue stiffness by 2N/mm; US
confirmed no change in the muscle length.

Outcomes monitored during passive knee extension (verified passive with EMG) from 90 to 0
degrees
% increase in passive TA force (via force transducer)
Post I>Pre I*
Gastroc stiffness (via force transducer)
Post I > Pre I*
Fasicle length (US)
Increased in both groups: I=C

Note: When no warm up is undertaken (after icing), ice produces an approximate decrease
of:1.7cm to vertical jump, and an approximate increase of 0.23 seconds onto shuttle run, and 40
yard sprint times.

Single leg vertical jump

No warm up: I < No I*
Warm up: I< No I
Shuttle run agility
No warm up: I>No I*
Warm up: I>No I
40 yard sprint
No warm up: I>No I*
Warm up: I>No I

Note: Ice slowed down shuttle run by approximately 0.2 seconds, and decreased single leg vertical
jump by 1cm

-Co-contraction (side shuffle test with band resistance)

Post control > pre control* (20 minutes post treatment)
-Shuttle run
Post I 10 min > Pre I 10 min* (at 10 and 20 mins post treatment)
-Single leg vertical jump
Pre I 10 min > Post I 10 min (immediately post treatment)

Outcomes / Results

-Ice 30 min (CWI in 5C, to head of fibula)

-Heat 30 min (HWI in 42C to head of fibula)

-Ice 10 min (CWI in 8C up to gluteal fold)

then PNF training
-Hot 10 min (HWI in 44C up to gluteal fold)
then PNF training
-Control 10 min (standing) then PNF training

N=8 healthy males

Mean age 26 (+/2yrs)

N=45 healthy
21 female, 24 male
Age range: 18-25 yrs

N=24 healthy
athletes
Mean age: 20.7 (+/1.2 yr)

N=15 healthy
8 male, 7 female
Mean age: 22.8 (+/2.5yr)

Kubo, 2005
(Randomised
cross over
- states
conditions were
on separate
days)

Burke, 2001
(RCT)

Brodowicz,
1996
(RCT)

Benoit, 1996
(Cross over - 1
day between
conditions)

-Ice 20 min (CWI at 15C, 4 inches a

bove patella)
-Heat 20 min (HWI at 40C,
4 inches above patella)
-Control (no immersion)

-Control 20 minutes stretching protocol

-Ice during 20 minutes stretching (ice bags
secured with an elastic wrap to the posterior
thigh)
-Heat during 20 minutes stretching (hot packs
secured with an elastic wrap to the posterior
thigh)
All groups: both legs

Pre test Intervention

Subject / inclusion
criteria

Author
(Study type,
Factors
lowering risk
of bias)

Anterior tibial displacement (KT 1000 arthrometer at 89N or maximal force)

No differences between groups

Note: Icing during stretching resulted in a 10 degree greater increase in flexibility, compared to
stretching alone and heat and stretching

Increase in active SLR (sum of right and left sides in degrees)

I>heat=control*

Note: 5 days of PNF stretching alone increased hamstring length by 26 degrees; there were no
differences when muscle heating was used prior to stretching. When ice was applied prior to PNF
stretching, hamstring length was increased by 23.5 degrees over a 5 day period.

No differences between groups

Conditions were undertaken prior to PNF training (every day for 5 days)
Hamstring length (degrees)
Day 5 I >Day 1 I
Day 5 H > Day 1H
Day 5 C > day 1 C

Note: Ice increased passive torque around the ankle joint by 1Nm, whereas with heat there was a
decrease of 2.2Nm

Passive ankle joint movement (5 dg/sec) in prone, from 15 degrees PF to 30 degrees DF

Passive torque
Post I>Pre I
Post H<Pre H*

Outcomes / Results

Subject / inclusion
criteria

Pre test Intervention

Postural sway variability (medial to lateral)

Pre I<Post I (immediately*, 10 min*, 20 min* post I)

H reflex [amplitude / change (pre-post)]

I>C (15*,30*,45*,60* mins post injection)
I>TENS (45*, 60* mins post injection)

Ice 20 minute (CWI; 0-4C, stirred every few

minutes, immersion to 5-7cm below tibial
tuberosity)

-Ice 30 mins (1.5 L bag of crushed ice x 2,

over anterior and posterior knee)
-TENS
-Control

N= 15 Post grade 1
lateral ankle sprain
(past 4-7 days)
Aged: mean: 21.3
(=/- 3.54 ) 18-29 y

N=30 induced injury

(Experimental knee
joint effusion using
60 ml sterile saline)
19 male, 11female
Age 21.8 (2.4y)

Kernozek, 2008
(Observational)

Hopkins,
2002b
(RCT)

Note: I resulted in an increase in H reflex beyond baselines from 15-60 minutes.

confirmed no change in the muscle length.

EMG activity (during recumbent stepping) post injury

-VL: I>C at 30 min and 60 min post effusino
-No differences in integrated EMG activity
DECREASE in peak torque post injury
I<C at 30 min post effusion
DECREASE in Peak power post injury
I<C at 30 and 60 min post effusion

Note: POST INJURY:

-I=26 Nm higher peak torque
-I=0.5ms-1 higher MFCV
-I=15% higher EMG RMS

Decrease in quadriceps peak torque post injury (normalised to body mass (nm/kg x 100)
I<C*
Decrease in MFCV post injury (n=10 only)
I<C*
Decrease in EMG (RMS) post injury (% of baseline)
I<C

-Ice 30 min (1.5L of crushed ice directly onto

the anterior knee, secured with elastic wrap)
-Control 30 min (sham ice bag and elastic
wrap onto anterior knee)

-Ice 20 min (3 plastic bags of crushed ice,

around knee joint)
-Control

N=45 induced injury

(experimental knee
joint effusion using
55ml of saline)
26 male, 19 female
Mean age: 21 (+/- 2
yrs)

10 male, 6 female
I: mean age 36.5
(10.6); C: mean age
34.3 (11.7)

N=16 induced injury

(Experimental knee
joint effusion using
saline to 50mmHg
intra-articular
pressure)

Outcomes / Results

Hopkins,
2006b
(RCT)

Rice, 2009
(RCT)

RANGE OF MOVEMENT / MECHANICAL PROPERTIES

Author
(Study type,
Factors
lowering risk
of bias)

Summary of cooling durations used in studies assessing Primary outcomes (Key correlates of physical performance)

Figure 2

I: Ice
C: Control (note: control = no intervention unless otherwise indicated)
H: Heat
CWI: cold water immersion
HWI: hot water immersion
PF: plantar flexion
MVC: maximum voluntary contraction
H Reflex: Hoffman reflex
VM: Vastus Medialis
VL: Vastus Lateralis
VI: Vastus Intermedialis
RF: Rectus Femoris
mf MRI: muscle functional magnetic resonance imaginge
CAR: Central activation ratio
I/M: Intramuscular
JPS: joint positional sense
TA: tendo Achilles
US: Ultrasound
PNF: Proprioceptive Neuromuscular Facilitation
SLR: Straight Leg Raise
MFCV: Muscle Fibre Conduction Velocity
RMS: Root Mean Squares
GRF: Ground reaction force
In the outcome column, the greater than (>) and less than (<) symbols refer to the absolute differences between groups/conditions (eg. Ice vs Control).
In cases where comparisons were made within groups before and after conditions, Pre I = pre intervention value (before), and post I = post intervention value (after).
* indicates a significant difference between groups (time point is immediately after treatment unless otherwise indicated).

Chapter 6
Which components of PRICE are effective in the clinical management of acute
soft tissue injury?
What is known in this area: Protection, rest, ice, compression and elevation (PRICE), is synonymous with acute
soft tissue injury management. The original guidelines1 recommended that all components of PRICE should be
employed immediately after acute soft tissue injury, and gave specific recommendations for its practical use. Much
of the recommendations were based on expert consensus and there was little high quality empirical evidence in this
area.
Aim: To update the clinical evidence for using PRICE in the management of acute soft tissue injury.
Clinical question: Which components of PRICE are effective in the clinical management of acute soft tissue injury?
Objective: To review the current literature and determine which components of PRICE are effective in the clinical
management of acute soft tissue injury based on an injured human model?
PROTECTION/REST: There is moderate quality evidence that functional treatment is more effective than cast
immobilisation after ankle sprain for a range of important outcomes, but there are little long term differences.
There is also moderate evidence that this approach is most effective in less severe sprains. There is moderate
evidence that semi-rigid supports are more effective than elastic bandages, and low quality evidence that DTG
is ineffective in terms of short term recovery. There is conflicting evidence on the most effective form of external
support used in conjunction with functional treatment (semi-rigid, taping, lace up or focal compression). There is
very low quality evidence that 3 weeks immobilisation after primary shoulder dislocation, is similar to using a period
of immobilisation guided by patient discretion and comfort, in terms of long term re-injury rate. There is very low
quality evidence that a supervised active treatment intervention is effective an effective approach after closed elbow
dislocation in an active population. There is very low quality evidence that 24 hrs of immobilisation in end of range
flexion, followed by isometric exercise is an effective approach after quadriceps contusion in an active population.
ICE: Based on a closed soft tissue injury model, there is moderate evidence that cold therapy is effective at
decreasing short pain after acute ankle injury and general soft tissue contusion; there is low quality evidence that
shorter intermittent applications are sufficient to induce short term analgesia. Based on a post surgical model, there
is high quality evidence that cold therapy provides effective short term analgesia, but has little benefit in terms of
other critically important outcomes. Evidence from questionnaires and case studies confirm that cold therapy does
have the capacity to do harm; most cases involved short term complications with a full recovery. There were isolated
cases of permanent scarring and loss of function. There are few reports of adverse events within lab based and
clinical studies. The risk of inducing adverse events is low, if clear instructions for practical application are provided
and evidence based rationale considered.
COMPRESSION: There is very low quality evidence that compression is ineffective after muscle injury. There is
moderate evidence that semi-rigid supports are more effective than elastic bandages, and low quality evidence that
DTG is ineffective in terms of short term recovery. There is conflicting evidence on the most effective form of external
support used in conjunction with functional treatment.
ELEVATION: There is low quality evidence that elevation alone is marginally better than compression and elevation
in terms of decreasing ankle volume. There is also low quality evidence that limb volume returns to baseline levels
quickly after returning the extremity to a dependent position. There is also moderate evidence that tissue volume
(swelling) will increase after orthopaedic surgery, regardless of the magnitude of home based limb elevation.

Which components of PRICE are effective in the clinical management of acute soft tissue injury based
on a human model?
Protection, rest, ice, compression and elevation (PRICE), is synonymous with acute soft tissue injury management. The
original guidelines1 recommended that all components of PRICE should be employed immediately after acute soft
tissue injury. There was little high quality empirical evidence in this area eg. meta-analysis (MA), systematic review
(SR), randomised controlled trials (RCT), and much of the recommendations were based on expert consensus. The main
recommendations from the original guidelines1 were as follows:

Protection and rest should be used for at least 3 days post injury, with longer periods advised according to injury severity.
Only isometric type exercises should be undertaken during the acute stages. Compression should be applied immediately
after injury, with continuous use over the first 72 hours post injury only. Compression should be applied with uniform
pressure, from proximal to distal, with additional intermittent compression permitted. Elevate for as long as possible in the
first 72 hours, but avoid immediate restoration of the injured body part to a gravity dependant position. Compression and
elevation should not be undertaken simultaneously. Where possible ice should be applied in the immediate stages post
injury; the optimal protocol is chipped ice (with damp barrier) for durations of 20-30 minutes, repeated every 2 hours.
This protocol should be altered based on levels of body fat, and the presence of superficial nerve at the application site.
Athletes were advised not to return to competition immediately after applying ice.
Our primary aim was to update the original ACPSM guidelines.1 The objective of this chapter was to review of the current
literature to determine which components / or combinations of PRICE are clinically effective in the management of acute
soft tissue injury.
Methods
Search strategy
We searched the titles generated from the main search strategy outlined in Appendix Table 1-4. Relevant studies were
extracted, with exclusions made based on titles, abstracts or full text versions.
Inclusion criteria
Study type: Systematic review, randomised or observational studies (published since 1996).
Participants: Injured humans with acute soft tissue injury. We did not consider models using complete tendon rupture,
however no otehr restrictions were made on the type or severity of injury.
Intervention: Any component or combination or PRICE initiated within the acute phases of injury. Protection and rest were
defined as any intervention unloading, immobilising or modifying activity after injury the injured tissue. No restrictions
were made on mode, duration or dosage of ice, compression or elevation.
Outcomes: Any clinical based outcome from a study based on injured human participants was considered to be of critical
importance.
Comparison: Intervention could have compared with anything other than surgery.
Note: Individual studies were not considered if they were already included within an eligible systematic review or metaanalysis.
Grading of evidence
Five criteria were used to grade the quality of evidence in each sub-group:
- Study design
- Internal validity (Risk of bias)
- Consistency (across studies, in terms of the size and direction of effect estimates)
- Directness (external validity)
- Publication bias
All five criteria were assessed by two independent reviewers. Initially study designs were used to categorise evidence into
the following categories: high quality (MA or SR, RCT), moderate (CT) low quality (observational) or very low quality (other
study design). Evidence was then upgraded, or downgraded according to the remaining four criteria. Internal validity was
assessed using The AMSTAR3 scoring scale for meta-analyses or systematic reviews (Appendix Table 9). For all original
research, internal validity was assessed using the Cochrane risk of bias tool2 (Appendix Table 10); using the criterion:
sequence generation, allocation concealment, assessor blinding, incomplete outcome data, other (adequate detail
provided on PRICE intervention AND details of co-interventions). Blinding of participant or caregivers were not considered
based on the nature of the interventions. Consistency, directness and publication bias were assessed qualitatively by each
reviewer. In the event that evidence was downgraded (or upgraded); the rationale was provided (as detailed overleaf).

Evidence Summary
As we were potentially dealing with a large number of heterogenous studies, we created separate evidence summaries
according to the type of soft tissue affected, and treatment comparison. The following definitions were used as a final
quality grading:4

High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the
estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the
estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

RESULTS
Figure 1 provides a summary of included studies by primary intervention and injury type. Detailed study characteristics are
summarised in Figure 2. Risk of bias summary is provided in Figures 3 and 4. Excluded studies (with reasons) are listed in
Appendix Table 5. Confidence intervals are 95% unless otherwise stated.
Protection/Rest: Ankle sprain
Four SRs were eligible for inclusion.5-8 Three compared immobilisation with functional treatment after acute ankle
sprain;5,6,8 one compared various forms of functional management of ankle sprain.7 All reviews fulfilled at least 85%
of criterion on AMSTAR.3 This included stringent selection criteria, data extraction, and literature searches. All reviews
undertook and considered the scientific quality of included studies, and when applicable, an appropriate method of
meta-analysis was undertaken. None of the studies used graphical or statistical tests to consider the threat of publication
bias. Searching up to 1998, Pijnenburg et al.5 included 27 randomised and quasi randomised studies of acute ligaments
ruptures, with the majority confirming injury severity via arthrogram or stress radiograph. They classified 5 included
studies as high quality, and undertook separate sub-group analysis on this data. Note that the authors considered <3
weeks of cast immobilisation to be a functional treatment. Kerkhoffs and colleagues undertook two Cochrane reviews.6,7
Both reviews included acute injury to the lateral ankle, based on physical examination, stress radiograph or arthrogram.
Literature searching was undertaken up to 2000, and a total of 30 RCTs were included. The smallest and most recent
review,8 excluded any studies with >80% drop out, leaving 9 eligible RCTs. All four reviews5-8 pooled data for MA.
There were 6 relevant RCTs 9-14 (and one related protocol15), published after the search limits on the aforementioned
SRs. All undertook adequate sequence generation. In one study, allocation concealment was based on a predetermined
rotation system,9 and another was unclear;10 all others employed either off site concealment, or opaque sealed envelopes.
Blinding of outcome assessor was limited to one study.14 In all cases the intervention was clearly described, and although
co-interventions were used, they were deemed to be standardised across groups. Ankle sprain severity differed across
RCTs, they included: grade 1 or 2;9,11 or grades 2 and 3.10,12 One study13 which included all severities sub-grouped by
injury grade, using stratified randomisation. No formal grading was undertaken in Lambs study;14 however, if we consider
that their primary inclusion criteria was unable weight-bear for at least 3 days post injury, this might have been restricted
to more severe sprains.
Cast immobilisation vs Functional
SR

Pooling the results of 10 studies, Pijnenburg et al.5 found that functional treatment was associated with less time lost from
work (15 days (range 12-18) compared with casting (38 days (range 28-48); and significantly lower levels of residual pain
(RR: 0.67 CI: 0.5-0.9), and complaints of giving way (RR: 0.7 CI: 0.5-0.9) at 6 weeks. Similarly, Kerkoffs et al.6 found
a number of significant effects in favour of functional treatment. These included return to sport (n=5 studies pooled: RR
1.86 (CI 1.2; 1.9); days to return to sport [n=3 studies pooled; WMD: 4.8 days (95% CI: 1.5-8.3)], swelling [n=3 studies
pooled; RR: 1.7 (CI: 1.17-2.59)] and patient satisfaction [n=6 studies pooled; RR: 4.25 (CI: 1.17-2.59)]. No differences
were found between groups in terms of pain, giving way, re-injury. Jones and Amendola8 did not find any significant
difference based on pooled data, however there were trends that functional treatment was superior in terms of time to
return to work, subjective instability and re-injury rate. Notably the two reviews that undertook sub-group analysis on the
high quality studies (n=11;6 n=55) found no significant differences between groups.
RCT
In a large RCT14 10 days of POP immobilisation resulted in significantly better function up to 3 months post injury,
compared to DTG. There were similar patterns that casting was also superior to rigid bracing, however, the differences
were less pronounced. Beynonn et al.13 used a stratified approach, randomising by injury grade. They found no differences
in outcome between casting for 10 days, and functional treatment with a rigid brace, in participants with grade 3 (severe)
injury. However, functional treatment resulted in a significantly faster return to function and sporting activity in participants
with grade 2 ankle sprains. None of the studies found any significant differences between groups at 6-12 month follow
up.13,14
Evidence summary: There is moderate quality evidence that functional treatment is more effective than cast
immobilisation after ankle sprain for a range of important outcomes, but there are little long term differences
between the interventions. There is also moderate evidence that this approach is most effective in less severe
sprains. [The findings were deemed to be consistent across reviews for the majority of outcomes. Downgraded
from high to moderate quality due to: the lack of effect based on sub-group analysis on high quality studies
within two of the reviews/inconsistent findings from the large RCT by Lamb et al.14] Note: A controlled trial
by Samato et al.48 (high risk of bias) showed very low quality evidence that Cast immobilisation with weightbearing is less effective in more severe sprains].
Functional vs functional
A number of studies have evaluated the effect of using different methods of external support, to accompany early
mobilisation after sprain. Kerkhoffs et al.7 included two RCTs comparing functional treatment with elastic bandaging or
semi rigid support. Pooled results found a faster return to work in favour of semi-rigid based on dichotomous (RR: 4.3 CI:
2.4-6); and continuous scales (WMD: 9.6 days (6.3;12.9) and less complaints of giving way (RR: 8; CI: 1.03-62), all in
favour of the semi rigid bracing. Lace up braces were associated with less swelling when compared to elastic bandaging
(n=1; RR: 5.5 CI 1.7; 17.8), taping (n=2; RR: 4.07, CI 1.2; 13.7), and semi rigid braces (n=1; RR: 4.19; CI: 1.3; 13.98).
No differences were noted between taping and semi-rigid bracing (n=2 studies).
Three related RCTs10,12,14 found further trends that semi rigid bracing is superior to elastic bandaging. Lamb et al.14 found
that semi-rigid supports resulted in higher levels of subjective function at 3 months compared to DTG. A small study by
Boyce et al.12 found no differences between groups in terms of pain or swelling, however, rigid bracing did result in higher
levels of function at 10 days post injury. Leanderson et al.10 recorded a range of outcomes; however, the only significant
difference was that braced participants completed a figure of eight running test faster at 10 weeks. Furthermore, Watts
and Armstrong11 found that adding DTG to standard advice was of little additional benefit during recovery after grade 1
and 2 sprains. Follow up at 10 days also found that DTG was associated with higher levels of analgesic consumption. A
critical limitation was the 60% loss to follow up reported in this study. One small study9 compared two different methods
of focal compression, with a semi rigid support. No short term differences were found in pain, swelling or function.
Evidence summary: There is moderate evidence that semi-rigid supports are more effective than elastic bandages
[downgraded based on consistency of results across all outcomes], and low quality evidence that DTG is
ineffective in terms of short term recovery [downgraded due to risk of bias11]. There is conflicting evidence on
the most effective form of external support used in conjunction with functional treatment (semi-rigid vs taping
vs lace up vs focal compression). There is very low quality evidence (controlled study49, high risk of bias) that
weight-bearing mobilisation with semi-rigid support is superior to a walking cast.

