Documente Academic
Documente Profesional
Documente Cultură
DAVID M. KIETRYS, PT, PhD, OCS1 KERSTIN M. PALOMBARO, PT, PhD, CAPS2 ERICA AZZARETTO, PT, DPT3
RICHARD HUBLER, PT, DPT4 BRET SCHALLER, PT, DPT5 J. MATHEW SCHLUSSEL, DPT MARY TUCKER, PT, DPT6
TTSTUDY DESIGN: Systematic review and meta- 23 to 40 points, with a mean of 34 points (scale
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
M
muscle called myofascial trigger points. When pal-
pated, active myofascial trigger points cause local patients with upper-quarter MPS, with an overall
or referred symptoms, including pain. Dry needling effect favoring dry needling. The findings of 2 stud- yofascial pain syndrome
involves inserting an acupuncture-like needle into ies that compared dry needling to sham or placebo (MPS) is a common
a myofascial trigger point, with the goal of reducing treatment provided evidence that dry needling can
condition associated
pain and restoring range of motion. decrease pain after 4 weeks in patients with upper-
with myofascial trigger
TTOBJECTIVE: To explore the evidence regarding quarter MPS, although a wide confidence interval
for the overall effect limits the impact of the effect. points (MTrPs).27 MTrPs are a
Journal of Orthopaedic & Sports Physical Therapy
1
School of Health Related Professions, Rutgers, The State University of New Jersey, Stratford, NJ. 2Institute of Physical Therapy Education, Widener University, Chester, PA.
3
NovaCare Rehabilitation, Philadelphia, PA. 4NovaCare Rehabilitation, Sewell, NJ. 5Comber Physical Therapy, Williamsburg, VA. 6SportsCare of America, Bridgewater, NJ. The
authors certify that they have no affiliations with or financial involvement in any organization or entity with a direct financial interest in the subject matter or materials discussed
in the article. Address correspondence to Dr David Kietrys, 40 East Laurel Road, UEC Suite 2105, Stratford, NJ 08084. E-mail: kietrydm@shrp.rutgers.edu t Copyright 2013
Journal of Orthopaedic & Sports Physical Therapy
620 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy
searching, n = 2 MPS, n = 11
T
Recent article published online Of remaining, did he studies included in this sys-
(ahead of print), n = 1 not include DN as tematic review and meta-analysis
an intervention had human subjects, were random-
group, n = 5 ized controlled trials (had a control or
Retained for literature review, n = 12 comparison group), had a dry-needling
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
when compressed.13,16,18,20,33,36,42 Palpat- has been suggested that the gate control were not randomized controlled trials
ing an MTrP or inserting a needle into theory of pain may play a role.14 Dry nee- were excluded. Next, articles that did
an MTrP may elicit a localized twitch dling causes stimulation of alpha-delta not involve subjects with upper-quarter
response, defined as a brisk contrac- nerve fibers, thus activating the enkepha- myofascial pain and articles that did not
tion of muscle fibers in or around the linergic inhibitory dorsal horn interneu- include dry needling as an intervention
MTrP.13,16,18,33,36,42 Localized twitch re- rons and causing opioid-mediated pain group were excluded.
sponses are more easily elicited when suppression.2 Dry needling may correct Our initial search produced a sys-
sensitive loci within an MTrP are identi- levels of several chemicals in the affected tematic review and meta-analysis re-
fied and targeted.16-19 muscles, including bradykinin, calcito- garding dry needling and acupuncture
nin gene-related peptide, and substance in the management of MTrP pain.41 A
Dry Needling P.10 Needling of MTrPs is also theorized hand search of that review produced 2
Trigger-point dry needling is a procedure to disrupt reverberatory central nervous articles that met our inclusion criteria
in which an acupuncture-like needle is system circuits.30 that were not previously identified with
inserted into the skin and muscle in the A previously published systematic our electronic search. All other key refer-
location of an MTrP.11 Needles are re- review of 7 studies of acupuncture/dry ences,1,4,9,15,17,20-23,26,43 as well as 1 other sys-
moved once the trigger point is inactivat- needling for the management of MTrPs tematic review8 on the topic, were hand
ed. Dry needling is typically followed by in various body regions (including the searched but did not yield any additional
stretching exercises.14 The actual mecha- upper quarter, low back, and lower ex- articles. One article39 published online
nism of effect of dry needling is still being tremity) found limited evidence in 1 study (ahead of print) in November 2012 was
debated. The localized twitch response that dry needling had an overall effect added to the review.
that often occurs may interrupt motor compared to standardized care.41 Meta- Retained articles were scored inde-
end-plate noise, eliciting an analgesic analysis of 4 studies comparing dry nee- pendently for internal validity using the
effect.10 Eliciting a localized twitch re- dling to a sham (placebo) treatment did evaluation guidelines for rating the qual-
sponse and stretching exercises relax the not show statistical significance between ity of an intervention study (the Mac-
journal of orthopaedic & sports physical therapy | volume 43 | number 9 | september 2013 | 621
sensus score. If the reviewers could not combined. In a meta-analysis, Kamanli et adults over 60 years of age. Duration of
reach a consensus score to within 1 point, al26 and Itoh et al23 both assessed the ef- symptoms varied among studies; par-
an additional reviewer was used to adju- fects of dry needling in comparison to 2 ticipants in 8 of the studies had chronic
dicate the score. If a consensus could still different treatments at 4 weeks. The data symptoms ranging from 3 months23 to
not be reached, the lower score was as- for each of these other treatments were 63 months39 in duration. One study9
signed. In addition, the studies reviewed entered separately; therefore, these 2 included participants whose shoulder
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
were assigned a level-of-evidence rating studies are each represented twice in the symptoms started following a stroke. The
as described by Sackett et al.34 All authors meta-analysis of dry needling compared study by Ilbuldu et al21 included only fe-
(except K.M.P.) participated in extraction to other treatments at approximately 4 male participants, whereas all other stud-
of relevant data related to MacDermid weeks. ies appear to have included individuals of
Quality Checklist scoring. We used 2 points on a 0-to-10 VAS both genders.
