Pain Medicine, 23(3), 2022, 515–525

doi: 10.1093/pm/pnab188
Advance Access Publication Date: 11 June 2021
Review Article

Dry Needling Versus Trigger Point Injection for Neck Pain

Symptoms Associated with Myofascial Trigger Points: A Systematic
Review and Meta-Analysis

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

Marcos J. Navarro-Santana, PT, MSc,*,† Jorge Sanchez-Infante, PT, MSc,‡
 mez-Chiguano, PT, MSc,§Joshua A. Cleland, PT, PhD,¶
Guido F. Go
sar Ferna
Ce ~ as, PT, PhD,k,kk Patricia Martın-Casas, PT, PhD,*,** and
ndez-de-las-Pen
Gustavo Plaza-Manzano, PT, PhD*,**

*Department of Radiology, Rehabilitation and Physiotherapy, Universidad Complutense de Madrid, Madrid, Spain; †Rehabilitacio n San Fernando,
Madrid, Spain; ‡Performance and Sport Rehabilitation Laboratory, Faculty of Sport Sciences, University of Castilla-La Mancha, Toledo, Spain; §Clınica
Dinamia Fisioterapia, Madrid, Spain; ¶Doctor of Physical Therapy Program, Department of Public Health and Community Medicine, Tufts University
School of Medicine, Boston, Massachusetts, USA; kDepartment of Physical Therapy, Occupational Therapy, Physical Medicine and Rehabilitation,
Universidad Rey Juan Carlos (URJC), Madrid, Spain; kkC atedra Institucional en Docencia, Clınica e Investigacion en Fisioterapia: Terapia Manual,
Puncion Seca y Ejercicio Terap
eutico, Universidad Rey Juan Carlos, Alcorc
on, Madrid, Spain; **Instituto de Investigaci
on Sanitaria del Hospital Clınico
San Carlos, Madrid, Spain

Correspondence to: Cesar Fernandez de las Pe~nas, PT, PhD, Facultad de Ciencias de la Salud, Universidad Rey Juan Carlos, Avenida de
Atenas s/n, 28922 Alcorcon, Madrid, Spain. Tel: þ 34 91 488 88 84; Fax: þ34 91 488 89 57; E-mail: [email protected]/[email protected].

Funding source: No funds were received for this study.

Conflicts of interest: The authors declare no conflicts of interest.

Study registration: OSF Registry—https://doi.org/10.17605/OSF.IO/3H6GS.

Abstract
Objective. To examine the effects of dry needling against trigger point (TrP) injections (wet needling) applied to TrPs associ-
ated with neck pain. Methods. Electronic databases were searched for randomized clinical trials in which dry needling
was compared with TrP injections (wet needling) applied to neck muscles and in which outcomes on pain or pain-
related disability were collected. Secondary outcomes consisted of pressure pain thresholds, cervical mobility, and
psychological factors. The Cochrane Risk of Bias tool, the Physiotherapy Evidence Database score, and the Grading
of Recommendations Assessment, Development, and Evaluation approach were used. Results. Six trials were in-
cluded. TrP injection reduced pain intensity (mean difference [MD ] –2.13, 95% confidence interval [CI] –3.22 to –1.03)
with a large effect size (standardized mean difference [SMD] –1.46, 95% CI –2.27 to –0.65) as compared with dry nee-
dling. No differences between TrP injection and dry needling were found for pain-related disability (MD 0.9, 95% CI –
3.09 to 4.89), pressure pain thresholds (MD 25.78 kPa, 95% CI –6.43 to 57.99 kPa), cervical lateral-flexion (MD 2.02 ,
95% CI –0.19 to 4.24 ), or depression (SMD –0.22, 95% CI –0.85 to 0.41). The risk of bias was low, but the heteroge-
nicity and imprecision of results downgraded the evidence level. Conclusion. Low evidence suggests a superior effect
of TrP injection (wet needling) for decreasing pain of cervical muscle TrPs in the short term as compared with dry
needling. No significant effects on other outcomes (very low-quality evidence) were observed. Level of Evidence.
Therapy, level 1a.

Key words: Dry Needling; Trigger Point Injection; Lidocaine: Meta-Analysis; Cervical Pain; Pain; Systematic Review

C The Author(s) 2021. Published by Oxford University Press on behalf of the American Academy of Pain Medicine. All rights reserved.
V
For permissions, please e-mail: [email protected] 515
516 Navarro-Santana et al.

Introduction To the best of the authors’ knowledge, no meta-

Neck pain is a common musculoskeletal condition showing analysis has specifically compared the effects of TrP in-
a point prevalence of 20% and a lifetime prevalence of jection (wet needling) with the effects of dry needling in
70% in the general population [1]. The Global Burden of the management of neck pain symptoms associated with
Disease Study identified neck pain as the fourth highest con- TrPs. Additionally, more clinical trials have been pub-
dition for number of years lived with disability [2]. lished since the publication of the meta-analysis by Liu
The etiology of mechanical neck pain is not completely et al. [8]. Therefore, an updated analysis of the current
understood, and it seems to be multifactorial. One hypothe- literature comparing the effects of TrP injection (wet nee-
sis associated with the development of neck pain is the pres- dling) vs dry needling is needed.
ence of myofascial trigger points (TrPs). A TrP is defined as The purpose of the present systematic review and
a hypersensitive spot in a taut band of skeletal muscle that, meta-analysis was to compare dry needling with TrP in-
when stimulated, induces referred pain symptoms and also jection (wet needling), applied over neck-shoulder muscle