Elevation vs elevation and compression

Tsang and colleagues16 measured the changes in ankle volume associated with 30 minutes of elevation, or combined
elevation and compression treatment, in twelve student subjects with acute ankle sprain. Both interventions significantly
decreased limb volume, with trends towards larger changes (10mL difference) in the elevation only group. Interestingly,
they also considered the effect of an immediate return to a gravity dependent position after treatment. In both groups,
limb volume returned to pre treatment values after 5 minutes of sitting.
Evidence summary: There is low quality evidence that elevation alone is marginally better than compression
and elevation in terms of decreasing ankle volume. There is also low quality evidence that limb volume returns
to baseline levels quickly after returning the extremity to a dependent position. [Downgraded based on risk of
bias and small number of studies (unable to judge for consistency)]
Soft tissue injury
Ice and compression
Much of the clinical literature on the effectiveness of ice and compression is based on postsurgical models. In a SR,17
only 6 out of 24 included RCTs used subjects with closed soft tissue injury. These studies generally had a number of
methodological limitations (mean PEDro score 4.3; range 3-8) and sample size ranged from 30-143. This review found
evidence from single, small RCTs that cold therapy is more effective that heat, contrast therapy and no treatment, in terms
of reducing swelling and pain after ankle sprain; but there is no evidence to suggest enhanced function or long term
outcome. Of note, the authors found wide variation in the mode and treatment dosage of cooling used across studies;
they note that often this was limited to between 20 and 90 minutes of cooling which may not be sufficient to achieve a
worthwhile clinical effect.
Two further RCTs have been added to the evidence base recently. A large RCT by Airaksinen and colleagues18 found
that 4 daily applications of a topical cooling gel, reduced pain in comparison to placebo gel. These results were based
on general minor soft tissue injuries and cannot be generalised to more moderate or severe injuries. Bleakley et al19
compared the effectiveness of two different cooling doses in a group of 89 participants with acute ankle sprain. An
intermittent 10 minute application was associated with lower levels of pain at week 1, in comparison to a standard 20
minute application. There were no differences between groups at 3 month follow up. One SR20 considered the effects of
compression after soft tissue injury; the included studies were all previously included within other reviews.6,7
Evidence summary: There is moderate evidence that cryotherapy is effective at decreasing short pain after
acute ankle injury and general soft tissue contusion. [Downgraded based small number of studies (unable to
judge consistency)]. There is low quality evidence that shorter intermittent applications are sufficient to induce
short term analgesia. [Downgraded based on risk of bias, and small number of studies (unable to judge
consistency)]
Post surgery
The remainder of recent RCTs have focused on the effectiveness of various combinations of cooling and compression after
post orthopaedic surgery (eg. ACL reconstruction, TKA, THA, CTR). Comparison across studies is greatly limited by clinical
heterogeneity. Bleakley et al.17 focused on a wide range of treatment comparisons, however few definitive conclusions
were found. Much of this related to the wide variations in terms of the injury type, treatment intervention and comparison
group. In addition, the internal validity of studies was low (3.4/10), based on PEDro. There was evidence from a single
RCTs that cold therapy is effective in terms of short term analgesia after minor knee surgery, but had no effect on function,
swelling or range of movement. The majority of orthopaedic studies compared ice and compression against, compression
only. The compression interventions normally consisted of a compressive cooling device filled with room temperature
water. Six out of eight studies found that ice and compression was no more effective than compression alone; two found
significantly larger decreases in short pain with ice and compression. The limitations of using an orthopaedic setting to
assess the effectiveness of cooling were noted by the authors; of particular concern is the use of post surgical dressing or
socks which risk mitigating the cooling effect. Only one study in this review included a measure of skin temperature under
the barriers, which was reported to be a modest 28C.
A related review21 focused on the effectiveness of cryotherapy after ACL reconstruction only. Using similar comparisons,
they pooled data from six studies using random effects modelling. They found cryotherapy had a significant effect over
placebo in terms of pain relief, but there were no overall differences for drainage or range of motion.

Further studies have been undertaken in this area since the publication of these SRs. Holmstrm and Hrdin22 compared
48 hours of basic analgesia (paracetamol), epidural anaesthesia or cryocuff after Uni-compartmental knee arthroplasty.
Sequence generation and allocation concealment were not described and blinding was not undertaken. Results found
no difference between groups in wound drainage, subjective pain, swelling, ROM and function. Both the epidural and
cryocuff groups needed significantly less additional pain relief (morphine) during the first 24 hours post surgery, compared
to the control. In a randomised controlled design, Smith et al.23 found little differences in swelling, ROM, bleeding or
pain, when using a commercial cooling systems, or a compression bandage over the first 24 hrs after TKR. There was no
allocation concealment or blinding of outcome assessment. A further limitation was that management of both groups was
similar after 24hrs, with both receiving cooling with an ice bag for up to 72 hrs post surgery. Using a similar population,
a controlled trial50 found significant differences in favour of a cooling group compared to compression only, in terms of
short term pain, ROM and post operative swelling.
There are few clinical studies of elevation after injury. We found one RCT24 measuring the effect of post operative hand
elevation, based on a population of n=43 participants after carpal tunnel decompression. One group received a normal
sling (arm by side with elbow flexed to 90 deg), and the other undertook high elevation with the hand above the heart.
Water volumetry found that both groups had increases in hand volume at five days post operatively. There were few
meaningful differences between groups, and the mean increase in the high elevation group was approximately 2 ml less
than in the sling control.
Evidence summary: There is high quality evidence that cold therapy provides effective short term analgesia,
but has little benefit in terms of other important clinical outcomes. There is also low quality evidence that tissue
volume (swelling) will increase after orthopaedic surgery, regardless of the magnitude of home based limb
elevation (Downgraded based on risk of bias, and small number of studies (unable to judge consistency). [Note:
Post surgical dressings could prevent adequate tissue cooling, and consequently, it is difficult to extrapolate
some of this evidence to a closed soft tissue injury model].
Other Evidence
Ankle sprain
A single case study47 on a female athlete with a grade 2 ankle sprain (clinical diagnosis) employed a range of PRICE
interventions coupled with progressive active rehabilitation and strengthening exercise. This approach resulted in a fast
return to sports (2 weeds) with excellent long term functional outcome at approximately 1 year.
Shoulder dislocation
Hovellius et al.25 focused on long term outcomes after primary shoulder dislocation. A fixed sling immobilisation time of
21 days, was compared with a control using immobilisation for any time up to 21 days. The duration of immobilisation
in the control was made at patients discretion and comfort; 87/104 patients in this group, continued with a sling for just
7 days. Of note, there were a number of protocol violators in the fixed time group ie. individuals who removed their sling
prior to the recommended 21 days. The primary outcome, injury reoccurrence, was recorded over a 25 year period. The
duration of immobilisation did not seem to a factor in the risk of re-injury, or requirement for surgical intervention. Of note
this was a mixed methods study, with only some of the recruitment centres using randomisation. In addition, allocation
concealment and intention to treat were not undertaken.
Elbow dislocation
In a case series design,26 20 military personnel with closed (simple) posterior elbow dislocation (reduced within 1 hour)
were treated with a supervised active treatment intervention, without slings or immobilisation. Co-interventions included:
compression, icing, elevation, electrical stimulation, active elbow and hand exercises were undertaken from day one,
with resistance added from day 2, and swimming from day 3. Maximum range of movement was achieved within 19
days (range 3-30) post injury, with all subjects reaching extension to within 5 of their injured side. No patients were lost
to follow up, and long term outcome was described as excellent, with minor subjective complaints. One re-dislocation
occurred during contact sports at 15 months.
Quadriceps contusion
Another prospective case series27 was undertaken on a military population, with acute quadriceps contusion. Inclusion
criteria which were, inability to continue participation in sport, and lack of pain free straight leg raise. Each of the 47
participants were braced in 120 degrees of knee flexion for 24 hours, followed by progressive isometric exercises. The
mean time for return to athletic activity was 3.5 days (range 2-5). Follow up between 3 and 6 months found 1 case of

Muscle injury
In a non randomised, controlled design,28 40 participants with acute muscle injury were treated with either immediate
compression (80mm/Hg), or control. The control group received rest and ice or in certain cases moderate compression
(40 mm/Hg) after 10-30 minutes post injury. There were no differences between groups in terms of subjective recovery,
range of movement, serum CK levels or ultrasonic findings.
Evidence summary: There is very low quality evidence that 3 weeks immobilisation after primary shoulder
dislocation, is similar to using a period of immobilisation guided by patient discretion and comfort, in terms of
long term re-injury rate. There is very low quality evidence that a supervised active treatment intervention is
effective an effective approach after closed elbow dislocation and ankle sprain in an active population. There is
very low quality evidence that 24 hrs of immobilisation in end of range flexion, followed by isometric exercise
is an effective approach after quadriceps contusion in an active population. There is very low quality evidence
that compression is ineffective after muscle injury [Very low quality based on observational design; risk of bias
and small number of studies (unable to judge consistency)]
Adverse effects
Characteristics of studies reporting adverse effects are provided in Figure 5. Two recent surveys focused on practitioners
experiences of adverse events relating to electrophysical agents, with a primary focus on cryotherapy. The target populations
were athletic trainers29 and physiotherapists30 with respective response rates of 30% to 72%. One report estimated that
cryotherapy accounted for a larger number of complications than heat, electrical stimulation and therapeutic exercise.29
The severity of these events varied but included skin burns, frostbite, intolerance or allergy. A smaller survey of private
practitioners estimated that on average one physiotherapist observed a cold related adverse incident every 5 years.30
In conjunction, we found a number of case reports of adverse events associated with cryotherapy; the most common
was skin burn/damage,30-36 with others reporting nerve damage,37,38 cold urticaria39 and a compartment syndrome.40
It is difficult to determine definitive risk factors for such adverse events based on case study evidence. Notwithstanding
this, there were a number of common clinical circumstances across reports which merit further investigation. Perhaps the
most frequent trend was that cryotherapy was usually applied continuously or for prolonged periods,33-35,38 and there
was one report of a patient falling asleep during treatment.32 Another common event was patients compressing the
cooling modality between their lower leg and a table; indeed this resulted in severe skin damage or nerve palsy with short
application times of 20-30 minutes.30,31,37 In conjunction with previous reports, both incidences of nerve palsy occurred
after cooling around a superficial nerve eg. common peroneal37 or around foot and ankle.38 In both cases symptoms were
still present for between 6 and 12 months. One survey noted that cold allergy was the most commonly observed adverse
event associated with cryotherapy.29 We found just one report of a cold allergy which occurred during an experimental
study;39 this was consistent with cold urticaria (wheal formation and skin discolouration), but symptoms had fully resolved
within 48 hours.
Cooling was applied directly to the skin in three cases,31,37,38 and two 33,34 did not specify if a barrier was used at the cooling
interface. Previous surveys highlight that in clinical practice it is common to use a barrier between the cooling surface
and the skin.1 Much of the rationale is to minimise the risk of adverse effects, and in particular, preventing excessive
tissue cooling. Dry or thick dressings are associated with excessive skin insulation, and can limit skin temperatures to just
27C41 even after applications of up to one hour. In contrast, damp barriers may do little to affect cooling efficiency when
compared to direct application onto the skin.42,43 We have previously discussed that optimal skin temperature reductions
can be achieved if our primary objective is to achieve analgesia. Furthermore, there is considerable evidence that shorter
applications times are sufficient. We must also consider that skin burns result from excessive cooling; cell death can occur
at a threshold of around minus 10C.44 Based on Chapter 245 such excessive tissue temperatures seem unlikely with
crushed ice, particularly in cases where it is melting during treatment and the skin/cooling interface remains at a constant
temperature of around 0C. Perhaps the mechanism for producing excessively skin temperature reductions relate to
concomitant mechanical compression and reduction in superficial blood flow. Pragmatically this is also more likely with
prolonged or uncontrolled treatment durations, or in circumstances of patient or practitioner error.
Note: Evidence from questionnaires and case studies confirm that cold therapy does have the capacity to do
harm. In most cases this was short term and a full recovery was made; there were isolated cases of permanent
scarring and loss of function. There are few reports of adverse events within lab based and clinical studies. The
risk of inducing adverse events is low, if clear instructions for practical application are provided and evidence
based rationale considered.

References
1. Kerr KM, Daley L, Booth L for the Association of Chartered Physiotherapists in Sports Medicine (ACPS M). Guidelines for the
management of soft tissue (musculoskeletal) injury with protection, rest, ice, compression and elevation (PRICE) during the first
72 h. London: Chartered Society of Physiotherapy, 1998.
2. Higgins JPT, Altman DG (editors). Chapter 8: Assessing risk of bias in included studies. Section 8.5. In: Higgins JPT,
Green S (editors), Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.2 [updated September 2009].
Available from www.cochrane-handbook.org.
3. Shea BJ, Hamel C, Wells GA, Bouter LM, Kristjansson E, Grimshaw J, Henry DA, Boers M. AMSTAR is a reliable and valid measurement
tool to assess the methodological quality of systematic reviews. Journal of Clinical Epidemiology. 2009 Oct;62(10):1013-20. Epub
2009 Feb 20.
4. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. for the GRADE Working Group. GRADE: an emerging
consensus on rating quality of evidence and strength of recommendations. British Medical Journal. 2008;336(7650):924-6.
5. Pijnenburg AC, Van Dijk CN, Bossuyt PM, Marti RK. Treatment of ruptures of the lateral ankle ligaments: a meta-analysis. Journal
of Bone and Joint Surgery (American). 2000 Jun;82(6):761-73.
6. Kerkhoffs GM, Rowe BH, Assendelft WJ, Kelly K, Struijs PA, van Dijk CN. Immobilisation and functional treatment for acute lateral
ankle ligament injuries in adults. Cochrane Database of Sysematic Reviews 2002;3:CD003762.
7. Kerkhoffs GM, Struijs PA, Marti RK, Assendelft WJ, Blankevoort L, van Dijk CN. Different functional treatment strategies for acute
lateral ankle ligament injuries in adults. Cochrane Database Systematic Reviews 2002;3:CD002938.
8. Jones MH, Amendola AS. Acute treatment of inversion ankle sprains: immobilization versus functional treatment. Clinical
Orthopaedics and Related Research 2007;455:169-72.
9. Guskiewicz KM, Riemann BL, Onate JA. Comparison of 3 methods of external support for management of acute lateral ankle
sprains. Journal of Athletic Training. 1999 Jan;34(1):5-10.
10. Leanderson J, Bergqvist M, Rolf C, Westblad P, Wigelius-Roovers S, Wredmark T. Early influence of an ankle sprain on objective
measures of ankle joint function. A prospective randomised study of ankle brace treatment. Knee Surgery, Sports Traumatology,
Arthroscopy. 1999;7(1):51-8.
11. Watts BL, Armstrong B. A randomised controlled trial to determine the effectiveness of double Tubigrip in grade 1 and 2 (mild to
moderate) ankle sprains. Emergency Medical Journal 2001;18:46-50.
12. Boyce SH, Quigley MA, Campbell S. Management of ankle sprains: a randomised controlled trial of the treatment of inversion
injuries using an elastic support bandage or an Aircast ankle brace. British Journal of Sports Medicine. 2005;39:91-96.
13. Beynnon BD, Renstrm PA, Haugh L, Uh BS, Barker H. A prospective, randomized clinical investigation of the treatment of first-time
ankle sprains. American Journal of Sports Medicine. 2006;34:1401-1412.
14. Lamb SE, Marsh JL, Hutton JL, Nakash R, Cooke MW; Collaborative Ankle Support Trial (CAST Group). Mechanical supports for
acute, severe ankle sprain: a pragmatic, multicentre, randomised controlled trial. Lancet. 2009;373:575-81.
15. Lamb SE, Nakash RA, Withers EJ, Clark M, Marsh JL, Wilson S, et al. Collaborative Ankle Support Trial research team. Clinical and
cost effectiveness of mechanical support for severe ankle sprains: design of a randomised controlled trial in the emergency
department. BMC Musculoskeletal Disorders 2005;13:6:1.
16. Tsang KK, Hertel J, Denegar CR. Volume decreases after elevation and intermittent compression of postacute ankle sprains are
negated by gravity-dependent positioning. Journal of Athletic Training. 2003 Dec;38(4):320-324.
17. Bleakley C, McDonough S, MacAuley D. The use of ice in the treatment of acute soft-tissue injury: a systematic review of randomised
controlled trials. American Journal of Sports Medicine. 2004 Jan-Feb;32(1):251-61.
18. Airaksinen OV, Kyrklund N, Latvala K, Kouri JP, Grnblad M, Kolari P. Efficacy of cold gel for soft tissue injuries: a prospective
randomized double-blinded trial. American Journal of Sports Medicine. 2003 Sep-Oct;31(5):680-4.
19. Bleakley CM, McDonough SM, MacAuley DC: Cryotherapy for acute ankle sprains: a randomised controlled study of two different
icing protocols. British Journal of Sports Medicine 2006;40:700-705.
20. Currie A, Cooke MW. Compression. Evidence Based Sports Medicine 2007, Second edition: Blackwell Publishing pg. 208-222.
21. Raynor MC, Pietrobon R, Guller U, Higgins LD. Cryotherapy after ACL reconstruction: a meta-analysis. Journal of Knee Surgery.
2005 Apr;18(2):123-9. Review.
22. Holmstrm A, Hrdin BC. Cryo/Cuff compared to epidural anesthesia after knee unicompartmental arthroplasty: a prospective,
randomised and controlled study of 60 patients with a 6-week follow up. Journal of Arthroplasty. 2005 Apr;20(3):316-21.
23. Smith J, Stevens J, Taylor M, Tibbey J. A randomised controlled trial comparing compression bandaging and cold therapy in post
operative total knee replacement surgery. Orthopaedic Nursing. 2002; 21(2):61-6.
24. Fagan DJ, Evans A, Ghandour A, Prabhkaran P, Clay NR. A controlled trial of postoperative hand elevation at home following
day-case surgery. Journal of Hand Surgery (British). 2004 Oct;29(5):458-60.
25. Hovelius L, Olofsson A, Sandstrm B, Augustini BG, Krantz L, Fredin H, Tillander B, Skoglund U, Salomonsson B, Nowak J, Sennerby
U. Nonoperative treatment of primary anterior shoulder dislocation in patients forty years of age and younger: a prospective twentyfive-year follow up. Journal of Bone and Joint Surgery (American). 2008 May;90(5):945-52.
26. Ross G, McDevitt ER, Chronister R, Ove PN. Treatment of simple elbow dislocation using an immediate motion protocol. American
Journal of Sports Medicine. 1999 May-Jun;27(3):308-11.

27. Aronen JG, Garrick JG, Chronister RD, McDevitt ER. Quadriceps contusions: clinical results of immediate immobilization in 120
degrees of knee flexion. Clinical Journal of Sports Medicine. 2006 Sep;16(5):383-7.
28. Thorsson O, Lilja B, Nilsson P, Westlin N. Immediate external compression in the management of an acute muscle injury. Scandinavian
Journal of Medicine and Science in Sports. 1997 Jun;7(3):182-90.
29. Nadler SF, Prybicien M, Malanga GA, Sicher D. Complications from therapeutic modalities: results of a national survey of athletic
trainers. Archives of Physical Medicine and Rehabilitation 2003;84(6):849-53.
30. Cuthill JA, Cuthill GS. Partial leg thickness burn to the leg following application of a cold pack. Physiotherapy 2006; 92, 61-5
31. OToole G, Rayatt S. Frostbite at the gym: a case report of an ice pack burn. British Journal of Sports Medicine.
1999 Aug;33(4):278-9.
32. Graham CA, Stevenson J. Frozen chips: an unusual cause of severe frostbite injury. British Journal of Sports Medicine
2000;34(5):382-3.
33. Keskin M, Tosun Z, Duymaz A, Savaci N. Frostbite due to improper usage of an ice pack. Annals of Plastic Surgery
2005;55(4):437-8.
34. McGuire DA, Hendricks SD. Incidences of frostbite in arthroscopic knee surgery postoperative cryotherapy rehabilitation.
Arthroscopy. 2006 Oct;22(10):1141.e1-6.
35. Lee CK, Pardun J, Buntic R, Kiehn M, Brooks D, Buncke HJ. Severe frostbite of the knees after cryotherapy. Orthopedics.
2007 Jan;30(1):63-4.
36. Selfe J, Hardaker N, Whitaker J, Hayes C. Thermal imaging of an ice burn over the patella following clinically relevant cryotherapy
application during a clinical research study. Physical Therapy in Sport 2007;8: 153-158.
37. Moeller JL, Monroe J, McKeag DB. Cryotherapy induced common peroneal nerve palsy. Clinical Journal of Sports Medicine. 1997
Jul;7(3):212-6.
38. Hiness PR, Hvaal K, Engebretsen L. Severe hypothermic injury to the foot and ankle caused by continuous cryocompression
therapy. Knee Surgery Sports Traumatology and Arthroscopy 1998;6(4):253-5.
39. Dover G, Borsa PA, McDonald DJ. Cold urticaria following an ice application: a case study. Clinical Journal of Sports Medicine.
2004 Nov;14(6):362-4.
40. Khajavi K, Pavelko T, Mishra AK. Compartment syndrome arising from use of an electronic cooling pad. American Journal of Sports
Medicine 2004;32(6):1538-41.
41. Ibrahim T, Ong, SM. Saint Clair Taylor, GJ. The effects of different dressings on the skin temperature of the knee during cryotherapy.
Knee 2005;12(1):21-3.
42. Janwantanakul P. Different rate of cooling time and magnitude of cooling temperature during ice bag treatment with and without
damp towel wrap. Physical Therapy in Sport 2004;5:156-161.
43. Kennet J, Hardaker N, Hobbs S, Selfe J. Cooling efficiency of 4 common cryotherapeutic agents. Journal of Athletic Training
2007;42(3):343-8.
44. Gage AA. What temperature is lethal for cells? The Journal of Dermatologic Surgery and Oncology 1979;5:459-460.
45. Bleakley CM, Glasgow PD, Philips P, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists
in Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and
Elevation, 2011. Chapter 2: What is the magnitude and depth of cooling associated with ice?
46. Wilke B, Weiner RD. Postoperative cryotherapy: risks versus benefits of continuous-flow cryotherapy units. Clinics in Podiatric
Medicine and Surgery 2003; 20: 307-322.
47. Glasoe WM, et al. Weight-bearing immobilisation and early exercise treatment following a grade 2 lateral ankle sprain. Journal of
Orthopaedic and Sports Physical Therapy 1999;29(7):394-9.
48. Samoto et al. Comparative results of conservative treatments for both isolated anterior talofibular ligament (ATFL) injury and injury
to both the ATFL and calcaneofibular ligament of the ankle as assessed by subtalar arthroscopy. Journal of Orthopaedic
Science 2007;12: 49-54.
49. Nyska M, Weisel Y, Halperin N, Mann G, Segal D. Controlled mobilisation after acute ankle inversion injury. Journal of Sports
Tramatology and Related Research 1999;21(2):114-120.
50. Morsi E. Continuous flow cold therapy after total knee arthroplasty. Journal of Arthroplasty 2002;17(6):718-22.

Figure 1
Summary of the evidence base by primary intervention and injury type

NOTE: In some cases external supports/bandaging could have been used for dual purpose (protection/ rest and compression)
this depicted by the translucent compression bar; this figure includes the individual studies within eligible SR

Kerkhoffs,
2002b
SR (up to 2000)

Kerkhoffs,
2002a
SR (up to2002)
(Includes
Karlsson 1996)

Pijnenburg,
2000
SR (19661998)

ANKLE SPRAIN

Study

F vs F
(F= early mobilisation with any of the
following: Elast, Semi-R; Tape; Lace up)

N=9 RCT

-Injury: Acute injury to lateral ankle ligaments

(diagnosed by physical examination, stress
radiograph or arthrogram)

N=21 RCT / quasi RCT

Mean validity score: 9.1 10 / 18

Immob vs functional

F vs Min / no Rx (eg. elastic bandage;

note 3/52 in cast immobilisation was
defined as functional)

-Injury: Acute ruptures of ankle (in most trials

an arthrogram or stress radiographs or both
were used)
-Excluded: Recurrent ankle injury; chronic
instability; follow up <60% or unclear or
follow up exclusively with questionnaire
-Stats: Relative risks, data pooling with
random effects modelling
-Follow up: 6/12 - 3.8 y

-Injury: Acute injury to lateral ankle ligaments

(diagnosed by physical examination, stress
radiograph or arthrogram)

F vs CAST

Intervention

N=27 RCT

Inclusion criteria

Figure 2
Characteristics of included

-SUBGROUP ANALYSIS: Based on high quality studies (n=11 RCT), differences

noted for return to work only

-Number returning to sports: F >Immob (RR 1.86; CI: 1.2 - 1.9) [n=5 RCT]
-Days to return to sport
F<IMM (WMD: 4.8 days (95% CI: 1.5-8.3) [n=3RCT]
-Time to return to work
F<IMM (WMD: 8 days (95% CI 6.3-10.1)
-Swelling (short term): F<IMM (RR: 1.7; CI 1.17-2.59) [n=3 RCT]
-Patient satisfaction: F>IMM (RR 1.8; CI 1.09-3.07) [n=6 RCT]
No significant differences in: Pain, giving way (subjective instability), recurrent
sprain, swelling

Note: higher levels of complications with tape

- Return to work: Semi-R>Elast; (RR 4.3; CI 2.4-6) [n=2 RCT]

-Time to return to work: Semi rigid<Elast (WMD 9.6 days; CI: 6.3-12.9)
[n=1 RCT]
-Giving way: Semi-R<Elast (RR 8; CI: 1.03-62) [n=1 RCT]
-Swelling: Lace up<Elast (short term*) (RR 5.5; CI: 1.7-17.8)[n=1 RCT]; Lace
up<Tape (RR 4.07; CI: 1.2-13.7) [n=2 RCT]; Lace up<Semi-R (RR 4.19; CI: 1.2613.98) [n=1 RCT]
-All outcomes: Tape =semi rigid [n=2 RCT]; Elast=tape

-Time lost from work: F < CAST (15 days; range: 12-18 vs 38 days; range: 2848)
[n=10 RCT]
-Residual pain: F<CAST (6/52*) (RR: 0.67 CI: 0.5 to 0.9) [n=10 RCT]; F< min/
no Rx (6/52*) (RR: 0.53 CI: 0.27 to 1.02) [n=3 RCT]
-Giving way: F<CAST (6/52*) (RR: 0.69 CI: 0.5 to 0.94) [n=10 RCT]; F< min/no
Rx (6/52*) (RR: 0.34 CI: 0.17 to 0.71) [n=3 RCT]
-SUBGROUP ANALYSIS: No differences based on high quality studies (n=5
RCT)

Results (follow up)

Beynnon,
2006
RCT

Lamb, 2009
RCT

N=9 RCT

Jones, 2007
SR (up to
December
2005)

-Excluded: Previous sprain, previous

abnormal gait, fracture current or within
previous 12 months, syndesmosis injury,
burns laceration, 16 yr, pregnant, chronic
illnesses

-Injury: First time ankle ligament injury <72

hrs (Stratified by Grade 1,2 and 3 (Bergfield,
1986)

N=172

-Excluded: >7days since injury, <16 yrs,

fractures (not including flake fractures),
contraindication to immobilisation,

-Injury: Severe ankle sprain (unable to WB at

least 3/7 after injury

N=584

-Injury: Ankle sprains

Exclusion: <80% patients available for follow
up, non randomised
-Stats: Relative risk, data pooling
Follow up: 3/12 3.8 y

Inclusion criteria

Study

All groups: Week 1: Crutches as needed,

ice, elevation, mobility, and non impact
C/V, progressing to Stage 2: contrast
therapy, progressive dynamic exs,
strength, and finally Stage 3: agility,
sports specific drills.

-Elastic wrap (G1 and 2)

-Air-stirrup (all grades)
-Air-stirrup and elastic wrap (Grade 1
and 2)
-CAST (with WB) (Grade 2 and 3)

-ELAST (DTG)
- CAST
-Semi-R (Aircast)
-Semi-R (Bledsoe)
All groups: elevated and immobilised in
tubular compression for 2-3 days; and
10 days, crutches, advise on elevation,
and pain relieving medications

Immobilisation vs some form of early

functional treatment (early motion,
bandage, tape, brace, airstirrup, aircast,
wrap and airstirrup)

Intervention

Function (Karlsson), re-injury, jump, toe raise, ROM

No differences between groups (6/12)

-Time to normal walking (days / subjective)

Grade 1: Elast and Air < Elast=Air*
Grade 2: Elast and Air =Elast=Air < CAST*
Grade 3: CAST=Air
-Time to normal stair climbing
Grade 1: Elast and Air < Elast=Air*
Grade 2 : Elast and Air =Elast=Air < CAST*
Grade 3: Air<CAST (not sig)
-Time to full athletic / recreational activity
Grade 1: Elast and Air < Elast*
Grade 2:Elastic=Air and Elast < CAST*
Grade 3: CAST=Air

Note: At 3/12, subjective function was 8-9% higher with CAST or Semi rigid
support compared with DTG

FAOS, SF, Functional limitation profile: Tubi=CAST=Aircast=Bledsoe (9/12)

-FAOS (subscales function, pain, symptoms, and activity): CAST > Tubi (1/12*;
3/12*); Aircast >tubi (3/12* for function subscale only)
-SF12: CAST>Tubi (3/12*); Aircast > Tubi (4/12*)

Note: Pooled data was not significant for any outcome, however aside from patient
satisfaction, there were trends in favour of functional Rx

-% return to work
(RR: 1.06; CI 0.98-1.150) [n=414 patients]
-Days to return to work/sport: F<Immob [4/5 RCT]
-Patient satisfaction: Immob>F [n=2] (RR 0.6 CI 0.3-1.2) [n=598 patients]
-C/o Subjective instability:
F<Immob [3/5 RCT], (RR; 1.01 CI 0.72-1.42) [n=414 patients]
-Re-injury rate: F<Immob [5/6 RCT], (RR 0.81 CI 0.58-1.12)

Results (follow up)

Guskiewicz,
1999
RCT

Leanderson,
1999
RCT

Watts, 2001
RCT

Boyce, 2005
RCT

N=89

Bleakley, 2006
RCT

-Injury: Grade 1 or 2 ankle sprain (clinical

exam), <24 hrs post injury

N=30

-Injury: Grade 2 / 3 ankle sprains (Gordon,

1968 )(<24 hrs)

N=73

-Excluded: <16 yrs, mental illness, bony

injury, multiple injury, >24 hrs, no telephone,
patients whom the treating physician felt
would benefit from cast immobilisation

-Injury: Grade 1 and 2 ankle sprains, 24 hrs

N=400 (197)

-Excluded: <16 y, fracture

-Injury: Mean 3-4 hr old ankle sprain

(moderate/severe), 38% sports related

N=50

-Excluded: Grade 3, >72 hours

-Injury: Grade 1 and 2 ankle sprain, 50%

sporting population

Inclusion criteria

Study

-Rigid support (in built focal

compression)
-Semi-R (Aircast: alternating
compression pressures)
- Elastic bandage (with felt horseshoe)
All groups: standardised rehab from
day 1. Crutches, rehab, cryo, abstinence
from NSAIDs. Progressed according to
functional level attained.

-Compression bandage
-Semi-R (Aircast)
All groups: early mobilisation and
WBing, no physiotherapy

-DTG
-No DTG
All groups: Advice sheet on exercises
and analgesia

-Elastic
-Semi-R (Aircast)
All groups: standardised advise sheet on
RICE

All groups: standardised rehab and

mode of I

-I : Standard (20 min, x 3 per day)

-I: Intermittent (10 min on, 10 min off,
10 min on, x 3 per day)

Intervention

-Pain, swelling (volumetric analysis), function

No differences between groups (day 1,2,3,5,7)

Clinical exam, JPS, Isokinetic strength

No differences: Bandage v sSemi-R (1,3,5/7; 2,4,10/52)

-Time of figure of 8 run

Semi-R<Bandage (10/52*)

Note: 24% more patients took painkillers during the first week when using DTF

-Analgesic consumption
No DTG<DTG (7/7*)
-Pain (sleep disturbance), mobility, return to work
No differences(7/7)

-Function (modified Karlsson)

Semi-R>Elastic (10/7*; 1/12*)
-Pain, swelling
No differences (10/7, 1/12)

-Pain
Intermittent<Standard (1 week*)
Intermittent=standard (1-4/52)
-Function
Intermittent=standard (1-4/52)
-Swelling
Intermittent=standard (1-4/52)

Results (follow up)

-Injury: L ankle inversion injury, grade 2

based on clinical assessment, no history of
previous sprain

N=1 female athlete, aged 17y

Airaksinen,
2003
RCT

-Injury: sports related soft tissue injury

N=74

GENERAL SOFT TISSUE INJURY

Glasoe, 1999
Case study

-Injury: Acute Grade 3 ankle sprains

N=36

All groups: 4 times/day for 14 days

-cold gel
-placebo get

-Pain
Cold<P (week 1*, 2*, 4*)
-Function (disability)
Cold<P (week 1*, 2*, 4*)
-Patient satisfaction
Cold>P (week 4*)

Participation in sport 2 weeks post injury, limited clinical problems (4 weeks), no

re-injury at approximately 1 year follow up.

Immediate care: I/C immediately after

injury, 20 min every 2 hours, NWBing
Up to 1 week: Immobilised in boot with
WBing (1 week), active ROM, IC post
exercise
1 week onwards: resistive exercise,
bracing, progressive rehab

All groups: WBing as soon as pain

allowed

-Walking ability
Semi>CAST (week 3*)
Semi=CAST (day 3, month 6)
-Function / Return to work
Semi>Immob (week 3*)
Semi=CAST (day 3, month 6)
-Clinical assessment (swelling, ROM, balance, tenderness)
Semi<CAST (week 3*)
Semi=CAST (day 3, month 6)
Physiotherapy sought
Semi<CAST (month 6*)

-Function/Pain/Alignment (AOFAS)
ATFL>combined (month 6*, 1 year*, final* (mean 5 years)
-Joint laxity
ATFL>combined (month 6*, 1 year*, final* (mean 5 years)

Results (follow up)

-CAST and WBing: 3 weeks in short leg

walking cast
-Semi-R and WBing: 3 weeks early
walking, aircast external support

Both groups: 1 week IMM with WB,

followed by 1 week in functional brace
with WB and physical therapy, jogging
permitted after 6 weeks, return to sport
after 10 weeks

-Injury: Acute lateral ankle ligament injury;

diagnosis by clinical assessment, stress X-Ray,
and arthrography (mean 3.5 days post
injury)

Nyska, 1999
CT

-Isolated ATFL injury

-Combined ATFL and CFL injury

N=55 injuries (54 participants) (grouped by

injury severity)

Samoto, 2007
CT

Intervention

Inclusion criteria

Study

N=12

Tsang, 2003
RCT

-Injury: primary shoulder dislocation (verified

radiologically), reduction performed/verified
by an
experienced surgeon

N=257 shoulders

-Knee braced in 120 deg flex continuously for

24 hrs, crutch walking, followed by pain-free
stretching and isometrics. Thigh padding for
remainder of athletic season.