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

motor phenomena [3]. It has been shown that, of all of the TrPs that reproduce neck pain of musculoskeletal origin,
neck-shoulder musculature, the upper trapezius is most af- with regard to effects on pain intensity, pain-related dis-
fected by TrPs in patients with mechanical neck pain [4]. In ability, pain sensitivity, and cervical range of motion.
fact, the referred pain elicited by active TrPs from the upper
trapezius muscle often reproduces symptoms associated
with insidious neck pain or traumatic neck pain [3]. Methods
Different therapeutic approaches are advocated for
the management of TrPs, with needling interventions be- This systematic review and meta-analysis adheres to the
ing the most commonly used [5]. There are two needling Preferred Reporting Items for Systematic Reviews and
procedures used for the management of TrPs: wet and Meta-Analyses (PRISMA) statement [14]. The interna-
dry needling [3]. Wet needling (also called TrP injections) tional OPS Registry registration link is https://doi.org/10.
refers to procedures that include the injection of a sub- 17605/OSF.IO/3H6GS. The methods used in the present
stance (usually a local anesthetic) into a TrP through a review and meta-analysis have been used in previous
hypodermic beveled-cutting-edge needle, whereas dry meta-analyses and have been described in detail previ-
needling is defined as a “skilled intervention using a thin ously [15–17]. We will briefly summarize the most rele-
filiform needle to penetrate the skin that stimulates myo- vant aspects here.
fascial TrPs, muscles, and connective tissue for the treat-
ment of musculoskeletal pain disorders” [6].
Current evidence supporting the effects of either nee-
Systematic Literature Search and Selection
dling intervention is conflicting. The review by Cagnie et Criteria
al. recommended dry needling for the treatment of Electronic literature searches were conducted in the
patients with upper trapezius TrPs; however, no meta- MEDLINE, CINAHL, PubMed, PEDro, Cochrane
analysis was conducted [7]. The meta-analysis by Liu et Library, SCOPUS, and Web of Science databases from
al. recommended dry needling for the management of their inception to July 10, 2020. If possible, searches
neck-shoulder myofascial pain at short- and mid-term were restricted to randomized clinical trials. This system-
follow-ups [8]. These authors included a comparison be- atic review and meta-analysis included randomized clini-
tween TrP injections and dry needling and found that cal trials in which any form of dry needling was
TrP injection (wet needling) was more effective than dry compared with any form of TrP injections (wet needling)
needling 1 month after the intervention [8]. in adults with neck pain due to myofascial TrPs.
A previous review found a trend toward a greater ef- Acupuncture was excluded. The primary outcome of the
fect of lidocaine injection for reducing pain over dry nee- trial had to include pain intensity or pain-related disabil-
dling in individuals with neck-shoulder pain [9]; ity. Pressure pain sensitivity (e.g., pressure pain thresh-
however, the analysis of that study has been questioned olds), cervical range of motion, or psychological factors
[10]. The meta-analysis investigated only changes in pain (e.g., depression, anxiety) were considered as secondary
intensity [9]. One review investigating the effects of TrP outcomes. The search strategy for each database can be
injection (wet needling) and dry needling in people with seen in Supplementary Data Table 1.
temporomandibular pain found inconclusive evidence in
favor of either needling intervention [11]. Similarly, two
recent meta-analyses reported low evidence supporting Screening, Selection Process, and Data Extraction
the use of local anesthetic injections for decreasing pain Articles identified from the different databases were inde-
in individuals with head-neck myofascial pain [12, 13]. pendently reviewed by two authors, as previously de-
These meta-analyses included patients with multiple di- scribed [15–17]. A standardized form was used for
agnoses, including headache, temporomandibular pain, extracted data from each trial. Data extraction was con-
and neck pain. ducted by two authors, as previously described [15–17].
Dry Needling vs Trigger Point Injection for Neck Pain 517

Table 1. Characteristics of the sample of the included studies

Study Diagnosis Group Total (Men/Women) Age, y, mean6SD Pain Duration, mo, mean6SD
Hong et al. [27] Myofascial pain G1: LI with LTR þ 35 (10/25) 41.6611.4 10.665.8
syndrome home program
G2: DN with LTR þ 23 (6/17) 41.8612.8 8.164.5
home program
G1a: LI with LTR þ 26 (7/19) 42.2612.1 10.265.6
home program*
G2a: DN with LTR þ 15 (4/9) 41.7614.4 7.664.7
home program*
G1b: LI without LTR 9 (3/6) 39.969.6 11.766.7
þ home program
G2b: DN without 8 (2/6) 42.1610.2 9.164.2

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

LTR þ home
program
Kamanli et al. [28] Myofascial pain G1: DN 10 (NR) 37.268.1 32.5622.0
syndrome G2: LI injection 10 (NR) 37.369.75 49.2634.9
G3: bolutinum toxin- 9 (NR) 38.365.25 50.6619.9
A injection (N/A)
Ay et al. [30] Myofascial pain G1: DN þ home 40 (14/26) 38.169.8 34.3640.95
syndrome exercises
G2: LI þ home 40 (14/26) 37.2610.1 30.6637.2
exercises
Ga et al. [29] Diagnosis of neck G1: DN þ self- 18 (1/17) 79.266.8 >6 months
pain and myofascial stretching
pain syndrome G2: LI þ self- 21 (2/19) 75.968.7 >6 months
stretching
Eroglu et al. [31] Diagnosis of neck G1: DN 20 (1/19) 33.7568.1 48 (range 2–120)
pain and myofascial G2: LI injection 20 (6/14) 32.8569.05 36 (range 3–120)
pain syndrome G3: flurbiprofen injec- 20 (0/20) 34.5568.3 24 (range 1–72)
tion (N/A)
Raeissadat et al. [33] Diagnosis of neck G1: DN 20 (4/16) 41.666.8 4.661.4
pain and myofascial G2: ozone injection 22 (6/16) 37.668.8 4.461.3
pain syndrome G3: LI injection (N/A) 20 (4/16) 39.467.7 4.461.2
Ibrahim et al. [32] Myofascial pain G1: DN 20 (10/10) 70.0564.9 >3 months
syndrome G2: LI injection 20 (9/11) 68.263.2 >3 months

DN¼ dry needling; LI¼ lidocaine injection; LTR¼ local twitch response; NR¼ not reported; N/A¼ not applicable to the meta-analysis.
*Groups included in the meta-analysis due to the time period analysis.