N=47 (mean age 19.2; range 18-22 yrs)

Aronen, 06
Case Series

-Injury: quadriceps contusion during sport,

unable to continue sport, unable to perform
SLR

-C (30 minutes of compression bandage

on maximum stretch = 40-80mmHg, within
5 minutes of injury. Moderate compression
(40mmHg) after 30 minutes, rest and elevation
-Rest and elevation (in some cases, non
maximal compression after 10-30 minutes)

-Reduction within 1hr post injury, next day

active ROM exs under supervision, progressing
to strengthening, supplemental treatments:
E-stim, cryotherapy, compression

N=40 athletes with contusion or distension

injuries (thigh or calf)

N=20 (age: 18-24 y)

-injury: closed posterior elbow dislocation
-excluded: fracture (aside from small
avulsion fracture), non posterior dislocation,
spontaneous reduction

Protocol violators were analysed in a separate

group (n=41)

-Immobilisation (sling 21 days 28 weeks)

-Partial immobilisation (sling until patient was
comfortable) (n=37 for <5/7, n=50 for 7/7,
n=16 for 2/52, n=1 for 3/52)

-Elevation (vertical)
-Elevation and intermittent compression (max
inflation to diastolic BP, 45s inflation;15s
deflation)
All groups: 30 minutes Rx

Intervention

Thorsson, 97
CT

MUSCLE INJURY

Ross, 1999
Case series

ELBOW DISLOCATION

Hovellius, 08
OB

SHOULDER DISLOCATION

-Injury: Inversion ankle sprain, 2-4 days old

Inclusion criteria

Study

-Time to full active flexion

All subjects: 72 hrs
-Time to unrestricted activities
3.5 days (range: 2-5 days)
-Myositis ossificans (X-Ray)
1/46

Note: Maximum compression group reached full subjective recovery

approximately 5/7 faster than control

-ROM, Serum CK, Ultrasonography (echostructure normal Y/N,

type), Time to subjective recovery, isometric strength:
No significant difference between groups

-Mean final elbow ROM

4 deg through to 139 deg
-Time to final ROM
19 days
-Final extension within 5 deg of un-injured side
All patients
-Re-dislocation (1 y follow up)
1/20

-Re-injury
Sub-grouped into age at time of dislocation (yrs) and analysed
Treatment group was not a factor in the risk of re-dislocation at 25 y
follow up

Note: 10 mL greater decreases in volume with E alone

-Change in ankle volume between pre and post (volumetric

analysis)
Both groups ankle volume decreased after Rx; E > E and C (immediate),
No significant difference between groups
5 minutes post Rx, volume returned to baseline levels in both groups

Results (follow up)

-Injury: recovery post Unicondylar knee

arthroscopy

-I/C (cryocuff, gauze barrier, automated

cooling device, 48 hrs)
-EDA (2.5-5 mg/mL bupivacaine)
-Analgesics and standardised
rehabilitation only (as below)
All groups: paracetamol, supplementary
morphine when required, standardised
rehabilitation

N=60

Holmstrom,
2005
RCT

Pain, bleeding, swelling, ROM, function

No differences between groups (day 1-7, six weeks)

Morphine consumption
EDA=I/C<Control (day 1*)

Note: 11-13mL increase in volume over 5/7 post surgery

-Swelling (volumetric analysis)

Both groups increased post surgery
E (high) = E(simple) (day 5)

-E (high sling, above heart, 45 deg full

elevation)
-E (simple sling, shoulder neutral, elbow
at 90 deg)

-Pain
I<placebo* [n=6 RCT]
-Post operative drainage
I=control / room temperature [n=4 RCT]
-ROM
I=control / room temperature [n=4 RCT]

Note: Few definitive conclusions due to heterogeneous data, wide variations in

intervention mode dosage, treatment comparison

-Swelling
I<H=C (day 5*) [n=1]
-Pain
I ex < ex alone (week 1*) [n=1]
I/C <C (week 1*) [n=2]
I<placebo (week 1*) [n=1]
All outcomes
I/C=C [n=6]
I/C=no intervention [n=4]

Results (follow up)

I vs No I (control)
I vs Room temperature (compression)

I vs H, contrast, E-stim, No Rx, placebo

I and ex vs ex alone
I/C vs I
I/C vs C

Intervention

N=43
-Injury: recovery post carpal tunnel
decompression
-Excluded: Revision surgery, concurrent
disease, post traumatic CTS, unable to use
sling or understand its use

Inclusion: srthroscopically assisted ACL

reconstruction treated post operatively with
cold therapy, comparison to no cold therapy
or room temperature device
Stats: effect size estimates and data pooling
using random effects modelling
Follow up: Short term

-Injury: recovery post ACL

N=7 RCT

-Injury: acute STI, recovery post orthopaedic

surgical procedure, pain, swelling, function,
ROM
-Inclusion: English language
Stats: Effect sizes, no data pooling
Follow up: 1/7 to 1/12

N=24 RCT/CT

Inclusion criteria

Fagan, 2004
RCT

Raynor, 2005
SR (up to
November
2002)

Bleakley,
2003; 2007
SR (up to April
2005)

SURGICAL

Study

-Injury: Bilateral TKR, 6 weeks between

surgery

N=60 (30 patients bilateral)

-I/C (continuous cold flow cooling device,

immediately post op, over post surgical
dressing, C from crepe bandage, skin
temperature maintained at 2C for first 2 h post
surgery, then 12C continuously for 6 days )
-C (crepe bandage)

-C (Robert Jones for 24 hrs, followed by ice

bags in linen towel for 24-48 hrs)
-I (Robert Jones for 6 hrs; cryopad technology
for 18 hrs; followed by ice bags in linen towel
for 24-48 hrs)

Intervention

Pain
-VAS:I<C (mean over day 1-6*)
-Analgesic consumption: I<C (mean over day1-6*)
ROM
I>C (week 1*,2*)
I=C (week 6)
Swelling/Blood loss
I<C (day 6*)

Length of hospital stay, blood loss, swelling, ROM (flexion), pain

C=I

Results (follow up)

Y
Y
Y
Y

Kerkhoffs, 2002b

Jones, 2007

Bleakley 2003, 2007

Raynor, 2005

Y: AMSTAR criterions fulfilled; : AMSTAR criterion not fulfilled3

Kerkhoffs, 2002a

Pijnenburg, 2000

AMSTAR CRITERION

Figure 3
Review authors judgements on AMSTAR criterion item for each included SR

10

11

Y: years; F: Functional treatment; Rx: treatment; TKR: Total knee replacement; CAST: Cast immobilisation; Elast: Elastic bandaging; Semi-R: Semi Rigid support; WB: weight-bearing ; C/V: Cardiovascular training; Ex:
exercise; Sx: Subjective; I: Ice; H: Heat; E-Stim: Electrical Stimulation; FAOS: Foot and ankle score; AOFAS: American orthopaedic and foot society; SF-12: Short form 12 (mental and quality of life scoring scale); FLP:
Functional limitation profile (UK version of the Sickness Impact Profile); DTG: Double tubigrip; Bergfeld J, Cox J, Drez D et al. Symposium: management of acute ankle sprains. Contemp Orthop 1986;13:83-116. ;
Gordon BL (1968). Standard nomenclature of athletic injuries. American Association of Orthopaedic Surgeons, Chicago, pg. 7.

Morsi, 2002

N=84

Smith, 2002
RCT

-Injury: recovery post unilateral TKR

Inclusion criteria

Study

Figure 4
Risk of bias summary: review authors judgements about each risk of bias item2
for each included clinical study

Details of Cochrane risk of bias tool2 provided in Appendix Table 10

N=4 cases of frostbite during post operative

care

N=1 female with soft tissue injury to left

gastrocnemius

N=1female with acute synovial arthritis

Maquire, 2006
(case series)
(data extracted
from abstract
only)

Cuthill, 2006
(case study)

Keskin, 2005
(case study)

at knee

N=1 male post bilateral patellar tendon

repair

Lee, 2007
(case report)

Duration: approximately 2 hrs to knee

Barrier:Direct?

Mode: Cold pack

Duration: approximately 30 minutes,

Barrier: dishcloth

Mode: Commercial cold pack, leg

resting on chair, with cold pack between.

Duration?

Barrier?

Mode: Cold cuff device

Duration: continuous?

Barrier: Dressing and paper towel

interface

Mode: Cooling pad (water and ice

circulated through the cooling pad)

3 days post operative cryotherapy, 2

weeks of home cryotherapy

Duration: 20 min

Barrier: None

Mode: Commercial gel pack (Surface of

gel pack = -17C at start of treatment)

Lab based experiment

N=1 healthy male, aged 43 years, recruited

into a research study

Selfe, 2007

(case report)

Details of intervention

Subject / inclusion criteria

Study

SKIN BURN

Figure 5
Characteristics of studies

-Grade 2 and 3 frostbite on medial aspect of knee

-Treated with minor debridement and daily dressings for 10 days.
Full recovery reported.

-On removal of pack, a large hard and dusky purple area noted with blanching of
skin.
-After 1 hr this became painful with erythema.
-Next day calf was swollen with blistering (3% of body surface area) = superficial
and deep partial thickness burns treated conservatively.
-14 days off work, with full skin coverage taking 12 days.

-Frostbite at top of patella.

-Speculation that the area of frostbite in all four cases corresponded with an area
of the cooling cuff that was modified for care ACL surgery, of note, none of the
cases had ACL surgery.

-At 2 weeks noted skin changes

-Admitted for debridement and dressing (two occasions), final treatment
involved further debridement and rectus abdominus microvascular transplants
on both knees.

Blanched skin over patella with a surrounding zone of erythema immediately after
pack removal. Area became red and hot over next 3 hours, with discomfort in bed
later that night. Skin colour and sensation returned to normal the next day, no long
term problems.

Details of adverse outcomes

N=1 female aged 59 yrs, calf strain

OToole, 1999
(Case study)

N=1 female; 59 yrs; closed dislocated

bimalleolar fracture left ankle ; clinical trial,
use of continusous cryocompression for
preoperative swelling

N=1 with hamstring soreness; aged: 22 yrs

Hoiness, 1998
(Case study)

Moeller, 1997
(Case study)

NERVE DAMAGE

Mode: Bag of frozen chips,

N=1 female age 42 yrs; 4 week history of

vague discomfort in foot. Previous treatment
was NSAID; and 3 days previously icing.

Graham,
2000
(Case study)

Duration: 20 minutes

Barrier: direct to skin

Mode: Ice pack compressed between

table and distal hamstring.

Duration: Continuous cryocompression

for 5-7 days before surgery.

Barrier: directly to the skin, removed

once per day to check skin

Mode: Cryocuff with elevation and

immobilisation

Duration: 20 minutes.

Barrier: direct application

Mode: ice pack placed between a foot

stool and the patients elevated leg

Duration: fell asleep, at least 40

minutes?

Barrier: Towel

Details of intervention

Subject / inclusion criteria

Study

-Unable to move foot after removal of ice pack

-Grade 3 muscle activity (dorsiflexion, toe extension and eversion), and deep
tendon reflexes were 1 / 4 at Achilles and 2 / 4 at knee after 24hrs and numbness
over the anterolateral aspect of the left leg.
-Symptoms persisted for 1 week. Nerve conduction studies at common peroneal
nerve found axonotmesis with 50-70% axonal loss compared to the opposite leg.
-Minimal strength improvements at 4 months.
-Full recovery noted at 12 months.

-6 months: Clawing of toes and marked dystrophy of the foot, with discoloration
and slight hyperaemia of the skin. Dorsal skin on foot was anaesthetic.
Neurograph showing damage to all major nerves in the foot. Severe stiffness of
T/C joint, and subchondral fracture of talus.

-On the 6th day pre surery, patient c/o coldness and numbness on foot.
-On assessment the foot was cold anaesthetic and glassy looking, with reduced
microcirculation.
-No sign of thromboembolism, and doral pedis and posterior tibial blood flow
were normal on Doppler.
-Blistering and epidermolysis developed in next few days.
-Patient underwent surgery external fixation for bony injury.

-Over 24 hrs large blistered area developed.

-Presented to casualty 4 days later, where a diagnosis of a superficial partial
thickness burn (1% of body) was made.
-Treated conservatively, with healing after 10 days.

-Erythema on removal of chips, developing into to discolouration and pain the next
day. -Paraesthesia of index and middle toes.
-At three days there was 4 x 3 cm area of skin necrosis (third degree frostbite
deep dermal layer) on the dorsum of the foot, surrounded by erythema
and blistering.
-Debridement under anaesthetic, and treatment with a split thickness skin graft,
with wound healing satisfactorily.
-Destruction of the superficial and deep peroneal nerves to 2 and 3rd toes.

Details of adverse outcomes

Subject / inclusion criteria

Barrier: direct?

Had previously used ice for a knee injury in

the past

Cuthill, 2006
(questionnaire)

N=111 physiotherapists in Scotland (private

practice)
72 % response rate

Duration: 30 minute duration

Mode: 1 kg ice to shoulder, air removed

secured with an elastic bandage

Duration: continuous, refilled with ice

every 3 hrs, and once during the night.

Barrier: Dressing between cooling unit

and skin, aside from at day 4 (direct
contact at posteriomedial calf).

Mode: Electronic cooling pad, 33F,

Details of intervention

N=1 male college student, 19 yr; 9.7 %

body fat; participating in a research study

N=1 male aged 35 years; Osteoarticular

transfer system (OATS) surgery knee;
discharged with cryotherapy instructions

QUESTIONNAIRE

Dover, 2004
(case study)

ALLERGY

Khajavi, 2004
(Case study)

COMPARTMENT SYNDROME

Study

Total years of practice from physiotherapists responding to questionnaires = 725 yrs

Incidence rate of cold induced injury estimated at 1 per physiotherapist every 5 yrs.

% of therapists with experience of witnessing cold induced injury

49%: 0 injuries
20%: 1 injury
23%: 1-5 injuries
9%: 5-10 injuries

-After 10 minutes of icing: subjective complaint of headache, feeling woozy,

dehydrated.
-Skin temperature did not attenuate as per other subjects with in the study.
-Lowest temperature reported was 23.4C, in comparison to 12C in other subjects.
-Unusually skin temperature began to increase after 10 minutes of treatment (total
increase of 4.2C)
-After removal of the ice pack (30 minutes): wheal formation and discoloration
around the shoulder consistent with cold urticaria. Similar to previous reaction at
knee after icing.
-Resolved within 48 hrs.

Day 5 post operatively pain was noted in the calf, mild to moderate calf swelling,
and large degree of erythema on posteromedial calf.
-No evidence of DVT on ultrasound examination. High compartment pressures
of 53mmHg in anterior and lateral compartments of leg, and 37mmHg and 20
mmHg in superficial and deep posterior compartments respectively.
-Emergency fasciotomy performed, no muscle debridement necessary initially (small
portion of muscle debrided in superficial compartment 2 days later).
-Full recovery at 33 month follow up.
-Full recovery noted at 12 months.

Details of adverse outcomes

Subject / inclusion criteria

N=3012 certified athletic trainers

110 podiatrists

Nadler, 2003
(Questionnaire)

Wilke, 2003
(Questionnaire)

QUESTIONNAIRE

Study

Survey on the use of continuous cold

flow units, and the incidences of
complications

No details on the circumstances of the

complications eg. duration, mode etc.

Descriptive questionnaire on the

frequency and types of complications
associated with therapeutic modalities

Details of intervention

-29 responses
-More than half used continuous cold flow units.
-No serious complications noted with ice bags, several minor complications noted
with ice bags and continuous flow units
-5 serious complications noted with continuous flow units: (Erythema and blistering
avascular necrosis of talus; cyanotic forefoot full recovery; erythema, oedema full
recovery; cyanotic forth toe amputation 4th toe; cyanotic forefoot amputation of
4th and 5th digits)

-30% response rate: of which 26% reported a complication.

-Cryotherapy accounted for the highest percentage of complications (43%)
-Other types of modalities listed were heat, electrical stimulation, therapeutic exercise
and other
N=86 allergy
N=16 pain or intolerance
N= 23 burn
N=6 frostbite
N=2 Raynauds
N=11 skin rash
N=3 Diaphoresis
N=1 unspecified

Details of adverse outcomes

Executive Summary:
Methods, Clinical Recommendations and Implications
Background
Soft tissue injury is a common problem in sport, recreational and physical activities. The quality of acute stage management
of a soft tissue injury is thought to be an important determinant for short and long term recovery. In the early stages, soft
tissue injuries are characterised by an acute inflammatory response. This is manifested clinically by the presence of pain,
redness, swelling and loss of function. Traditionally, the clinical management of soft tissue injury has placed most emphasis
on minimising or controlling acute inflammation. Protection, Rest, Ice, Compression and Elevation (PRICE) remains one of
the most popular approaches for the management of acute injuries and in particular the control of inflammation. Some
or all components of PRICE continue to be combined, based on the type or severity of soft tissue injury.
Original ACPSM guidelines
Soft tissue injury is a common problem in sport, recreational and physical activities. The quality of acute stage management
of a soft tissue injury is thought to be an important determinant for short and long term recovery. In the early stages, soft
tissue injuries are characterised by an acute inflammatory response. This is manifested clinically by the presence of pain,
redness, swelling and loss of function. Traditionally, the clinical management of soft tissue injury has placed most emphasis
on minimising or controlling acute inflammation. Protection, Rest, Ice, Compression and Elevation (PRICE) remains one of
the most popular approaches for the management of acute injuries and in particular the control of inflammation. Some
or all components of PRICE continue to be combined, based on the type or severity of soft tissue injury.
Aim
Our aim was to review the pathophysiological rationale and clinical effectiveness for using PRICE in acute soft tissue injury
management, and produce clinical guidelines for its use.
Methods
In January 2009, ACPSM members from a range of disciplines volunteered to form a working group responsible for
guideline development. The guideline development process followed the Grading of Recommendations, Assessment,
Development, and Evaluation (GRADE) approach.2 The following milestones were made and addressed in order: 1)
develop specific clinical questions; 2) literature searching; 3) address importance of outcomes (critical; important but not
critical, not important); 4) extract data/grade quality of evidence; 5) develop final recommendations.
Group consensus in January 2009 produced 5 clinical questions relating to using PRICE, or its individual components to
treat acute soft tissue injury (Figure 1). Extensive literature searching (electronic, hand searching) was undertaken by at
least two researchers, and was completed in October 2009. Evidence relevant to each clinical question was extracted,
graded, summarised and reviewed by the working group, and external experts. Data from over 250 relevant studies were
extracted and summarised.
Final guideline recommendations and relevant clinical implications were made by consensus discussion in September
2010. The strength of the recommendation (for or against the intervention) was graded as strong (definitely do); weak
(probably do), or uncertain (indicating that the panel made no specific recommendation for or against interventions).
Factors that influenced decisions on the strength of recommendations were: quality of evidence, balance between
desirable and undesirable effects based on the values and preferences in terms of health outcome benefits, burden
and expense.

Figure 1

March 2009: Establish working group

Key for Figure 1

Clinically relevant questions:
1 What is the magnitude and depth of cooling associated with
ice?
2 Can PRICE decrease the inflammatory response after acute
soft tissue injury?
3 What effect does mechanical loading have on inflammation
and soft tissue healing after acute injury?
4 Do the physiological effects of local tissue cooling affect
function, sporting performance and injury risk?

Consensus on clinically relevant questions

Lit searching (96 Oct 09)
1). TEMPERATURE RED (I/C)
2). INFLAMMATION (PRICE)
3). MECHANICAL LOAD VS PR
4).SAFE RETURN TO SPORT (I/C)
5). CLINICAL REVIEW OF
RATIONALE FOR (PRICE)

5 Which components of PRICE are effective in the clinical

management of acute soft tissue injury?

Grading of outcomes
Evidence profiles

Studytype

Directness (external validity)

Quality (Risk of bias)

Pathophysiological rationale

Consistency (pattern of effects

and volume of research

Figure 2
Overview of relevant studies used to inform panel consensus

Sept 2010: Final

Recommendations

Evidence profile

Costs

Benefits vs Harms

Values and preferences

H: Human study (injured model)

A: Animal study (injured model)
EIMD: exercise induced muscle damage

Summary of Evidence and Clinical Implications

Protection And Rest
Unloading injured tissue
The rationale for protection and rest after an acute soft tissue injury is to minimise bleeding, and prevent excessive distension or re-rupture at
the injury site. Animal models confirm that periods of protection/rest (tissue unloading) are required in the acute stages after soft tissue injury
and that aggressive ambulation or exercise should be avoided.3 Previous guidelines suggested isometric exercise can be undertaken in the
immediate stages after injury; 1 we found very low quality clinical evidence to support this.4

Loading injured tissue and stress induction

There is moderate quality clinical evidence that functional treatment, which incorporates early mobilisation and weight-bearing, is more
effective than cast immobilisation after ankle sprain (particularly less severe sprains), for a range of important short term outcomes. There is
a dearth of evidence in this area for other types of soft tissue injury. Evidence from basic science models concur that progressive mechanical
loading (in comparison to immobilisation and/or unloading) in animals is more likely to restore the strength and morphological characteristics
of collagenous tissue after injury.3 The mechanism may be that mechanical loading upregulates gene expression for key proteins associated
with tendon tissue healing.3 It is not clear whether this is consistent for all other soft tissue, particularly in human models. Notwithstanding this,
exercise or tissue loading should perhaps be more formally recognised as an essential component of soft tissue injury management. Although
functional exercise already plays a key role in rehabilitation, much of its rationale relates to maintenance of mobility and strength during
recovery. We would suggest that progressive mechanical loading could also play an important role in tissue healing, and restoration of tissue
morphology and biomechanics after injury. Developing evidence based guidance on the optimal timing, dosage and nature of initiating loading
after injury is important, and we must ascertain whether this should vary according to the type of soft tissue affected (eg. muscle vs tendon
vs ligament).

Clinical relevance
The optimal nature and duration of protection/rest is not clear and ultimately depends on injury severity. We would suggest that for the majority
of soft tissue injuries, as a minimum, movements replicating the injuring force are avoided in the acute phases. There is potential that excessive
protection/rest (tissue unloading) will do harm. Animal models confirm that 2-3 weeks of protection/rest result in adverse changes to tissue
biomechanics and morphology.3 This threshold may not be clinically applicable however, and the changes may be reversible.
Mechanical loading after injury may be reparative or damaging. Clinically, practitioners should be aware of the importance of making a
safe transition from protection/rest to tissue loading after injury. Until further evidence is forthcoming, we would suggest that the transition
should continue to be made carefully, within the limits of pain, with full consideration of the injury severity and nature of damage (tensile vs
compression) and affected tissue (ligament vs muscle vs tendon). When applicable, therapeutic exercises should provide progressive loading
on the injured structures, based on speed, range and number of repetitions.

ICE
The basic premise for cooling after injury is to extract heat from the body tissue to attain various clinical benefits. These include: limiting the
extent of injury by decreasing tissue metabolism thereby reducing secondary cell death, providing analgesia, and facilitating rehabilitation.5,6

Cold induced analgesia

We found most evidence that ice can provide adequate pain relief after injury. There is moderate clinical evidence that cold therapy is effective
at decreasing short term pain after acute ankle injury and general soft tissue contusion, and low quality clinical evidence that shorter intermittent
applications are sufficient to induce short term analgesia.4 Based on a post surgical model, there is high quality evidence that cold therapy
provides effective short term analgesia.4 These clinical findings are supported by basic scientific evidence; and we found numerous laboratory
based studies confirming that skin temperature is reduced to less than 13C after 5-15 minutes of cooling.5 This is the currently accepted
threshold for inducing effective local analgesia. Importantly, not all modes of cryotherapy cool equally or at the same rate.5 We would suggest
that in most clinical scenarios, crushed ice (ice cubes in a plastic bag) should be the mode of choice. Primarily, this is the most widely researched
mode of cooling. Furthermore, it is assured that crushed ice (when melting) will be close to 0C throughout treatment; in contrast the surface
temperature of commercial gel packs is more difficult to estimate which risks ineffective cooling or extreme temperature gradients. Finally,
crushed ice reaches threshold temperatures for analgesia faster than other modes of cooling.5

Ice and inflammation

There is moderate clinical evidence that ice has little effect on other clinical outcomes including recovery time, function, and swelling (limb girth).
There may be many reasons for this lack of effect. Primarily, the majority of studies in this area have used post surgical models; few have used
closed soft tissue injury. We must also acknowledge that it can be difficult to induce clinically significant temperature reductions on deep muscle
and joints after injury; and currently accepted threshold temperatures for reducing cellular metabolism (5C-15C) seem unattainable.5 There
is very low quality evidence from human studies that ice may reduce metabolism or aspects of the inflammatory response;6 the effect that ice
has on biomarkers of inflammation and muscle damage after EIMD in humans is inconsistent.6
There is evidence from animal models to suggest that ice can have a consistent effect on some important cellular and physiological events
associated with inflammation after injury. This includes cell metabolism, white blood cell activity within the vasculature, and potentially apoptosis
(all reduced6). The relative benefits of these effects have yet to be fully elucidated and it is difficult to contextualize within a human model. A key
factor to consider is that large temperature reductions can be readily induced at the site of injury in an animal model.
Another important mechanism associated with cooling in the immediate stages after injury is the vasoconstrictive effect it may have on the
vasculature. The original PRICE guidelines found consistent evidence that various modes of cooling decreased blood flow in human models.
We found further evidence that cooling reduced blood flow in healthy human tissue, with no evidence of cold induced vasodilation.6 Other cold
induced effects included reduced oxygen saturation, and facilitated venous capillary outflow in healthy tendon models.6 There is a dearth of
related studies on injured humans; and the effect of ice on microcirculation in injured animal models is inconsistent.6

Figure 3
Summary of the effects of PRICE on white blood cell activation and secondary damage

*Animal model (n=3 studies) found cooling decreased % of adherent and rolling WBCs based on intravital microscopy; 4 animal models found lower levels of WBCs after
cooling based on histological analysis of excised tissue, one found increased numbers of macrophages after cooling6
Animal model (n=1 study) found immediate exercise increased neutrophils and macrophages (ED1+ and ED2+) based on histological analysis of excised tissue3
Animal model (n=1 study) found cooling inhibited the loss of mitochondrial oxidative function6
Animal model (n=1 study) found cooling decreased number of apoptotic muscle cells6

Figure 4
Summary of the effects of PRICE on biochemical markers of
inflammation and oxidative stress

Animal model (n=1 study) found immediate exercise increased neutrophils and macrophages (ED1+ and ED2+) based on histological analysis of excised tissue 3
Animal model (n=1 study) found that mechanical loading significantly increased expression compared to unloaded 3
Animal model (n=1 study) found trends that mechanical loading decreased expression compared to unloaded 3
Animal model (n=1 study) found trends that mechanical loading decreased expression compared to unloaded 3

Targeting deep tissue injury

Based on basic scientific and clinical evidence, we would suggest that in some clinical scenarios, cold induced benefits may be limited to
analgesia; particularly in situations associated with higher levels of body fat, deep soft tissue injury or when insulating barriers are necessitated
(eg. post surgical models) at the cooling interface. We do however acknowledge that the current threshold temperatures for achieving analgesia
(skin temperature of <13C) or metabolic reduction (reduction to 5-15C at the injury site) are not definitive and may be subject to change
based on emerging understanding of the biochemical and physiological events surrounding acute injury and inflammation.7 We can also
not discount the clinical value of inducing modest deep temperature reductions, or using ice to prevent injured tissue from heating up above
baseline levels.5
The previous guidelines1 recommended applying ice every 2 hours during the acute phases. For certain injuries, it may be more effective to
increase the cooling dose by decreasing the time between applications. This will maximise the potential to lower deep tissue temperature.
However, we must consider that prolonged continuous ice applications risk side effects; potentially the safest balance is to apply more regular
intermittent applications. Lab based evidence shows that short periods of ice pack removal allow skin temperatures to re-warm (thereby
avoiding extreme skin or nerve temperature reductions), without re-warming deep tissue temperatures.5

Figure 5

-Circles represent values from participants with 21-40mm skin-fold at

the point of cryotherapy application
-Lines represent values from participants with <20 mm skin-fold at the
point of cryotherapy application
-Arrows overlap the area representing potentially optimal INTRAMUSCULAR temperature reduction of 5-15C.
-All data is based on crushed ice, applied directly onto the skin surface 5
-Three studies applied additional external compression 5

Figure 6
Mean Re-warming rates in Muscle (Intramuscular depths 1-3cm)

Blue line: 1-1.5 cm IM depth

Green line: 2 cm IM depth
Red line: 3cm IM depth
Dashed line: Exercising post treatment

Combining ice and exercise

Cryokinetics is a technique which uses ice to facilitate therapeutic exercise after injury. This practice may be gaining popularity within the UK, and was
not fully discussed within the previous guidelines.1 There remains a dearth of related clinical evidence for cryokinetics; however there are patterns
from basic scientific reports which align well with its general concept. Both healthy and injured models show isolated joint cooling or short periods
of muscle cooling can have an excitatory effect on muscle activation.8 Certain soft tissue injuries are associated with varying degrees of muscle
inhibition, and cooling may allow patients to access motor neurones that may be otherwise unattainable after injury. There is no evidence to suggest
that cryokinetics is safe in the acute phases of injury. The risk is that cooling is erroneously used to extensively anaesthetise an injured area to facilitate
aggressive rehabilitation. Most recommendations concur that cryokinetics is most appropriate after the acute phases. We would suggest that
the exercises included in a cryokinetics treatment package should be relevant to the stage of tissue healing.

Pitch-side management and return to play

There are many occasions, particularly in sport when ice is applied prior to physical activity. Often this involves very brief applications during a
break in play, where the cold sensation is used as a counter irritant for pain. This technique is usually effective, and is associated with minimal
tissue cooling. Other scenarios involve longer duration icing (eg. half time, rolling substitutions or between events), and more significant levels
of cooling can result prior to initiating or returning to sporting activity. It is important that clinicians consider the effect that this may have on local
tissue physiology. We found a large number of studies relevant to this area (n=46); the corollary was that although cooling induced a number
of changes to various physiological systems relevant to athletic performance, some of which were statistically significant, their clinical relevance
was generally questionable.8 However, we would suggest that longer periods of intense cooling (>10 minutes) may negatively influence some
correlates of athletic performance (eg. vertical jump, sprinting speed), and consequently affect injury risk.8 Cooling of the hands, feet, or other
areas that typically have less body fat may be most at risk. The effect that cold has on other important correlates of functional performance such
as precision and sporting performance per se is not clear. It is likely that any risks are negated with a progressive warm up prior to returning
to activity.

Adverse effects
Based on evidence from case reports and questionnaires, it is clear that cold therapy does have the capacity to do harm. A small number of
reports described isolated cases of permanent scarring on the skin and temporary and permanent loss of muscle function.4 There are few
reports of adverse events within laboratory controlled and clinical studies. We feel that the risk of inducing adverse events with cooling is low,
provided contraindications are fully checked, and clear instructions for practical application and dosage are given.

Compression And Elevation

Acute injury and oedema
Healthy tissue maintains a dynamic balance of fluid movement between the vascular system and interstitial tissue space. Much of this resides in
the fine interaction between hydrostatic and osmotic forces at the capillary bed. Acute soft tissue injury and the ensuing inflammatory response
can affect hydrostatic and osmotic pressures, disrupting this balance. Around the injury site, changes in blood viscosity and flow tend to increase
the hydrostatic pressure within the blood vessel; and macromolecules accumulate within the interstitial space reversing the osmotic gradient.
The net effect is increased fluid movement from the blood vessel, into the surrounding interstitial space resulting in oedema formation.

Elevation
One of the primary reasons for using elevation or compression after injury is to try to restore the pressure gradients within the affected tissue.
This is done through external mechanical pressure (compression) or gravity (elevation). The physics and physiology underpinning the rationale
for elevation are perhaps more clear-cut. Elevating an acutely injured ankle decreases the gravitational force exerted on the column of blood
between the heart and foot, thereby decreasing the hydrostatic pressure within the vasculature, minimising oedema and the resistance to
venous and lymphatic flow around the injured tissue.

Despite clear rationale, the optimal duration or angle of elevation remains contentious. If our primary clinical goal is to limit tissue oedema and
avoid the pain and discomfort relating to increased tissue pressures, it seems logical to elevate as much as possible, for as long as possible.
Clinical evidence for elevation is poor however; we found one relevant post surgical model showing no effect.4 Other models found foot and
ankle volumes cannot be decreased for a prolonged period of time with elevation, as volumes quickly return to baseline on returning to sitting
or standing.4 This has been termed, the rebound effect. Intermittent return to a gravity dependent position is inevitable in most body parts
after injury, particularly at distal body segments (eg. ankles). Clinically, a graduated return to standing after elevation is perhaps most likely to
minimise rebound swelling and discomfort. This is in accordance with the previous guidelines.1

Compression pressures
Despite a relatively large number of studies into the effectiveness of external supports and bandages, much of the evidence for compression is
conflicting.4 The exact physiological rationale has yet to be fully explicated, and there is little evidence for an optimal compressive force. The
rationale for using high levels of compression pressure is to minimise initial tissue haemorrhage after injury. We found very low quality evidence
that this approach is ineffective after muscle injury. This was based on one controlled study comparing immediate compression (bandaging)
at 80mmHg, with rest/ice.4 A related rationale for compressing is to prevent oedematous fluid from accumulating within the interstitial space.
The external mechanical pressure is thought to increase the hydrostatic pressure of the interstitial fluid thereby forcing fluid from the injury site
towards the capillary, lymph vessels, or tissue spaces away from the traumatised area. Smaller external forces may achieve this effect; many
standard protocols recommend pressures between 15 and 35mm/Hg.9,10

Limb shape
Double tubigrip (DTG) provides external pressures of around 15-25 mmHg, but has no clinical benefit after ankle sprain.4 Some have expressed
concern that as DTG does not conform to the shape of the ankle its compressive force is mitigated, and may even encourage swelling to
accumulate. Indeed it is well accepted that compression devices/stockings used in other areas of health care must be properly fitted and
graduated, to work effectively eg. in prevention of deep venous thrombosis (DVT). Notwithstanding this, soft tissue injuries can affect most
body parts, and it is difficult to design a bandage which conforms to all bony prominences and allows for anatomical variation. A number of
commercial supports have been designed to provide focal compression around the ankle, similarly felt padding can be used to fill in gaps
around body prominences, and focus compression at certain tissues. This approach seems pragmatic; horseshoe shaped felt around the ankle
malleolus may prevent pressure peaks around the bony prominences11 at the ankle, and may prevent oedematous accumulation around the
more pain sensitive injured ligaments. Currently, there is little supporting clinical evidence for focal compression.4

Compression, body position and ambulation

Elevation and compression are commonly applied together in clinical practice; the rationale is usually to achieve an enhanced effect on local
tissue pressure gradients. The previous guidelines1 found some evidence that elevation alone reduced ankle volume more effectively, than a
combination of elevation and compression post sprain. There was low quality evidence that both elevation alone, and concomitant compression
and elevation significantly decreased limb volume after acute ankle sprain, however there were trends towards larger decreases with elevation
only.4 Some12 have suggested that combining elevation and compression induces too great a change in interstitial pressure which collapses
some of the finer lymphatic capillaries, adversely affecting their drainage function. Clearly, pressure profiles within the venous and lymphatic
system will vary according to body position. Compressive force may therefore need to be reduced in a lying or elevated position.
There is both retrospective and prospective evidence that intermittent compression is most effective at resolving oedema post foot fracture.13,14
Clinical evidence shows moderate evidence that semi-rigid supports are more effective than elastic bandages after ankle sprain;4 the physiological
reasons for this are not clear. Some semi-rigid ankle supports are designed that their external pressure increases on weight-bearing which
creates an oscillating force (from proximal to distal) on the vasculature and lymphatic systems during ambulation. It is an interesting concept
that walking and movement may be needed to maximise the effectiveness of some compression devices.

Blood flow velocity and shunting

Compression is used effectively in many areas of health care to treat venous insufficiency, prevention of ulcer recurrence and DVT.15 The most
popular mode is graduated stockings (GCS) which increase blood flow velocity, and shunt blood from the superficial to the deep venous
system.16 The mechanical pressures generated with GCS range from 14-17mmHg (Class 1) up to 25 to 35 mmHg (Class 3);16 these values are
quite similar to those currently used and recommended after acute soft tissue injury. Increased blood flow velocity and venous shunting may
only be warranted at certain times after acute injury. Clinically related evidence in this area is limited to EIMD; in these models, the immediate
use of graduated compression garments (10-15 mmHg) had an inconsistent effect on biomarkers of muscle damage and inflammation.6 Note
we did not consider their effects on sports recovery per se within the scope of these review guidelines.

Optimal compression
There is little evidence to suggest an optimal level of compression pressure, or whether a constant or intermittent approach is preferable.
This perhaps depends on our desired physiological rationale which may include: stopping or minimising the initial tissue haemorrhage;
preventing / reversing the accumulation of oedema within the interstitial space; or enhancing venous and lymphatic return. In all cases, too
little compression may not have the desired effect whereas too much could be deleterious. Perhaps ideally, the level of compression should be
guided by the tissue pressures within the injured tissue.12 It is unlikely that one approach (mode, dosage) will be equally effective across the soft
tissue injury spectrum. It is most likely that factors such as: distance of the injured body part from the heart; injury severity and depth; time after
injury; tissue vascularity and depth of the injured tissue, should influence our choice and dosage of compression, and elevation.