Assessment of Methodological Quality and Risk Data Synthesis and Analysis

of Bias Data synthesis considered the following follow-up peri-
The risk of bias (RoB) of the included trials was indepen- ods: short-term follow-up (1 to 12 weeks), mid-term
dently assessed by two authors using the Cochrane Risk of follow-up (12 to 24 weeks), and long-term follow-up
Bias assessment tool [18]. The RoB was applied and classi- (>24 weeks), if data were available.
fied according to the Cochrane Collaboration’s tool [18]. We extracted the sample size, means, and standard
The methodological quality of the studies was evaluated deviations for each variable. When the trial reported only
with the Physiotherapy Evidence Database (PEDro) scale standard errors, they were converted to standard devia-
[19]. A trial was considered to be of high quality when the tions. When necessary, the mean scores and standard
PEDro score was 6 out of 10 points [19]. deviations were estimated from graphs. Also, if the trial
presented nonparametric values (medians and interquar-
Level of Evidence tile ranges), they were converted to means and standard
The Grading of Recommendations Assessment, deviations as appropriate [22, 23].
Development, and Evaluation (GRADE) approach was used Cervical range of motion was pooled just for lateral-
to summarize the quality of the evidence [20]. The level of flexion, as it was the only motion assessed in more than
evidence was classified as high (when all items were nega- two studies. When the trial calculated the total range of
tive), moderate (when one item showed serious risk), low motion or either side separately, the mean was used in
(when two items showed serious risk or one item very serious the quantitative analysis.
risk), or very low (when three or more items had serious risk A random-effects model was used to determine the
or two or more items had very serious risk) [21]. overall effect size (standardized mean difference [SMD]).
518 Navarro-Santana et al.

Effect sizes on all outcomes were calculated at short-term Methodological Quality

follow-up (1 to 12 weeks), as no mid- or long-term data The methodological quality scores ranged from 5 to
were available. Effect sizes (SMD) were classified as large 8 (mean: 6.4, standard deviation: 1.4) out of a maximum
(0.8), moderate (from 0.5 to 0.79), or small (from 0.2 of 10 points. Four studies (57%) were considered to be
to 0.49) [24]. P values <0.05 were considered statisti- of high methodological quality (6 points), and the
cally significant. remaining three (43%) were considered to be of low
The heterogeneity of the studies was assessed with the methodological quality (<6 points). The most frequent
I2 statistic and classified as considerable heterogeneity (I2 methodological quality bias was lack of blinding of the
from 75% to 100%), substantial heterogeneity (I2 from participants, as only one study was able to report this
50% to 90%), moderate heterogeneity (I2 from 30% to item . Table 2 represents the details of the PEDro scale of
60%), or not relevant/important heterogeneity (I2 from each trial.
0% to 40%) [25].

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

Risk of Bias
The details of the risk-of-bias assessment of the included
Results trials are displayed in Figure 2. No trial was able to blind
the therapists, and most trials had an unclear or high
Study Selection
RoB in the item of blinding assessors and participants. In
A total of 557 potential studies were identified for re-
general, the RoB of the included trials in the present
view. After removal of duplicates, 324 studies remained.
meta-analysis was low.
Of those, 316 were excluded after their titles or abstracts
had been read, which left eight articles for full-text analy-
Effects of TrP Injections vs Dry Needling on Pain
sis [26–33]. One article was excluded because it mea-
sured the intensity of pain during each needling and Pain-Related Disability
The meta-analysis found a statistically significant effect
intervention [26]. Finally, a total of seven trials [27–33]
(P < 0.001) for reducing pain
were included in the systematic review, and six trials
(mean difference [MD] –2.13, 95% confidence in-
[27–30, 32, 33] were included in the quantitative analy-
terval [CI] –3.22 to –1.03, n ¼ 6 trials, Figure 3) of TrP
ses (Figure 1).
injection (wet needling) vs dry needling, with a large
effect size (SMD –1.46, 95% CI –2.27 to –0.65,
Study Characteristics n ¼ 291, Z ¼ 3.54, P ¼ 0.0004) but also with consider-
The characteristics of the participants of the included able heterogeneity (I2 ¼ 91%) between studies at short-
studies are shown in Table 1. Supplementary Data Table term follow-up. One study investigating the immediate
2 summarizes the characteristics of the needling interven- differences reported no significant effect of TrP injec-
tions applied in each trial. All studies targeted active TrPs tion (wet needling) vs dry needling (MD –0.58, 95%
(i.e., those that referred pain that reproduced the CI –1.20 to 0.04) [27]. Similarly, no differences at
patient’s symptoms) with the needle, five targeted only mid-term follow-up (MD –0.28, 95% CI –0.64 to
upper trapezius TrPs [24, 26, 29, 30], and the remaining 0.08) were observed [30]. Table 3 summarizes the
two targeted active TrPs in all neck muscles identified main results of the included studies.
[28, 31]. Five trials reported the presence of local twitch Since only one trial investigated changes in pain-
responses during the needling intervention [27, 29–32]. related disability between TrP injection (wet nee-
All clinical trials included a group receiving lidocaine in- dling) and dry needling, a meta-analysis was not pos-
jection as a TrP injection (wet needling) and a group re- sible. No significant between-group differences were
ceiving dry needling . Kamanli et al. [28] also included found (MD 0.90, 95% CI –3.09 to 4.89) for pain-
one group receiving botulinum toxin injection, and related disability [33].
Raeissadat et al. [33] included one group receiving oxy-
gen injection. Only the comparison between lidocaine in- Effects of TrP Injection vs Dry Needling on
jection and dry needling was pooled in the quantitative Secondary Outcomes
analysis. All trials except one [31] applied just one nee- The meta-analysis found that TrP injection (wet nee-
dling treatment session. All studies included neck pain as dling) exhibited a nonsignificant effect (MD 25.78 kPa,
the primary outcome, whereas just one included pain-re- 95% CI –6.43 to 57.99 kPa, n ¼ 101, Z ¼ 1.57, P ¼ 0.12,
lated disability [33]. Pressure pain thresholds and depres- n ¼ 3 trials) for increasing pressure pain thresholds as
sive levels were assessed in three studies, and cervical compared with dry needling, with moderate heterogene-
range of motion in lateral-flexion was assessed in four ity (I2 ¼ 54%) between studies (Figure 4A). Similarly, no
studies. All included studies investigated the effects at significant immediate differences (MD 22.55 kPa, 95%
short-term follow-up, ranging from 2 to 4 weeks (mean: CI 22.37 to 67.47 kPa) in pressure pain threshold were
3 6 1 week) [27–30, 32, 33]. found [27].
Dry Needling vs Trigger Point Injection for Neck Pain 519