We have already discussed the effects that compression can have on blood flow velocity. Intermittent or sequential compression pressures may
maximise this effect, with increases in flow of up to 200%.17 Such large changes can induce a number of effects at a tissue level including:
increased shear stress, arteriovenous pressure gradient and peripheral resistance, with subsequent increases in proteins, nutrients, and growth
factors to the injury site. Shear stresses on the endothelial walls also stimulate nitric oxide production, causing further vasodilatation. Again, the
optimal time after injury for inducing these effects should be considered.
The importance of introducing controlled mechanical loading during rehabilitation has already been discussed. There is potential that
compression (particularly intermittent compression) may provide another medium for inducing cyclic loading during recovery.17 Related animal
models have found that early cyclic compression is effective for recovery and reduces certain aspects of the inflammatory response after EIMD.6
In other clinical populations, compressive garments have been used to assist or re-establish motor functioning;18 the mechanism is thought to
relate to a combination of biomechanical or neurophysiological effects. Currently there is preliminary evidence that compression garments can
improve motor function after EIMD.19 Although supervised rehabilitation is regarded as the primary method of re-training motor control and
sporting technique after injury, compressive supports may be a useful adjunct; based on their biomechanical support, and potential to enhance
cutaneous stimulation, sensory awareness and joint positional sense.
Swelling and joint effusion adversely affect sensori-motor control, joint function and range of movement after injury. These factors can prevent
progression of rehabilitation, particularly therapeutic exercise.8 Recent evidence suggests that compression and/or elevation have little long term
effect on tissue swelling, particularly in gravity dependent positions.4 Nonetheless, there could still be potential to use elevation or compression
to produce short term reductions in swelling/effusion, providing a therapeutic window to maximise the effect of rehabilitation exercises.
In recent years, various kinesiology taping techniques have become a popular method of managing oedema after acute soft tissue injury. The
application of tape usually follows the lymphatic anatomy; the rationale is to lift superficial tissue, creating space for optimal lymphatic function.
This seems to contradict the traditional approach of applying an external compressive force on an injury. It is unclear which of the approaches
is superior, or whether they can be used interchangeably or simultaneously.
The clinical effectiveness of other methods of equilibrating tissue pressures, and enhancing venous or lymphatic flow should also be considered.
Exercise (either systemic or isolated movement of a body part), creates intra-thoracic pressure changes and induces the skeletal muscle pump
which may enhance venous haemodynamics and lymphatic function after injury. The success of this approach may again be time dependent
(after injury), but aligns well with functional treatment, and the graduated transition to controlled tissue loading.

GENERAL RECOMMENDATIONS FROM PANEL CONSENSUS

-Definitely unload (Protection/Rest) soft tissue in the acute phases after injury. As a minimum, unloading should definitely include avoiding
movements in the plane of injury during the acute phases after injury. (1)
-Progressive mechanical loading should definitely be initiated after the acute phases. For the majority of soft tissue injuries, complete tissue
unloading (Protection and Rest) beyond the acute phases after injury should definitely be avoided. (2)
-Definitely apply ice after an acute soft tissue injury (3). In the absence of acute injury, it is probably safe to apply brief applications of ice for
pain relief before returning to exercise/sporting activity, without increased risk of injury or performance detriment (4).
-Probably use elevation after an acute soft tissue injury. After any elevation definitely return the body part to a gravity dependent position
gradually. (5)
-Probably use compression after an acute soft tissue injury. Probably DONT use high levels of compression with simultaneous elevation. (6)

Specific Recommendations And Key Clinical Implications

1)
2)

3)

4)

5)

6)

The duration of unloading (Protection/Rest) definitely depends on the severity of the injury. We suggest that sustained or repetitive low
level loads into the injuring plane should probably be avoided, and that external supports or bracing will help compliance with this.
We recommend that the details of mechanical loading and speed of progression are guided by: the severity of the injury, the injuring
mechanism (excessive tensile vs compressive forces), and the type of soft tissue injury (ligament vs tendon vs muscle). This approach
should definitely be supervised by a Chartered Physiotherapist.
We suggest that the primary clinical effect is a reduction in pain; reducing inflammation, swelling or secondary injury is not guaranteed,
and depends on the depth of injury, body part and % of overlying body fat. We recommend using crushed ice for at least 10 minutes to
reduce pain. A thin damp barrier at the cooling interface will definitely not mitigate the cooling effect. Dry thick bandaging will
definitively mitigate the cooling effect and clinical effectiveness. We recommend avoiding continuous application for longer than 20-30
minutes; we suggest that the optimal time between ice applications should be guided by pain and discomfort levels and can be less
than 2 hours. It is difficult to give specific recommendations on an optimal cooling dosage for effectively inducing deep tissue
temperature reduction; we suggest intermittent applications are the safest approach.
In most sporting environments, it is best practice to undertake a game related warm up prior to initiating or returning to physical activity
or sport. We would recommend that a short game related warm up is undertaken between completing a short period of icing
(<10 minutes), and returning to return to physical activity / sport.
There is no optimal duration of elevation; we suggest that distal body segments require longer periods. During periods of elevation
patients will probably feel less pain and discomfort as a result of lower tissue pressures. Tissue/limb girth will probably return to its pre
elevation girth even with a gradual return to a dependant position.
If the clinical aim is to reduce the net fluid loss from the vascular system after injury, then the compressive force of the bandage/
compressive modality should exceed interstitial fluid pressure. This value will definitely depend on the type of injury and affected body
part, and it is difficult to provide specific recommendations on an optimal pressure. We suggest using focal compression around bony
protuberances, and we recommend that the compressive modality configures to the shape of the body part, and provides a graduated
compressive force. Compression bandages and external supports could contribute to recovery through a combination of physiological
and therapeutic pathways. We suggest that most modes of compression provide additional benefits in terms of biomechanical support,
controlling range of movement, and reassurance after injury. In some cases, external support may simply be needed to corroborate the
severity of an injury to a patient, particularly in environments where they might risk an aggressive return to function.

Strong recommendation: We recommend or definitely

Weak recommendation: We suggest or probably
Values and preferences relevant to each recommendation are summarised in Figure 8.
We acknowledge that these recommendations are subject to change based on further research (Figure 8)

Figure 7

EVIDENCE PROFILE

Summary of clinical evidence profiles by type of soft tissue injury (The effect of PRICE on critically important outcomes)

High

Mod

Low

V.
Low

Functional
>IMM
(1) (+)

Short IMM =
Long IMM

Lig/Jt
PROTECTION / REST /
LOADING

Effective analgesia
post op
(2) (4) (+)
Effective analgesia
post ankle sprain
(2) (3)

Lig/Jt
ICE

Semi
Rigid>Elastic
(-)
No effect
DTG (-)
No effect
with
80mmHg
M

Elev > Comp (5)

Elev does not prevent

limb swelling (4) (5)
Lig/Jt
COMPRESSION

Lig/Jt
ELEVATION

M: Muscle injury; Lig/Jt; ligament or joint injury; T: Tendon injury

A > B: Treatment A is more effective than treatment B; A=B: Treatment B is equally as effective as treatment B; - (dash): No relevant clinical evidence
1). No differences in long term function; 2). No effect on other important clinical outcomes; 3). Immediate return to sport after cooling (>5 minutes) decreases
performance in some correlates of sporting performance ; 4) Post surgical model; 5) Based on Limb girth outcome
(+) Supporting evidence from basic science research; (-) Potential conflicting evidence from basic science research
EVIDENCE PROFILE:
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

What we know

BASIC SCIENCE

-A period of protection and rest is

required after the majority of soft
tissue injuries. Excessive protection/rest
and unloading will do harm (A).
-Mechanical loading / ambulation
after injury helps restore strength
and morphological characteristics of
collagenous tissue after tendon
injury (A).
-Mechanical loading can upregulate
gene expression for some proteins
associated with tendon tissue
healing (A).
-Excessive loading in the early stages
after injury will do harm (A).
-Exercise (either systemic or isolated
movement of a body part), creates
intra-thoracic pressure changes and
induces the skeletal muscle
pump which enhances venous
haemodynamics and lymphatic
function.

PROTECTION / REST /
TISSUE LOADING
COMPRESSION
-Provides support and confidence for
many patients. Also exerts an external
force on the tissue which could
decrease bleeding; and/or prevent
accumulation of oedema within the
interstitial space.
-Commonly used compression
bandages exert pressures between 15
and 60mmHg.
-In other areas of health care,
graduated compression stockings,
exerting pressures between 5 and
30mmHg; are used to increase local
blood flow, and shunt blood from the
superficial to deep venous system.
-Cyclic loading (massage) reduces
aspects of inflammation and is
effective for recovery after EIMD (A).
-Light compression garments have
an inconsistent effect on markers of
muscle damage, and some markers
of inflammation after EIMD (IH).

ICE
-Skin temperature can be reduced to
<13C within 5-15 minutes. Not all modes
of cooling are equally effective; crushed
ice provides effective cooling and is
probably safest. (HH, CCS)
-A damp barrier will not mitigate the
cooling effect on the skin; however a thick,
dry barrier will mitigate excessively and
limit clinical effectiveness. (HH)
-Excessive skin / tissue cooling results in
harm (CCS).
-Muscle and joint temperature reductions
are much smaller compared to those
obtained at the skin. The lowest
temperature recorded at a 1cm intramuscular (I/M) depth was approximately
20C; this was in a very lean population
(around <10mm skin fold), after 20
minutes of cooling (HH). Few studies have
reduced I/M temperature below 25C.
(HH)
Fat creates an insulating effect and limits
deep tissue temperature reductions (HH).
Deep tissue temperature remains the same
or continues to cool for up to 10 minutes
after ice pack removal (HH).
Ice reduces blood flow, reduces oxygen
saturation and facilitates capillary outflow
in healthy tendons (HH). It has an
inconsistent effect on microcirculation in
injured animals (A).
Ice has a consistent effect on cellular
and physiological events associated with
inflammation (A)
Short periods of cooling can have an
excitatory effect on muscle activation
(HH, IH).

PRICE and soft tissue injury: Risks, burdens and costs, what we know and what we dont know

Figure 8

-Elevating a body part decreases the

gravitational force on the column of
blood between that body part and the
heart, thereby decreasing hydrostatic
pressure, and facilitating venous and
lymphatic flow.

ELEVATION

RISKS, BURDENS
AND COST

WHAT WE KNOW

CLINICAL

-Ice application can effectively decrease

pain after surgery (High) and acute ankle
sprain (Mod).
-Removal for an ice pack for a short period
of time is safe and
effective (Low).
-Ice has little effect on recovery time,
function, swelling (Mod), but there are few
studies on closed soft
tissue injuries.
-Short periods of cooling can safely precede
physical activity; longer periods (>5-15
minutes), particularly over areas with little
body fat, could risk re-injury and affect
sporting performance (HH).

Risks: Direct side effects: skin burn/ nerve

palsy, non-specific advice from practitioner,
unable to tolerate cold sensation, difficulty
targeting deeper injuries, premature return
to activity/sport after excessive cooling, wide
range of treatment modes
Burdens: Patient compliance required,
crushed ice requires preparation for home
use, repetitive removal/reapplication,
melting ice can be messy, tolerating cold
sensation, Cost: Variable: higher relating to
commercial ice machine, gel packs have a
lower cost, crushed ice
is minimal.

-Protection/rest is a non specific term

which is interpreted as casting (with or
without weight-bearing), partial weightbearing, activity modification, external
support, bracing, bandaging or taping.
-Functional treatment (early
mobilisation/ambulation with a support)
is more effective than immobilisation/
non weight-bearing after acute ankle
sprain (Mod).
-Semi-rigid supports are more effective
than elastic bandages when used as an
adjunct to functional treatment (Mod).

Risks: Excessive or inadequate PR;

misunderstanding of what PR should
entail; currently non-specific terminology
relating to different advice from
practitioner, patient misinterpretation;
difficulty progressing into the active
rehabilitation stage, potential loss of
general fitness
Burdens: Potential lack of
independence, loss of earnings, patient
discipline, boredom, safe progression
into active loading/rehabilitation
requires increased therapist input/time
Cost: Variable: high relating to bracing,
crutches, casting; low relating to slings,
bandages, lower relating to unloading/
avoidance

-Elevation had no effect on upper

limb swelling after surgery (Low).
It decreased lower limb swelling
after ankle sprain (Low); this quickly
returns to baseline (pre-elevation
volumes) on return to a gravity
dependent position (Low).

Risks: Potential rebound, may

encourage excessive unloading, potential difficulty progressing into the
active rehabilitation stage, potential
loss of general fitness
Burdens: Potential lack of independence, loss of earnings, patient
discipline, boredom
Cost: No direct cost

-Light tubular bandaging is ineffective

after ankle sprain (Low).
-Maximal compression bandaging is
ineffective after muscle contusion (V.
Low).
-Semi-rigid supports are more effective
than elastic bandaging (Mod).
-Compression and elevation is less
effective than compression alone in
terms of reducing swelling (Low).

Risks: Excessive or inadequate compression; non-specific advice from

practitioner; wide range of treatment
options and associated pressures
Burdens: potential discomfort during
wear, repetitive removal/reapplication; pain with removal/reapplication, in some cases application skill is
important for ensuring effectiveness
and patient education or increased
therapist input/time is required
Cost: Variable: high relating to
pneumatic devices; low relating to
bandaging, however potential for
a high cumulative cost (eg. within a
department)

-The majority of basic science research

is based on injured tendon models
(surgical resection).
-How to make a safe transition from
protection/rest to progressive loading?
-The most effective form of external
support to compliment functional
management?
-The optimal timing, dosage, nature
and speed of induced load? And
whether this should vary according to
muscle, tendon or ligament tissue?

-The effect of cooling on key events

associated with inflammation eg.
oxidative stress?
-Whether the current threshold
temperatures for achieving analgesia
(skin temperature of <13C) or metabolic
reduction (reduction to 5-15C at the
injury site) will change based on emerging
understanding of inflammation?
-The magnitude or depth of temperature
reductions obtainable in smaller joints with
less surrounding adipose tissue (eg. wrist,
elbow, ankle), in the absence of tourniquet
and post operative dressings?
-The clinical relevance of inducing small or
moderate temperature reductions on deep
tissue in terms of inflammation, / local
cellular metabolism?
-The benefits and safety of using higher
dosages of cooling (eg. longer durations,
but with regular short periods of removal)
to target deep injuries?
-If ice can decrease the magnitude of
the inflammatory response or selectively
influence aspects of inflammation enough,
to have a clinical benefit after a significant
injury across all soft tissue types?
-Optimal mode of compression, its
associated force, and if this should
be constant or intermittent? Clear
physiological goal eg. preventing
bleeding, increasing interstitial
pressure, increasing venous return
or venous shunting? Whether this
should vary according to time from
injury, type of tissue injured?
-The clinical relevance of other
potential mechanisms of effect
including: limiting movement,
enhancing adherence with advice
(rest), enhanced sensori-motor
function, cyclic loading?
-If a temporary reduction in
swelling could provide a therapeutic
window for exercise?

-If elevation improves patient

comfort? Or decreases the risk of
adverse effects?
-How a temporary reduction in
tissue girth relates to actual joint
effusion?
-Optimal duration or angle of
elevation? Whether this should
vary according to the type of tissue
injured, or the body part?
-If a temporary reduction
in swelling could provide a
therapeutic window for exercise?

A: Based on evidence derived from animal models; HH: Based on evidence derived from un-injured human models; IH: Based on evidence derived from lab induced injury human
V. Low: Based on Very low quality evidence from injured human model; Low: Based on low quality evidence from injured human; Mod: Based on moderate quality evidence from injured human; High:
Based on high quality evidence from injured human; CCS: Clinical case study; EIMD: exercise induced muscle damage

-THERE IS A PARTICULAR DEARTH OF RELEVANT RESEARCH BASED ON ACUTELY INFLAMED TENDONS, AND MUSCLE STRAINS OR CONTUSION.
-WE ACKNOWLEDGE THAT OUR CURRENT RECOMMENDATIONS ARE SUBJECT TO CHANGE BASED ON FURTHER RESEARCH IN ANY OF THESE OUTLINED AREAS

NOTE:

WHAT WE
DONT KNOW

References
1. Kerr KM, Daley L, Booth L for the Association of Chartered Physiotherapists in Sports and Exercise Medicine (ACPSM). Guidelines for the
management of soft tissue (musculoskeletal) injury with protection, rest, ice, compression and elevation (PRICE) during the first 72 h.
London: Chartered Society of Physiotherapy, 1998.
2. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. for the GRADE Working Group. GRADE: an emerging
consensus on rating quality of evidence and strength of recommendations. British Medical Journal. 2008;336(7650):924-6.
3. Bleakley CM, Glasgow PD, Philips N, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists in
Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and Elevation,
2011. Chapter 5: What effect does mechanical loading have on inflammation and soft tissue healing after acute injury?
4. Bleakley CM, Glasgow PD, Philips N, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists in
Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and Elevation,
2011. Chapter 7: Which components of PRICE are effective in the clinical management of acute soft tissue injury?
5. Bleakley CM, Glasgow PD, Philips N, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists in
Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and Elevation,
2011. Chapter 3: What is the magnitude and depth of cooling associated with ice?
6. Bleakley CM, Glasgow PD, Philips N, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists in
Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and Elevation,
2011. Chapter 4: Can PRICE decrease the inflammatory response after acute soft tissue injury?
7. Scott A, Khan KM, Roberts CR, Cook JL, Duronio V. What do we mean by the term inflammation? A contemporary basic science
update for sports medicine. British Journal of Sports Medicine 2004;38(3):372-80.
8. Bleakley CM, Glasgow PD, Philips N, Hanna L, Callaghan MJ, Davison GW et al. for the Association of Chartered Physiotherapists in
Sports and Exercise Medicine (ACPSM). Management of acute soft tissue injury using Protection Rest Ice Compression and Elevation,
2011. Chapter 6: Do the physiological effects of local tissue cooling affect function, sporting performance and injury risk?
9. Thomas S. Bandages and bandaging: the science behind the art. Care Science and Practice 1990; 8 (2): 56-60.
10. Mayrovitz H, Delgado M, Smith J. Compression bandaging effects on lower extremity peripheral and sub-bandage skin perfusion.
Ostomy Wound Management 1998; 44 (3): 56-67
11. Cullum N, Roe B. Prevention of re-ulceration. In: Keachy J, ed. Leg Ulcers: Nursing Management: A Research-based guide. Balliere
Tindall, London.
12. Tsang KK, Hertel J, Denegar CR. Volume decreases after elevation and intermittent compression of postacute ankle sprains are negated
by gravity-dependent positioning. Journal of Athletic Training. 2003;38(4):320-324.
13. Thordarson DB, Greene N, Shepherd L, Perlman M. Facilitating edema resolution with a foot pump after calcaneus fracture. Journal of
Orthopaedic Trauma 1999; 13(1): 43-46.
14. Myerson MS, Juliano PJ, Koman JD. The use of a pneumatic intermittent impulse compression device in the treatment of calcaneus
fractures. Military Medicine 2000; 165(10): 721-725
15. Sachdeva A, Dalton M, Amaragiri SV, Lees T. Elastic compression stockings for prevention of deep vein thrombosis. Cochrane Database
of Systematic Reviews 2010, Issue 7. Art. No.: CD001484.
16. Bradley L. Venous haemodynamics and the effects of compression stockings. British Journal of Community
Nursing 2001;6 (4): 165-175.
17. Khanna A, Gougoulias N, Maffulli N. Intermittent pneumatic compression in fracture and soft tissue injuries healing. British
Medical Bulletin 2008; 88: 147-156.
18. Attard J, Rithalia S. A review of the use of Lycra pressure orthoses for children with cerebral palsy. International Journal of Therapy and
Rehabilitation 2004; 11: 120-6.
19. Pearce AJ, Kidgell DJ, Grikepelis LA, Carlson JS. Wearing a sports compression garment on the performance of visumotor tracking
following eccentric exercise: A pilot study. Journal of Science of Medicine 2009: 12: 500-502.