Publications potentially relevant identified Figure 4C) in depressive symptoms between TrP injection
by electronic search (n= 557) (wet needling) and dry needling were either observed at
short-term follow-up. This analysis showed moderate
Duplicated publications (n=233) heterogeneity (I2 ¼ 65%).

Quality of Evidence (GRADE)

Publications excluded based on Table 4 displays the details of GRADE assessment show-
review of title and abstract (n=316)
ing RoB, inconsistency of the results, indirectness of evi-
dence, imprecision of results, and high probability of
Publications selected for full- publication bias. The serious/very serious inconsistency
text evaluation (n=8) of the results (heterogeneity) and the serious/very serious
imprecision downgraded the evidence level for TrP injec-

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

tion (wet needling) to low or very low.

Adverse Events
Most studies did not report any serious adverse event
[27–30, 32, 33]. Only one study did not provide informa-
tion about adverse events [31] (Supplementary Data
Table 3). The most common adverse events with the ap-
Studies included in qualitative plication of TrP injections (wet needling) were post-nee-
synthesis (n =7) dling soreness, muscle pain, and discomfort after the
intervention [27–29]. Other adverse events included par-
Publications excluded based on full-
text review (n=1)
esthesia, fatigue, headache, hemorrhage, transient flare
Studies included in quantitative
synthesis (meta-analysis) reaction, and dizziness [27–30, 32, 33]. All of these ad-
Assess pain during needling (n=1)
(n =6) verse events did not need further treatment and disap-
peared after a few days.
Figure 1. PRISMA flow diagram. The most common adverse events with dry needling
application were post-needling soreness, pain, and dis-
Table 2. Score of randomized clinical trials with PEDro scale comfort after the intervention [27, 29]. One patient expe-
1 2 3 4 5 6 7 8 9 10 TOTAL
rienced a transient flare reaction after dry needling [33].
All these adverse events did not need further treatment
Hong et al. [27] Y N Y N N Y N N Y Y 5/10
Kamanli et al. [27] Y N Y N N N Y N Y Y 5/10
and disappeared spontaneously after a few days.
Ga et al. [29] Y N Y N N N Y N Y Y 5/10
Ay et al. [30] Y Y Y N N N Y Y Y Y 7/10
Eroglu et al. [31] Y N Y Y N Y Y N Y Y 7/10 Discussion
Raeissadat et al. Y Y Y Y N N Y N Y Y 8/10 TrP Injection (Wet Needling) or Dry Needling
[33]
Ibrahim et al. [32] Y N Y Y N Y Y Y Y Y 8/10
This meta-analysis compared the effects of TrP injection
(wet needling) vs dry needling for the management of
1 ¼ random allocation of participants; 2 ¼ concealed allocation; 3 ¼ simi- neck pain symptoms of musculoskeletal origin associated
larity between groups at baseline; 4 ¼ participant blinding; 5 ¼ therapist
with TrPs. We found low evidence suggesting that TrP
blinding; 6 ¼ assessor blinding; 7 ¼ fewer than 15% dropouts; 8 ¼ intention-
to-treat analysis; 9 ¼ between-group statistical comparisons; 10 ¼ point
injections (wet needling) with lidocaine had a superior ef-
measures and variability data. fect for reducing pain when compared with dry needling.
The RoB of the trials included in this meta-analysis was
generally low, but the inconsistency (heterogeneity) and
TrP injections (wet needling) did not show a signifi- imprecision of the results downgraded the evidence level
cant effect (MD 2.02 , 95% CI –0.19 to 4.24 , n ¼ 200, (GRADE).
Z ¼ 1.79, P ¼ 0.08, n ¼ 4 trials) as compared with dry The present meta-analysis is the first one specifically
needling for improving cervical lateral-flexion range of analyzing the impact of TrP injection (wet needling) vs
motion (Figure 4B). There was substantial heterogeneity dry needling on neck pain intensity, pain-related disabil-
between the trials (I2 ¼ 56%). The only trial investigating ity, pressure pain sensitivity, range of motion, and de-
immediate changes in cervical lateral-flexion motion pressive levels in people with TrPs associated with neck
reported no significant differences (MD 4.70 , 95% CI – pain. A previous meta-analysis did not find significant
0.30 to 9.70 ) between TrP injection and dry needling differences between these needling interventions [9], but
[27]. Table 3 summarizes main results of the studies. the results should be considered with caution [10]. In
No significant differences (SMD –0.22, 95% CI –0.85 contrast, Liu et al. [8] found a large effect for TrP injec-
to 0.41, n ¼ 139, Z ¼ 0.68, P ¼ 0.50, n ¼ 3 trials, tion (wet needling) when compared with dry needling
520 Navarro-Santana et al.

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

Figure 2. Plots of the RoB of the included studies.

Figure 3. Comparison (MD) between the effects of TrP injection (wet needling) and the effects of dry needling on pain intensity at
short-term follow-up.

(SMD 1.69; 95% CI 0.40 to 2.98) at 4 weeks. Our results considered 9–28 days after the intervention as mid-term
are similar (SMD –1.46, 95% CI –2.27 to –0.65) to those follow-up, when it may be more appropriate to be con-
previously reported by Liu et al. [8]; however, they sidered as short-term follow-up. In addition, our results
Dry Needling vs Trigger Point Injection for Neck Pain 521

Table 3. Results of the included studies

Short-Term Follow-Up (<12

Weeks after Treatment)
Study Outcome/Group Baseline Mean6 SD Mean6 SD
Hong et al. [27] Pain (NPRS, 0–10)
G1a 7.886 0.93 0.966 0.90 (2 wk)
G2a 7.806 0.83 4.936 1.44 (2 wk)
PPT (kPa)
G1a 272.626 55.89 361.866 61.78 (2 wk)
G2a 276.546 68.64 363.826 46.09 (2 wk)
Lateral flexion ( )
G1a 34.86 10.2 48.86 9.2 (2 wk)
G2a 35.76 15.3 47.76 18.2 (2 wk)

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

Kamanli et al. [28] Pain (VAS, 0–10)
G1 7.036 2.68 5.126 2.94 (4 wk)
G2 6.906 1.43 1.956 1.67 (4 wk)
G3 (N/A) 6.096 1.95 2.686 1.04 (4 wk)
PPT (kPa)
G1 302.046 43.14 371.676 76.49 (4 wk)
G2 313.816 41.18 427.566 81.39 (4 wk)
G3 (N/A) 314.796 55.89 389.326 76.49 (4 wk)
Depression (0–53)
G1 10.806 4.05 11.306 3.65 (4 wk)
G2 9.206 5.65 7.006 3.53 (4 wk)
G3 (N/A) 12.626 6.23 8.506 4.81 (4 wk)
Ga et al. [29] Pain (VAS, 0–10)
G1 6.986 1.32 3.826 2.47 (2 wk)
G2 6.436 2.08 3.466 2.47 (2 wk)
Lateral flexion ( )
G1 25.286 6.08 35.006 6.47 (2 wk)
G2 29.656 9.6 38.456 10.62 (2 wk)
Depression (GDS-SF)
G1 5.446 3.15 4.176 3.68 (2 wk)
G2 6.106 3.95 5.146 4.35 (2 wk)
Pain (VAS, 0–10)
G1 5.556 1.33 3.826 0.47 (4 wk)
G2 5.826 1.25 2.276 0.98 (4 wk)
Ay et al. [30] Right lateral flexion
G1 42.376 2.52 42.256 2.76 (4 wk)
G2 41.256 2.19 41.256 2.46 (4 wk)
Left lateral flexion
G1 42.626 2.52 43.206 2.39 (4 wk)
G2 41.126 2.50 42.126 2.50 (4 wk)
Lateral flexion (mean calculated)
G1 42.496 2.52 42.726 2.57 (4 wk)
G2 41.186 2.34 42.186 2.50 (4 wk)
Depression (BDI, 0–21)
G1 12.126 3.57 10.876 3.25 (4 wk)
G2 14.526 16.92 10.676 2.58 (4 wk)
Eroglu et al. [31] (Excluded from • VAS pain No differences were found between No significant differences were
meta-analysis) • PPT groups obtained between groups in
• Right-lateral flexion PPT, VAS, cervical range of
• Left-lateral flexion motion
• Right rotation
• Left rotation
Raeissadat et al. [33 ] VAS (0–10)
G1 6.36 0.7 3.26 0.8 (4 wk)
G2 (N/A) 5.76 0.9 2.46 1.5 (4 wk)
G3 6.26 0.9 2.56 1.1 (4 wk)
PPT (kPa)
G1 272.626 36.28 322.636 37.26 (4 wk)
G2 (N/A) 281.446 67.66 363.826 83.35 (4 wk)
G3 284.396 50.99 360.886 45.11 (4 wk)
Lateral flexion ( )
G1 32.86 4.7 33.96 3.9 (4 wk)

(continued)
522 Navarro-Santana et al.

Short-Term Follow-Up (<12

Weeks after Treatment)
Study Outcome/Group Baseline Mean6 SD Mean6 SD
G2 (N/A) 33.96 4.2 35.96 3.9 (4 wk)
G3 33.86 6.5 37.56 3.0 (4 wk)
Disability (NDI, %)
G1 46.36 9.1 40.86 7.3 (4 wk)
G2 (N/A) 49.66 11.4 36.86 9.8 (4 wk)
G3 51.06 7.0 39.96 7.9 (4 wk)
Ibrahim et al. [32] Pain (VAS, 0–10)
G1 7.036 2.68 6.126 2.94 (2 wk)
G2 7.426 0.82 2.86 1.1 (2 wk)

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

G ¼ group; VAS ¼ visual analog scale; NPRS ¼ numeric pain rating scale; NDI ¼ Neck Disability Index; PPT ¼ pressure pain thresholds; BDI ¼ Beck
Depression Inventory; GDS-SF ¼ Geriatric Depression Scale—Short; N/A ¼ group not included in the meta-analysis.