References for Figure 6

-Bender AL, Kramer EE, Brucker JB, Demchak TJ, Cordova ML, Stone MB. Local ice-bag application and triceps surae muscle temperature during treadmill walking.
Journal of Athletic Training 2005; 40(4): 271-275.
-Dykstra JH, Hill HM, Miller MG, Cheatham CC, Michael TJ, Baker RJ. Comparisons of cubed ice, crushed ice, and wetted ice on intramuscular and surface temperature
changes. Journal of Athletic Training 2009;44(2):136-41.
-Jutte LS, Merrick MA, Ingersoll CD, Edwards JE. The relationship between intramuscular temperature, skin temperature, and adipose thickness during cryotherapy and
rewarming. Archives of Physical Medicine and Rehabilitation 2001;82(6):845-50.
-Myrer JW, Measom G, Fellingham GW. Temperature changes in the human leg during and after two methods of cryotherapy. Journal of Athletic Training
1998;33(1):25-29.
-Myrer JW, Measom GJ, Fellingham GW. Exercise after cryotherapy greatly enhances intramuscular re-warming. Journal of Athletic Training 2000;35(4):412-416.
-Myrer WJ, Myrer KA, Measom GJ, Fellingham GW, Evers SL. Muscle temperature is affected by overlying adipose when cryotherapy is administered. Journal of Athletic
Training 2001 Mar;36(1):32-36.

Appendix Table 1:
Search Strategy 1

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.

*Cryotherapy/
cold compress.mp.
ice bath$.mp.
ice pack.mp.
exp Cold Temperature/th, mt, et, ae, me [Therapy, Methods, Etiology, Adverse Effects, Metabolism] (3068)
cryokinetic.mp.
cold pack$.mp.
exp Ice/
rice.mp.
PRICE.mp.
OR/ 1-10
H-Reflex/ or arthrogenic muscle response.mp.
exp Proprioception/ph, cl [Physiology, Classification]
Kinesthesis/ or joint positional sense.mp.
Blood Circulation/ or Blood Flow Velocity/ or blood flow.mp.
Edema/ or swelling.mp.
Metabolism/ph [Physiology]
Lymphatic System/ph, pp, in, pa [Physiology, Physiopathology, Injuries, Pathology]
limb girth.mp.
range of movement.mp.
Muscle Strength/ph [Physiology]
Neural Conduction/ph [Physiology]
Hemodynamics/ph [Physiology]
Skin Temperature/ph [Physiology]
Body Temperature/ or Oxygen Consumption/ or muscle temperature.mp.
joint temperature.mp.
OR/12-26
and 11 (1482)
from 28 keep 208

MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trial s (Ovid)

Appendix Table 2:
Search Strategy 2

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.

exp Casts, Surgical/ or exp Immobilization/ or immobilisation.mp.

exp Bandages/
compression.mp. or Intermittent Pneumatic Compression Devices/ or Stockings, Compression/
pop.mp.
tubigrip.mp.
exp Sprains and Strains/
exp Ligaments/
exp Contusions/
exp Tendon Injuries/
exp Tendinopathy/
exp Soft Tissue Injuries/
exp Weight-Bearing/
elevation.mp.
Athletic Injuries/
Rest/
OR/1-5
OR/6-14
16 and 17
limit 18 to (english language and yr=1996 - 2009) (2947)
from 19 keep (131)

MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trial s (Ovid)

Appendix Table 3:
Search Strategy 3
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.

exp Cryotherapy/
cold compress.mp.
ice pack.mp.
Cold Temperature/
exp Ice/
PRICE.mp.
exp Sprains and Strains/
exp Contusions/
exp Tendon Injuries/
exp Soft Tissue Injuries/
exp Athletic Injuries/
OR/1-6
OR 7-11
12 and 13
limit 14 to (english language and humans and yr=1996 -Current) (139)
from 15 keep (11)

MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trial s (Ovid)

Appendix Table 2:
Search Strategy 2
Related articles search on Pubmed (http://www.ncbi.nlm.nih.gov/pubmed/)
Reference

Number of related titles read

Pollard (2005)

145

Tsang (2003)

96

Agalfy (2007)

183

Hopper (1997)

190

Wassinger (2007)

119

Hopkins (2002)

225

Kubo (2005)

472

De Ruiter (2001)

202

Kimura (1997)

120

Yanagisawa (2007)

816

Glenn (2004)

146

Karunakara (1999)

244

Knobloch (2008)

94

Weston (1994)

116

Schaser (2007)

321

Howatson (2003)

184

Davies (2009)

291

Kraemer (2001)

468

Citation tracking
Undertaken on all incoming primary and review articles n=350

Appendix 4b
Literature searches from related reviews, previously completed by the authors

1. MeSH descriptor Cryotherapy, this term only

2. MeSH descriptor Cold Temperature, this term only
3. MeSH descriptor Immersion, this term only
4. #2 AND #3
5. #1 OR #4
6. ((cold or ice or contrast) NEAR/3 (immers* or bath or therapy)):ti,ab,kw
7. (cryotherapy):ti,ab,kw
8. (#5 OR #6 OR #7
9. MeSH descriptor Exercise, this term only
10. MeSH descriptor Muscle, Skeletal, this term only
11. MeSH descriptor Athletic Injuries, this term only
12. MeSH descriptor Soft Tissue Injuries, this term only
13. MeSH descriptor Creatine Kinase explode all trees
14. MeSH descriptor Physical Exertion, this term only
15. MeSH descriptor Muscle Fatigue, this term only
16. MeSH descriptor Muscle Cramp, this term only
17. MeSH descriptor Spasm, this term only
18. MeSH descriptor Muscle Rigidity, this term only
19. MeSh descriptor Sprains and Strains, this term only
20. MeSH descriptor Muscle Weakness, this term only
21. (sore* NEAR/3 musc*):ti,ab,kw
22. (DOMS):ti,ab,kw
23. (exercise induced and (muscle* NEAR/2 (damage* or injur*))):ti,ab,kw
24. MeSH descriptor Lactic Acid, this term only
25. (lactate* or lactic):ti,ab,kw
26. (#9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR#21
OR #22 OR #23 OR #24 OR #25)
27. (#8 AND #26)
Undertaken on Cochrane Central Register of Controlled Trials (Wiley InterScience interface), and modified for
Medline, EMBASE (Ovid)

Appendix Table 5:
List of excluded studies with reasons
Chapter 3: What is the magnitude and depth of cooling associated with ice/ice and compression?
Study

Reason for exclusion*

Akgun K, et al. Temperature changes in superficial and deep tissue

layers with respect to time of cold gel pack application in dogs. Yonsei
Medical Journal 2004;45(4):711-8.

Animal subjects

Barlas D, et al. In vivo tissue temperature comparison of cryotherapy

with and without external compression. Annals of Emergency Medicine
1996;28(4):436-9.
Kim YH, et al. The effect of cold air application on intra-articular and skin
temperatures in the knee. Yonsei Medical Journal 2002;43(5):621-6.
Pollitt C, et al. Prolonged, continuous distal limb cryotherapy in the horse.
Equine Veterinary Journal 2004; 36(3):216-20.
Daanen HA, et al. The effect of body temperature on the hunting
response of the middle finger skin temperature. Eur J Appl Physiol Occup
Physiol 1997;76(6): 538-43.

No measure of temperature /
insufficient outcome measure

Rubley MD, et al. Time course of habituation after repeated ice bath
immersion of the ankle. Journal of Sports Rehab 2003.
Singh H, et al. The efficacy of continuous cryotherapy on the
postoperative shoulder. J Shoulder Elbow Surg 2001;10: 522-25.
Speer KP, et al. The efficacy of cryotherapy in the post operative shoulder.
J Shoulder Elbow Surg, 1996;5: 62-8.
Anvari B, et al. A comparative study of human skin thermal response
to sapphire contact and cryogen spray cooling. IEEE Transactions on
Biomedical Engineering 1998;45(7):934-41.

Cold therapy intervention not

clinically relevant

Jay O, et al. Skin cooling on contact with cold materials: the effect of
blood flow during short-term exposures. Annals of Occupational Hygiene
2004;48(2):129-37.
Enwemeka CS, et al. Soft tissue thermodynamics before, during and after
cold pack therapy. Med Sci Sports Exerc 2002;34(1):45-50.

Unable to determine the depth at

which temperature was recorded

Muraoka T, et al. Effects of muscle cooling on the stiffness of the human

gastrocnemius muscle in vivo. Cells Tissues Organs. 2008;187(2):15260. Epub 2007 Oct 15.
Nosaka K, et al. Influence of pre-exercise muscle temperature on
responses to eccentric exercise. J Athl Train 2004;39(2):132-137.
Yanagisawa O, et al. Effects of cooling on human skin and skeletal
muscle. Eur J Appl Physiol 2007;100(6):737-45. Epub 2007 May 4.

Lindsell C, et al. Interpretation of the finger skin temperature response to

cold provocation. International Archives of Occupational & Environmental
Health 2001;74(5):325-35.

Hand immersion, and finger tip

temperature recorded only

Holcomb WR, et al. The effect of icing with the pro stim edema
management system on cutaneous cooling. J Athl Train 1996;31(2):
126-9

Concomitant electrical stimulation

Chapter 4: What effect does PRICE have on the inflammatory response after acute soft tissue injury?
Study

Reason for exclusion*

Fu FH, et al. The effects of cryotherapy on muscle damage in rats

subjected to endurance training. Scand J Med Sci Sports 1997;7(6):35862.

Injury model not relevant (animal)

Jia J, et al. Cold nerve injury is enhanced by intermittent cooling. Muscle

Nerve 1999;22(12):1644-52.
Locke M, et al. Cold stress does not induce stress proteins SP 25 and SP
72 in rat skeletal muscle. Cryobiology 2001;43(1):54-62.
Nemeth N, et al. Systemic and regional hemorheological consequences
of warm and cold hind limb ischemia-reperfusion in a canine model.
Clinical Hemorheology & Microcirculation 2004;30(2):133-45.
Thorlacius H, et al. Effects of local cooling on microvascular
hemodynamics and leukocyte adhesion in the striated muscle of
hamsters. J Trauma 1998;45(4):715-19.*
Wladis A, et al. Effects of induced hypothermia after soft tissue injury.
Arch Orthop Trauma Surg 2004;124(4):243-9. Epub 2003 Oct 7.
Bailey DM, Erith SJ, Griffin PJ, Dowson A, Brewer DS, Gant N, Williams
C. Influence of cold water immersion on indices of muscle damage
following prolonged intermittent shuttle running. Journal of Sports
Sciences 2007;25(11):1163-1170.
Halson SL, Quod MJ, Martin DT, Gardner AS, Ebert TR, Laursen PB.
Physiological responses to cold water immersion following cycling in
the heat. International Journal of Sports Physiology and Performance
2008;3:331-346
Ingram J, Dawson B, Goodman C, Wallman K, Beilby J. Effect of water
immersion methods on post-exercise recovery from simulated team sport
exercise. Journal of Science and Medicine in Sport 2009;12(3):417-421.
King M, Duffield R. The effects of recovery interventions on consecutive
days of intermittent sprint exercise. Journal of Strength and Conditioning
Research 2009;23(6):1795-1802.
Nemet D, Meckel Y, Bar-Sela S, Zaldivar F, Cooper DM, Eliakim A. Effect
of local cold-pack application on systemic anabolic and inflammatory
response to sprint-interval training: a prospective comparative trial. Eur J
Appl Physiol. 2009 Nov;107(4):411-7. Epub 2009 Aug 4.
Peake J, Peiffer JJ, Abbiss CR, Nosaka K, Okutsu M, Laursen PB, Suzuki
K. Body temperature and its effect on leukocyte mobilisation, cytokines
and markers of neutrophil activation during and after exercise. European
Journal of Applied Physiology 2008;102(4):391-401
Robey E, Dawson B, Goodman C, Beilby J. Effect of postexercise recovery
procedures following strenuous stair-climb running. Research in Sports
Medicine 2009; 17: 245-259.
Rowsell GJ, Coutts AJ, Reaburn P, Hill-Haas S. Effects of cold-water
immersion on physical performance between successive matches in
high-performance junior male soccer players. Journal of Sports Sciences
2009:1-9.
Ali A, Creasy RH, Edge JA. Physiological effects of wearing graduated
compression stockings during running. European Journal of Applied
Physiology 2010; 109: 1017-1025.
Trenell MI, Rooney KB, Sue CM, Thompson CH. Compression garments
and recovery from eccentric exercise: A 31P-MRS study. Journal of Sports
Science and Medicine 2006; 5: 106-114.

Injury / exercise model not relevant

(human)

Chapter 4: What effect does PRICE have on the inflammatory response after acute soft tissue injury?
Study

Reason for exclusion*

Duffield R, Portus M. Comparison of three types of full body compression

garments on throwing and repeat sprint performance in cricket players.
British Journal of Sports Medicine 2007; 41: 409-414

Injury / exercise model not relevant

(human)

Kraemer WJ, Volek JS, Bush JA, Gotshalk LA, Wagner PR, Gmez AL,
Zatsiorsky VM, Duarte M, Ratamess NA, Mazzetti SA, Selle BJ. Influence
of compression hosiery on physiological responses to standing fatigue in
women. Med Sci Sports Exerc. 2000 Nov;32(11):1849-58.
Duffield R, Edge J, Merrells R, Hawke E, Barnes M, Simcock D, Gill N. The
effects of compression garments on intermittent exercise performance and
recovery on consecutive days. International Journal of Sports Physiology
and Performance 2008;3(4):454-68.
Crowe MJ, OConnor D, Rudd D. Cold water recovery reduces anaerobic
performance. International Journal of Sports Medicine 2007;28(12):994998.

Outcomes not relevant

Heyman E, De Geus B, Mertens I, Meeusen R. Effects of four recovery

methods on repeated maximal rock climbing performance. Medicine &
Science in Sports and Exercise 2009;41(6):1303-1310.
Kuligowski LA, Lephart SM, Giannantonio FP, Blanc RO. Effect of whirlpool
therapy on the signs and symptoms of delayed onset muscle soreness.
Journal of Athletic Training 1998;33(3):222-228.
Yamane M, Teruya H, Nakano M, Ogai R, Ohnishi N, Kosaka M. Postexercise leg and forearm flexor muscle cooling in humans attenuates
endurance and resistance training effects on muscle performance and on
circulatory adaptation. Eur J Appl Physiol. 2006 Mar;96(5):572-80.
Kraemer WJ, Bush JA, Triplett-McBride NT, Koziris LP, Mangino LC et al.
Compression garments: Influence on muscle fatigue. Journal of Strength
and Conditioning Research, 1998; 12(4): 211-215.
Kraemer WJ, Bush JA, Newton RU, Duncan ND, Volek JS, Denegar CR
et al. Influence of a compression garment on repetitive power output
production before and after different types of muscle fatigue. Sports
Medicine, Training and Rehabilitation 1998; 82(2): 163-184.
Schaser, et al. In vivo analysis of micro circulation following closed soft
tissue injury. J Orthop Research 1999;17: 678-85.

No cold therapy intervention

Chapter 6: What other physiological effects are associated with tissue cooling, and do they affect sporting
performance and injury risk?
Bornmyr S, et al. Effect of local cold provocation on systolic blood pressure
and skin blood flow in the finger. Clin Physiol 2001;21(5):570-5.
Oksa et al. Combined effect of repetitive work and cold on muscle
function and fatigue. J Appl Physiol 2002; 92:354-361.
Petrofsky J. Muscle temperature and EMG amplitude and frequency during
isometric exercise. Aviat Space Environ Med 2005;76(11):1024-30.
Shibahara N, et al. The responses of skin blood flow, mean arterial
pressure and R-R interval induced by cold stimulation with cold wind and
ice water. Journal of the Autonomic Nervous System 1996;61(2):109-15.
Yona et al. Effects of cold stimulation of human skin on motor unit activity.
Jpn J Physiol 1997; 47:341- 8

Cold therapy intervention no relevant

Bandrivskyy A, et al. Wavelet phase coherence analysis: Application

to skin temperature and blood flow. Cardiovascular Engineering
2004;4(1):89-93.

Outcomes not relevant

Doering TJ, et al. Cerebral hemodynamics and cerebral metabolism

during cold and warm stress. American Journal of Physical Medicine &
Rehabilitation 1996;75(6):408-15.
Yanagisawa O, et al. The use of resonance imaging to evaluate the effects
of cooling on skeletal muscle after strenuous exercise. European Journal
of Applied Physiology 2003;89(1):53-62.
Bergersen, et al. Blood velocity and skin temperature during cold induced
vasodilatation. Journal of Physiology 1998;509. [Bergersen, 1999]

Double data publication

De Ruiter CJ, et al. Similar effects of cooling and fatigue on eccentric and
concentric force-velocity relationships in human muscle. J Appl Physiol
2001;90(6):2109-16.

Twitch contraction on muscle strength

outcome

De Ruiter CJ, et al. Temperature effect of the force/velocity relationship

of the fresh and fatigued human adductor pollicis muscle. Pflugers Arch
2000;440(1):163-70.
De Ruiter CJ, et al. Temperature effect on the rates of isometric force
development and relaxation in the fresh and fatigued. Exp Physiol 1999;
84:1137-50.
Gossen ER. Effect of temperature on post-tetanic potentiation in human
dorsiflexor muscles. Can J Physiol Pharmacol 2001;79(1): 49-58.
Yanagisawa O, et al. Evaluations of cooling exercised muscle with MR
imaging and 31MR spectroscopy. Med Sci Sports Ex 2003;35(9):1517-23.
Mutungi et al. Temperature dependent changes in the viscoelasticity of
intact resting mammalian (rat) fast and slow twitch muscle fibres. J Physiol
1998;508: 253-65.