Figure 4. Comparison between the effects of TrP injection (wet needling) and the effects of dry needling on (A) pressure pain
thresholds (MD), (B) cervical range of motion in latera-flexion (MD), and (C) depressive levels (SMD) at short-term follow-up.

were similar to those previously reported in people with (MCID) of 2.1 points described for subjects with me-
temporo-mandibular pain disorders associated with mas- chanical neck pain [34] and was superior to the MCID
ticatory TrPs [12, 13]. Current evidence would support (1.4 cm) determined by Bijur et al. [35]. This between-
that TrP injection (wet needling) may be effective for the groups MD suggests a potential clinical superiority of
management of pain associated with neck and head TrPs TrP injections (wet needling) vs dry needling; however, it
(low evidence); however, it should be considered that the should be considered that the lower bound of the confi-
effects were mostly observed at short-term follow-up (2 dence interval did not surpass the MCID. It is possible
to 12 weeks after treatment). that some individuals with myofascial TrPs associated
The pooled data reported an overall mean decrease of with neck pain symptoms exhibit more benefits from TrP
pain intensity of –2.13 points (95% CI –3.22 to –1.03) injections or dry needling than do others.
after TrP injections (wet needling). This between-groups We were unable to pool data for comparing the effects
MD reached the minimal clinically important difference of TrP injections vs dry needling for pain-related disability
Dry Needling vs Trigger Point Injection for Neck Pain 523

Table 4. Level of Evidence (GRADE) for effects of TrP injection (wet needling) and dry needling on pain intensity, pressure pain sen-
sitivity, cervical range of motion, and depressive levels in patients with neck pain

Number of Studies RoBInconsistency Indirectness of EvidenceImprecisionPublication BiasQuality of EvidenceMD or SMD (95% CI)
TrP injection (wet needling) vs dry needling on neck pain intensity
Overall effect (n ¼ 6) No Very serious No No No Low MD –2.13 (–3.22 to –1.03)*
(I2 ¼ 91%) SMD –1.46 (–2.27 to –0.65)*
TrP injection (wet needling) vs dry needling on pressure pain sensitivity
Overall effect (n ¼ 3) No Serious No Serious No Low MD 25.78 (–6.43 to 57.99)
(I2 ¼ 49%)
TrP injection (wet needling) vs dry needling on cervical lateral-flexion motion
Overall effect (n ¼ 4) No Serious No Very serious No Very low MD 2.02 (–0.19 to 4.24)
(I2 ¼ 74%)
TrP injection (wet needling) vs dry needling on depressive levels

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

Overall effect (n ¼ 3) No Serious No Very serious No Very low SMD –0.22 (–0.85 to 0.41)
(I2 ¼ 57%)

*Statistically significant (P < 0.05).

RoB: No ¼ most information is from results at low RoB; serious ¼ crucial limitation for one criterion, or some limitations for multiple criteria, sufficient to
lower confidence in the estimate of effect; very serious ¼ crucial limitation for one or more criteria sufficient to substantially lower confidence in the estimate of
effect.
Inconsistency: Serious¼ I2>40%; very serious¼I2>80%.
Indirectness of evidence: No indirectness of evidence was found in any study.
Imprecision (based on sample size): Serious¼n < 250 subjects; very serious¼n < 250, and the estimated effect is little or absent.
Publication bias (based on funnel plots): No publication bias was found. Funnel plots are not shown because of the small number of trials.

because this outcome was included in only one study. No clinicians need to consider the potential risks associated
between-groups differences were observed. Similarly, we with their application in each body area to which they
did not observe significant differences between TrP injection are applied.
(wet needling) and dry needling in changes observed in pres-
sure pain sensitivity, cervical range of motion in lateral- Strengths and Limitations
flexion, and depression. The results suggest that both nee- The results of this updated meta-analysis comparing the
dling interventions produced similar effects on these out- effects of TrP injections (wet needling) with the effects of
comes, although this conclusion should be considered with dry needling should be analyzed according to its
caution (very low evidence). strengths and weaknesses. Strengths of this meta-analysis
include a comprehensive literature search, methodologi-
Adverse Events cal rigor, data extraction, and statistical analysis. Among
Safety is an outcome highly relevant to the application of the limitations, we recognized that the number of trials
a needling intervention. Most studies reported the pres- included in the quantitative analysis was small (n ¼ 6),
ence of post-needling soreness after either TrP injections and only two were of high methodological quality.
or dry needling interventions. Boyce et al. reported that
Nevertheless, this is the largest number of trials included
minor adverse events after dry needling can be seen in up
in a meta-analysis on this topic . Additionally, the hetero-
to 37% of the patients, with bleeding (16%), bruising
geneity seen in the forest plots limits extrapolation of the
(7.7%), and pain during dry needling (5.9%) being the
results. This heterogeneity leads to the use of a random-
most frequent [36]. Post-needling soreness is attributed
effects model rather than the use of a fixed-effects model
mainly to tissue damage during needle insertion. It is im-
[37]. In fact, the results reported by Eroglu et al. [28]
portant to note that most trials included in the present
meta-analysis compared lidocaine TrP injections and dry were not included in the present meta-analysis, although
needling applied with a syringe needle instead of a solid- the inclusion and exclusion criteria were met, because no
filament acupuncture needle. The level of tissue damage post-intervention data were provided in that study and
induced by beveled-cutting-edge needles is higher than no answer from authors was obtained. Second, the dos-
that observed with the solid-filament needles commonly age (volume of lidocaine) and pH used during TrP injec-
used in dry needling [5]. In fact, one major adverse event tions (wet needling) was not clarified in most studies.
of needling thoracic and paraspinal muscles is the possi- Finally, no mid- or long-term data comparing TrP injec-
bility of needling the lung and creating a pneumothorax. tion (wet needling) vs dry needling are available.
In such a scenario, the length of the needling, instead of Therefore, large-scale, high-quality clinical trials includ-
the gauge, could be more relevant for a safe application ing longer follow-ups are necessary to determine the
of the needling procedure. Although TrP injections (wet advantages or disadvantages of TrP injections (wet nee-
needling) and dry needling seem to be safe procedures, dling) and dry needling.
524 Navarro-Santana et al.

Clinical and Research Implications 2. Vos T, Abajobir AA, Abate KH, et al. Global, regional, and na-
This meta-analysis found low evidence supporting the tional incidence, prevalence, and years lived with disability for
328 diseases and injuries for 195 countries, 1990–2016: A sys-
application of TrP injection (wet needling) for the treat-
tematic analysis for the Global Burden of Disease Study 2016.
ment of musculoskeletal neck pain associated with active Lancet 2017;390(10100):1211–59.
TrPs; however, several questions remain to be elucidated. 3. Simons DG, Travell JS. Myofascial Pain and Dysfunction. The
First, most studies investigated just short-term effects. Trigger Point Manual. 3rd edition. Philadelphia: Wolters
Further high–methodological quality randomized con- Kluwer; 2019.
trolled trials including mid- and long-term follow-ups are 4. Chiarotto A, Clijsen R, Fernandez-de-las-Pe~ nas C, et al.
Prevalence of myofascial trigger points in spinal disorders: A sys-
needed. Second, four trials investigated the isolated appli-
tematic review and meta-analysis. Arch Phys Med Rehabil 2016;
cation of lidocaine TrP injection or dry needling, which 97(2):316–37.
does not represent common clinical practice [28, 31–33]. 5. Dommerholt J, Fernandez-de-las Pe~ nas C. Trigger Point Dry
Similarly, most studies targeted only the upper trapezius Needling: An Evidence and Clinical-Based Approach. 2nd edi-

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

muscle, which, again, does not represent the clinical tion. London: Churchill Livingstone (Elsevier); 2019.
practice for patients with mechanical neck pain. In fact, 6. APTA. Description of dry needling in clinical practice: an educa-
tional resource paper. Alexandria, VA, USA. APTA Public
most studies applied the needling intervention unilater-
Policy, Pract Prof Aff Unit; 2013.
ally, when the bilateral presence of TrPs in the upper tra- 7. Cagnie B, Castelein B, Pollie F, et al. Evidence for the use of is-
pezius muscle is commonly observed in patients with chemic compression and dry needling in the management of trig-
neck pain symptoms [4]. Future clinical trials should ger points of the upper trapezius in patients with neck pain: A
identify whether adding TrP injection or dry needling to systematic review. Am J Phys Med Rehabil 2015;94(7):573–83.
multimodal and pragmatic approaches is more effective 8. Liu L, Huang QM, Liu QG, et al. Effectiveness of dry needling
for myofascial trigger points associated with neck and shoulder
not including it in this pain population. Third, we ob-
pain: A systematic review and meta-analysis. Arch Phys Med
served an inadequate reporting of injection or needling Rehabil 2015;96(5):944–55.
location; however, it is important to understand that 9. Ong J, Claydon LS. The effect of dry needling for myofascial
there is no exact anatomic location of TrPs, and most trigger points in the neck and shoulders: A systematic review and
muscles can exhibit multiple TrPs. Finally, a topic for fu- meta-analysis. J Bodyw Mov Ther 2014;18(3):390–8.
ture research may be the cost-effectiveness analysis of 10. Liu L, Huang QM, Wang LB. Whether lidocaine or dry needling
should be the favored treatment after meta analysis. J Bodyw
both TrP injection (wet needling) and dry needling inter-
Mov Ther 2014;18(4):517–8.
ventions. As lidocaine injections seem to be more expen- 11. Machado E, Machado P, Wandscher VF, et al. A systematic re-
sive than dry needling, this context requires future well- view of different substance injection and dry needling for treat-
conducted trials to evaluate their cost-effectiveness. ment of temporomandibular myofascial pain. Int J Oral
Maxillofac Surg 2018;47(11):1420–32.
12. Nouged E, Dajani J, Ku B, et al. Local anesthetic injections for
Conclusion the short-term treatment of head and neck myofascial pain syn-
On the basis of the available evidence to date, this sys- drome: A systematic review with meta-analysis. J Oral Facial
Pain Headache 2019;33(2):183–98.
tematic review and meta-analysis found low-quality evi-
13. Al-Moraissi EA, Alradom J, Aladashi O, et al. Needling thera-
dence suggesting that lidocaine TrP injection exhibits a pies in the management of myofascial pain of the masticatory
superior effect for reducing pain when compared with muscles: A network meta-analysis of randomised clinical trials. J
dry needling in patients with TrPs associated with neck Oral Rehabil 2020;47(7):910–22.
pain symptoms. Very low-quality evidence showed no 14. Moher D, Liberati A, Tetzlaff J, et al.; PRISMA Group.
significant differences in pain-related disability, pressure Preferred reporting items for systematic reviews and meta-
analyses: The PRISMA statement. PLoS Med 2009;6
pain sensitivity, and depressive levels. The RoB of the in-
(7):e1000097.
cluded trials was generally low, but the inconsistency 15. Navarro-Santana MJ, G omez-Chiguano GF, Cleland JA, et al.
(heterogeneity) and the imprecision of the results down- Effects of trigger point dry needling for nontraumatic shoulder
graded the level of evidence. Future trials investigating pain of musculoskeletal origin: A systematic review and meta-
mid-and long-term follow-up periods are needed to fur- analysis. Phys Ther 2021;101:pzaa216.
ther determine the effects of both needling interventions. 16. Fernandez-de-las-Pe~ nas C, Plaza-Manzano G, Sanchez-Infante J,
et al. Is dry needling effective when combined with other thera-
pies for myofascial trigger points associated with neck pain
Supplementary Data symptoms? A systematic review and meta-analysis. Pain Res
Manag 2021;2021:8836427.
Supplementary Data may be found online at http://painmedicine.oxfordjour- 17. Navarro-Santana MJ, Sanchez-Infante J, Fern andez-de-Las-
nals.org.
Pe~nas C, et al. Effectiveness of dry needling for myofascial trigger
points associated with neck pain symptoms: An updated system-
atic review and meta-analysis. J Clin Med 2020;9(10):3300.
References 18. Higgins JPT, Altman DG, Gøtzsche PC, et al.; Cochrane
1. Hoy D, March L, Woolf A, et al. The global burden of neck Statistical Methods Group. The Cochrane Collaboration’s tool
pain: Estimates from the Global Burden of Disease 2010 Study. for assessing risk of bias in randomised trials. BMJ 2011;343
Ann Rheum Dis 2014;73(7):1309–15. :d5928.
Dry Needling vs Trigger Point Injection for Neck Pain 525