DOMS study

In vitro

Jia, J et al. Cold nerve injury is enhanced by intermittent cooling. Muscle

& Nerve 1999; 22(12):1644-52.
Zhou L et al. Mechanism research of cryoanalgesia. Neurological
Research 1995;17(4):307-11.
Gao et al. 2000. Mechanism of cryoinjury in living cells

Induced injury at cellular level using

cold therapy

Gallen KG et al. Medicine & Science in Sports & Exercise 1999;

31(5):S295.

Platform presentation

Marvar PJ et al. Medicine & Science in Sports & Exercise 2002;

34(5):S100.
McCrory JL et al.Medicine & Science in Sports & Exercise 2005;
37(5):S278-S279.
Ransone, J. W. Medicine & Science in Sports & Exercise 1999; 31(5):S145.
Rogers J T, et al. Medicine & Science in Sports & Exercise 2003;
35(5):S265.
Wang HE et al. Medicine & Science in Sports & Exercise 2008;
40(5):S215.
doi: 10.1249/01.mss.0000322385.57749.b3
Weimar W et al. Medicine & Science in Sports & Exercise 2004;
36(5):S187.
Wells B L et al. Medicine & Science in Sports & Exercise 2002; 34(5):S101.
Wilson SM et al. Medicine & Science in Sports & Exercise 2006;
38(5):S375.

Chapter 5: What effect does mechanical loading have on inflammation and healing after acute
soft tissue injury?
Study

Reason for exclusion*

Demirhan M, et al. Tensile strength of ligaments after thermal shrinkage

depending on time and immobilization: in vivo study in the rabbit. J
Shoulder Elbow Surg 2005;14(2):193-200.

Not soft tissue injury

Eliasson P, et al. Unloaded rat Achilles tendons continue to grow but lose
viscoelasticity. J Appl Physiol 2007;103(2):459-63. Epub 2007 Apr 5.

Animal study not injured

Kanda K, et al. Adverse effect on rabbit anterior cruciate ligament after

knee immobilisation: changes in permeability of horseradish peroxidise.
Arch Orthop Trauma Surg 1998;117:307-311.
Matsumoto F, et al. Mechanical effects of immobilisation on the achilles
tendon. Arch Phys Med Rehabil 2003;84(5):662-7.
Narmoneva DA, et al. Altered swelling behaviour of femoral cartilage
following joint immobilisation in a canine model. J Orthop Res
2002;20(1):83-91.
Thornton GM, et al. Ligament creep recruits fibres at low stresses and can
lead to modulus-reducing fibre damage at higher creep stresses: a study
in rabbit medial collateral ligament model. J Orthop Res 2002;20(5):96774.
Gelberman RH, et al. The effect of gap formation at the repair site on the
strength and excursion of intrasynovial flexor tendons. J Bone Joint Surg
Am 1999;81(7):975-82.

Animal study surgical repair

Palmes D, et al. Achilles tendon healing: long term biomechanical effects

of post operative mobilisation and immobilisation in a new mouse model.
Journal of Orthopaedic Research 2002;20: 939-946.
Green DM, et al. Effect of early full weight bearing after joint injury
on inflammation and cartilage degradation. J Bone Joint Surg Am
2006;88(10):2201-9.
Hurschler C, et al. Scanning electron microscopic characterisation of
healing and normal rat ligament microstructure under slack and loaded
conditions. Connect Tissue Res 2003;44(2):59-68.

Osteochondral lesion

No or inadequate control group

Provenzano PP, et al. Healing of subfailure ligament injury: comparison

between immature and mature ligaments in a rat model. J Orthop Res
2002;20(5):975-83.
Yasunda T, et al. Unfavourable effect of knee immobilization on Achilles
tendon healing in rabbits. Acta Orthop Scand 2000; 71(1): 69-73.
Chapter 7. Clinical effectiveness of PRICE
Guillodo Y, et al. Treatment of muscle trauma in sports people (from
injury on the field to resumption of the sport). Ann Phys Rehabil Med.
2009;52(3):246-55. Epub 2009 Feb 23.

No relevant outcomes

Kullenberg B, et al. Post operative cryotherapy after total knee arthroplasty:

a prospective study of 86 patients. J Arthroplasty 2006;21(8):1175-9.
Kerkhoffs GM, et al. Functional treatments for acute ruptures of the lateral
ankle ligament. Acta Orthop Scand. 2003;74(1):69-77.
Kerkhoffs GM et al. -Immobilisation of acute ankle sprain. Arch Orthop
Trauma Surg 2001;121(8):462-71
Malone TR, et al. Nerve injury in athletes caused by cryotherapy J Athl
Train 1992;27(3):235-37. [Bassett, 1991]

Double data publication

Ardvol J, et al. Treatment of complete rupture of the lateral ligaments

of the ankle: a randomised clinical trial comparing cast immobilisation
with functional treatment. Knee Surg Sports Traumatol Arthrosc.
2002;10(6):371-7. [Jones et al. 2007]

Study included as part of a relevant

systematic review article

Karlsson J, et al. Early functional treatment of acute ligament injuries of

the ankle joint. Scand J Med Sci Sports 1996;6(6):341-5. [Kerkhoffs et al.
2002]
Costa et al. Randomised controlled trials of immediate weight-bearing
mobilisation for rupture of the tendo Achillis. Journal of Bone and Joint
Surgery 2006;88:69-77.
Josey RA, et al. Immediate, full weightbearing cast treatment of acute
Achilles tendon ruptures: a long term follow up study. Foot Ankle Int
2003;24(10):775-9.
Jung YB, et al. Active non operative treatment of acute isolated posterior
cruciate ligament injury with cylinder cast immobilisation. Knee Surg
Sports Traumatol Arthrosc 2008;16(8):729-33. Epub 2008 Apr 17
Saeki Y et al. Effect of local application of cold or heat for relief of
pricking pain. Nursing and Health Sciences 2002;4(3):97-105.
Stevens et al. Frostbite due to improper use of frozen ice pack.

*excluded from full text

Healthy subjects / injury model not

relevant

Methods

N=11 healthy

N=9 healthy

N=50 healthy

N= 30 healthy

N=15 healthy

N=20 healthy

N=12 healthy

Study

Selfe, 2009
Randomised cross
over

Kennet, 2007
Repeated measures
(24 hr wash out)

Kanlayanaphotporn,
2005
Repeated measures
(24 hr wash out)

Janwantanakul,
2004
Repeated measures
(24 hr wash out)

Merrick, 2003
Repeated measures
(48hr wash out)

Chesterton, 2002
Repeated measures
(3 day wash out)

Palmer, 1996
Repeated measures
(48 hr wash out)

Skin temperature

Appendix Table 6:

Ankle and
Thigh

Thigh

Thigh

Thigh

Thigh

Ankle

Knee

Body part

20 min

20 min

- Ice pack (damp towelling barrier)

- Gel pack (damp towelling barrier)
- Frozen peas (damp towelling barrier)
- Mixture of water/alcohol (damp towelling barrier)
- Ice bag (chipped ice in plastic bag)
- Ice bag with damp towel wrap (100% cotton towel single layer)
- Room temperature water bag

20 min

- 20 min
- 30 min
- 40 min

- Frozen gel pack (single layer damp towelling barrier)

- Frozen peas (single layer damp towelling barrier)
- Room temperature gel pack
Exercise (15 mins) then ice pack secured with elastic wrap, then simulated
showering, then reapplication of ice pack (as above)

30 min

20 min

- Crushed Ice (1 litre in terrycloth)

- Gel pack (direct)
- Frozen peas (plastic packaging in terrycloth)
- CWI (6L cold water with 1L crushed ice)

- Ice bag (crushed ice, 1kg, 4L)

- Wet ice (terry towelling inbuilt to allow melting water to contact skin)
- Flex-i-cold (Frozen gel pack with no phase change)
- Control (no treatment)
All ice groups: Elastic wrap to secure pack in place at 45mmHg.

20 min

Duration

-Crushed Ice (300 gm in plastic bag)

-Gel pack
-Commercial cooling 1
-Commercial cooling 2

Mode

Allocation
concealment

Factors lowering
risk of bias
(Cochrane tool)

Knee

N=18 healthy

N=9 healthy

Ibrahim, 2005
RCT

Tsang, 1997
Repeated measures
(24 hr wash out)

Metzman, 1996

-Synthetic cast
-Plaster cast

Forearm

Ankle

N=32 healthy
N=12 acute inversion
injury

Okcu, 2006
RCT

Body part

Methods

Study

Crushed ice

0.68kg of cubed ice applied over:

-Bare skin
-Dry towelling
-Dry elastic bandage

Cryocuff (with autochill) applied over:

-No dressing
-Thin adhesive dressing (Tegaderm)
- Bulky dressing (wool and crepe)

200 ml frozen ice pack x2 applied over:

- Robert Jones bandage
- Elastic support wrap
- Below knee plaster
- Synthetic below knee cast

Mode

30 min

2 hrs

2 hrs?

Duration

Sequence generation
Allocation
concealment

Sequence generation
Allocation
concealment

Factors lowering
risk of bias
(Cochrane tool)

Myrer, 2001
Observational

Jutte, 2001
Observational

Otte, 2002
Observational

Merrick, 2003
Repeated measures
(2 day wash out)

Long, 2005
Repeated measures
(2 day wash out)

Bender, 2005
Repeated measures
(2 day wash out)

N=12 healthy

Dykstra, 2009
Repeated measures
(4 day wash out)

Depth of measurement:
skinfold measurement plus
1cm or 3cm

N= 30 healthy

Depth of measurement:
2cm below adipose layer

N=15 healthy

Depth of measurement:
1 cm deep to adipose layer

N=47 healthy

Depth of measurement:
1 and 2 cm subadipose

N=15 healthy

Depth of measurement:
1 and 2 cm subadipose

N=6 healthy

Depth of measurement:
skinfold measurement plus
1.5 cm

N=16 healthy

Depth of measurement:
skin fold thickness plus 2 cm

Methods

Study

Muscle temperature

Appendix Table 7:

Calf (overall mean:

15.5+/-8.7mm); subgrouped into: <8mm (6.5
+/-1.4mm); 10-18 mm
(14.1 +/-2.6); >20 mm
(25.7+/-5.4)

Crushed ice

Ice cubes with 5 inch elastic wrap

Crushed ice

Thigh
(sub-grouped into:
0-10mm; 11-20mm; 2130mm; 31-40mm)
Thigh (21.2 +/-8.6mm)

-Ice bag
-Wet ice
-Flex-i-cold
All groups compression at 45-50mm/Hg)

Ice bag with compression wrap at 42-48mm Hg

Ice bag secured with plastic wrap (62 +/-6.8mm/Hg)

-Cubed
-Crushed
-Wetted.
All groups no elastic wrap

Mode

Thigh (19.3 +/-4.1mm)

Thigh (25.4 +/-2.7 mm)

Calf (21.1 +/-9.3mm)

Calf
(male: 12.8 +/-4.1;
female: 17.3 +/-3.4mm)

Body part
(mean skin fold mm)

20 min

30 min

Up to 58.6 min

32 min

Up to 54 min

30 min

20 min

Duration

Factors lowering
risk of bias
(Cochrane tool)

Myrer, 1997
RCT

Zemke, 1998
RCT

Myrer, 1998
RCT

N=28 healthy

Myrer, 2000
RCT

Depth of measurement:
1 cm below subcutaneous fat

N=16 healthy

Depth of measurement:
skinfold measurement plus
1cm

N=14 healthy

Depth of measurement:
1 cm below subcutaneous fat

N=32 healthy

Depth of measurement:
1cm below subcutaneous fat

Methods

Study

-Ice massage
-Ice bag

Crushed ice

Calf (male: 17.2 +/7.5mm; female: 20.9

+/-8mm)

20 min

15 min

20 min

20 min

Crushed ice

-Crushed ice
-CWI

Duration

Mode

Calf (ice massage: 0.78

+/-0.25 cm; ice bag:
0.59 +/-0.36cm)

Calf

Calf (16.6 +/-7.4mm)

Body part
(mean skin fold mm)

Sequence
generation

Sequence
generation

Sequence
generation

Sequence
generation

Factors lowering
risk of bias
(Cochrane tool)

Knee: lateral gutter

(Two layer bandaging)

Shoulder joint / AC joint

(Large bulky dressing)

Knee: medial gutter and

supra-patellar pouch
(Two layer bandaging)
Knee: medial gutter and
supra-patellar pouch
(Two layer bandaging)
Knee: suprapatellar pouch
(Round cotton bandage)

N=17
Knee arthroscopy (meniscal tears, OA,
loose bodies, plica, chondromalacia)

N=20 subjects with full thickness

rotator cuff tears repaired with a
corkscrew open repair

N=12
Knee arthroscopy (meniscal tears, OA,
loose bodies, plica, chondromalacia)

N=16
ACL reconstruction (middle third
patellar graft; standard two incision
technique, metal interface screws)

N=23
Knee arthroscopy

Martin, 2001
CT

Osbahr, 2002
RCT

Martin, 2002
CT

Glenn, 2004
CT

Sanchez-Inchausti,
2005

Observational/RM

-Commercial icing machine set to 5C;

-Commercial icing machine set to 10C
-Control (dressing only)

Knee: Suprapatellar pouch /

intracondylar notch
(5 sheets 20x30 gauze,
wrapped in elastic bandage)

N=21
ACL reconstruction (bone tendon bone
patellar graft; single incision technique)

Ohkoshi, 1999
RCT

Ice bag

-Cryocuff for first hour post surgery

- Cryocuff for second hour post surgery
Both groups water changed every 30 minutes

-Cryocuff immediately post surgery

- Cryocuff 1 hour post surgery
Both groups water changed every 30 minutes

- Polar care continuous cooling replenished

every 4-6 hrs
- Control (dressing only)

- Cryocuff immediately post surgery

- Cryocuff 1 hour post surgery
Both groups water changed every 30 minutes

- Cryocuff automatic pump re-chilling every 30 secs

- Control (dressing only)

Knee
(Sterile gauze pads and a
layer of sterile webril)

N=30
Arthroscopic knee surgery (medial or
lateral meniscectomy)

Zaffagnini, 1998
RCT

23

1-2

48

1.5

- Cryocuff rechilled every 15,30,45 minutes

- Control (dressing only)

Shoulder: glenohumeral
and subacromial spaces
(Double thickness 4x4
gauze pad and silk tape)

N=15
Shoulder arthroscopy and debridement
of the subacromial spaces

Duration
(hours)

Mode

Levy, 1997
RCT

Body part
(Barrier)

Methods

Study

Joint temperature reduction

Appendix Table 8:

Sequence
generation
Allocation
concealment
Blinded outcome
assessment

Sequence
generation

Allocation
concealment

Sequence
generation

Allocation
concealment

Sequence
generation

Factors lowering
risk of bias
(Cochrane tool)

Appendix Table 9: AMSTAR Scale

1. Was an a priori design provided?
The research question and inclusion criteria should be established before the conduct of
the review.

Yes
No
Cant answer
Not applicable

2. Was there duplicate study selection and data extraction?

There should be at least two independent data extractors and a consensus procedure
for disagreements should be in place.

Yes
No
Cant answer
Not applicable

3. Was a comprehensive literature search performed?

At least two electronic sources should be searched. The report must include years and
databases used (e.g. Central, EMBASE, and MEDLINE). Key words and/or MESH terms
must be stated and where feasible the search strategy should be provided. All searches
should be supplemented by consulting current contents, reviews, textbooks, specialized
registers, or experts in the particular field of study, and by reviewing the references in
the studies found.

Yes
No
Cant answer
Not applicable

4. Was the status of publication (i.e. grey literature) used as an

inclusion criterion?
The authors should state that they searched for reports regardless of their publication
type. The authors should state whether or not they excluded any reports (from the
systematic review), based on their publication status, language etc.

Yes
No
Cant answer
Not applicable

5. Was a list of studies (included and excluded) provided?

A list of included and excluded studies should be provided.

Yes
No
Cant answer
Not applicable

6. Were the characteristics of the included studies provided?

In an aggregated form such as a table, data from the original studies should be
provided on the participants, interventions and outcomes. The ranges of characteristics
in all the studies analyzed e.g. age, race, sex, relevant socioeconomic data, disease
status, duration, severity, or other diseases should be reported.

Yes
No
Cant answer
Not applicable

7. Was the scientific quality of the included studies assessed and documented?
A priori methods of assessment should be provided (e.g., for effectiveness studies if the
author(s) chose to include only randomized, double-blind, placebo controlled studies,
or allocation concealment as inclusion criteria); for other types of studies alternative
items will be relevant.

Yes
No
Cant answer
Not applicable

8. Was the scientific quality of the included studies used appropriately in

formulating conclusions?
The results of the methodological rigor and scientific quality should be considered
in the analysis and the conclusions of the review, and explicitly stated in formulating
recommendations.

Yes
No
Cant answer
Not applicable

9. Were the methods used to combine the findings of studies appropriate?

For the pooled results, a test should be done to ensure the studies were combinable,
to assess their homogeneity (i.e. Chi-squared test for homogeneity, I). If heterogeneity
exists a random effects model should be used and/or the clinical appropriateness of
combining should be taken into consideration (i.e. is it sensible to combine?).

Yes
No
Cant answer
Not applicable

10. Was the likelihood of publication bias assessed?

An assessment of publication bias should include a combination of graphical aids (e.g.,
funnel plot, other available tests) and/or statistical tests (e.g., Egger regression test).

Yes
No
Cant answer
Not applicable

11. Was the conflict of interest stated?

Potential sources of support should be clearly acknowledged in both the systematic
review and the included studies.

Yes
No
Cant answer
Not applicable

Appendix Table 10: Cochrane Risk of bias tool

1. Sequence generation
Was the allocation sequence adequately generated?
(Yes / No / Unclear)
2. Allocation concealment
Was allocation adequately concealed?
(Yes / No / Unclear)
3. Blinding
Was knowledge of the allocated intervention adequately prevented during the study?
Assessments will be made for:
a. Outcome assessors: primary outcomes: (Yes / No / Unclear)
4. Were incomplete outcome data adequately addressed?
(Yes / No / Unclear)
5. Free of other bias
a. Is the PRICE protocol clear and consistent between groups?
(Yes / No / Unclear)
b. Were no co-interventions used, or if present, were they standardised across groups?
(Yes / No / Unclear)

blinding of participant or caregivers were not considered based on the nature of the interventions.