19. Maher CG, Sherrington C, Herbert RD, et al. Reliability of the 29. Ga H, Choi JH, Park CH, et al. Acupuncture needling versus li-
PEDro scale for rating quality of randomized controlled trials. docaine injection of trigger points in myofascial pain syndrome
Phys Ther 2003;83(8):713–21. in elderly patients: A randomised trial. Acupunct Med 2007;25
20. Schünemann HJ, Oxman AD, Brozek J, et al. GRADE: Assessing (4):130–6.
the quality of evidence for diagnostic recommendations. BMJ 30. Ay S, Evcik D, Tur BS. Comparison of injection methods in myo-
Evid-Based Med 2008;13(6):162–3. fascial pain syndrome: A randomized controlled trial. Clin
21. Austin TM, Richter RR, Sebelski CA. Introduction to the Rheumatol 2010;29(1):19–23.
GRADE approach for guideline development: Considerations 31. Erolu PK, Yilmaz O,€ Bodur H, et al. A comparison of the efficacy
for physical therapist practice. Phys Ther 2014;94(11):1652–9. of dry needling, lidocaine injection, and oral flurbiprofen treat-
22. Wan X, Wang W, Liu J, et al. Estimating the sample mean and ments in patients with myofascial pain syndrome: A double-
standard deviation from the sample size, median, range and/or blind (for injection, groups only), randomized clinical trial.
interquartile range. BMC Med Res Methodol 2014;14(1):1–13.
Turkish J Rheumatol 2013;28:38–46.
23. Luo D, Wan X, Liu J, et al. Optimally estimating the sample
32. Ibrahim DA, Abdelrahem HA. Cervical region trigger point
mean from the sample size, median, mid-range, and/or mid-
Injection with dry needling versus wet needling by lidocaine in

Downloaded from https://academic.oup.com/painmedicine/article/23/3/515/6296615 by guest on 05 May 2022

quartile range. Stat Methods Med Res 2018;27(6):1785–805.
geriatric population: A comparative study. Ain-Shams J
24. Sullivan GM, Feinn R. Using effect size—or why the P value is
Anesthesiol 2019;11(1):4–9.
not enough. J Grad Med Educ 2012;4(3):279–82.
33. Raeissadat SA, Rayegani SM, Sadeghi F, et al. Comparison of
25. Deeks JJ, Higgins JPT. Chapter 9: Analyzing data and undertak-
ozone and lidocaine injection efficacy vs dry needling in myofas-
ing metanalyses. In: Higgins JPT, Churchill R, Chandl J, Cumpst
cial pain syndrome patients. J Pain Res 2018;11:1273–9.
MS, eds. Cochrane Handbook for Systematic Review of
34. Cleland JA, Childs JD, Whitman JM. Psychometric properties of
Intervention. Version 5.2.0 (updated June 2017). Cochrane;
2017. Available at: www.training.cochrane.org/handbook the Neck Disability Index and Numeric Pain Rating Scale in
(accessed July 20, 2020). patients with mechanical neck pain. Arch Phys Med Rehabil
26. Krishnan SK, Benzon HT, Siddiqui T, et al. Pain on intramuscu- 2008;89(1):69–74.
lar injection of bupivacaine, ropivacaine, with and without dexa- 35. Bijur PE, Latimer CT, Gallagher EJ. Validation of a verbally ad-
methasone. Reg Anesth Pain Med 2000;25(6):615–9. ministered numerical rating scale of acute pain for use in the
27. Hong CZ. Lidocaine injection versus dry needling to myofascial emergency department. Acad Emerg Med 2003;10(4):390–2.
trigger point: The importance of the local twitch response. Am J 36. Boyce D, Wempe H, Campbell C, et al. Adverse events associ-
Phys Med Rehabil 1994;73(4):256–63. ated with therapeutic dry needling. Int J Sports Phys Ther 2020;
28. Kamanli A, Kaya A, Ardicoglu O, et al. Comparison of lidocaine 15(1):103–13.
injection, botulinum toxin injection, and dry needling to trigger 37. Higgins J, Green S. Cochrane Handbook for Systematic Review
points in myofascial pain syndrome. Rheumatol Int 2005;25 Interventions. 2011. Available at: http://handbook.cochrane.org
(8):604–11. (accessed July 2020).