Two of the authors (D.M.K. and as a conservative cutoff value for clini- Intervention groups (independent
K.M.P.) worked as a team to extract cal meaningfulness of change in pain for variables), outcome measurements (de-
relevant data related to meta-analyses. between-group comparisons. Various pendent variables), and times to out-
Meta-analyses were performed with studies have reported a range of minimal comes are summarized in TABLE 3. Six of
Journal of Orthopaedic & Sports Physical Therapy
MetaAnalyst Version Beta 3.13 (Tufts clinically important difference values for the studies used a true control (placebo
Medical Center, Boston, MA), with a numeric or visual analog pain scales for or sham) group.4,21-23,39,43 One study used
continuous-variable random-effects patients with upper-quarter pathologies, the contralateral side of the participants
model. Four separate meta-analyses were including 1 point for patients with chron- as the control.20 Eight studies utilized
performed, with pain on a visual analog ic musculoskeletal pain,35 1.3 points for a variety of comparison groups (groups
scale (VAS) as the outcome measure: (1) neck pain,5 1.7 points for chronic pain,12 that received interventions other than dry
dry needling compared to sham or con- 2.17 points for shoulder pain,31 and 3.0 needling to MTrPs). Comparison groups
trol, immediate effects; (2) dry needling points for patients with neck/upper ex- included lidocaine injection,1,17,26 botuli-
compared to sham or control at 4 weeks; tremity/lower extremity pain.38 num toxin injection,26 laser,21 nonlocalized
(3) dry needling compared to other treat- acupuncture,22,23 and standard rehabilita-
ments, immediate effects; and (4) dry RESULTS tion (external support, positioning, exer-
needling compared to other treatments cise) for hemiparetic shoulder pain.9 The
T
at approximately 4 weeks. All studies that welve studies that met our in- comparison group in the study by Ga et
compared dry needling to other treat- clusion criteria1,4,9,15,17,20-23,26,39,43 are al15 received a treatment (intramuscular
ments provided data at 4 weeks, with the listed in chronological order in stimulation) that, technically, is a dry-
exception of the study by DiLorenzo et TABLES 1 through 6. Inclusion and exclu- needling technique, with subtle differ-
al,9 which measured outcomes at 21 days. sion criteria for participants in the re- ences in technique between the authors
These data were used in the comparisons viewed studies are described in TABLE 1. operational definitions of dry needling
at approximately 4 weeks. Outcomes at In all studies, subjects had symptoms at- and intramuscular stimulation. Times to
times other than immediately after and tributed to upper-quarter MPS, typically outcomes ranged from immediate4,17,20,22,43
approximately 4 weeks after treatment involving the neck or shoulder region. to 6 months,21 with 4 studies17,20,22,43 re-
were not considered in this review, due Etiology of pain was not consistent across porting only immediate effects.
to variability across studies in other studies. For example, DiLorenzo et al9 TABLE 4 describes the key findings,
times to outcomes. The VAS pain scores included subjects with shoulder pain fol- MacDermid Quality Checklist scores, and
622 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy
Head injury
Congenital cerebral palsy
Worsening or pre-existing internal derangement of shoulder ligaments or tendons
Adhesive capsulitis
Peripheral neuropathy
Complex regional pain syndrome
Shoulder fractures
Neglect syndrome
Decline participation
Kamanli et al26 A
t least 1 MTrP on CS, back, or shoulder muscles with disease of at T reatment in prior 8 wk
least 6 mo in duration MTrP injection within prior 2 mo
Cardiovascular or respiratory disease
Allergies
CS or shoulder surgery in prior year
Fibromyalgia
CS radiculopathy or myelopathy with severe disc or skeletal lesions
Uncooperative
Use of medications that prevent neuromuscular transmission
Motor neuron or neuromuscular junction disease
Pregnancy
Table continues on page 624.
level-of-evidence ratings. Scores for each published.3 Levels of evidence34 ranged ternal validity, as evidenced by relatively
of the 24 items on the MacDermid Qual- from 2b4 to 1b.1,9,15,17,20-23,26,39,43 Internal va- higher scores on the MacDermid Quality
ity Checklist are provided in TABLE 5. The lidity scores (MacDermid Quality Check- Checklist, were those by Tekin et al,39 Ga
criteria and description of the scoring list) ranged from 234 to 40,39 with a mean et al,15 and Irnich et al.22 The studies with
system for this tool have been previously of 34. The articles with the strongest in- the weakest internal validity were those
journal of orthopaedic & sports physical therapy | volume 43 | number 9 | september 2013 | 623
No significant differences in clinical presentation between 2 sides Contraindication for DN, such as local infection, serious medical problems, recent
multiple trauma, or pregnancy with threatened abortion
Condition that might interfere with pain/pain threshold assessment
CS or UE surgery
Itoh et al23 N eck pain for 6 mo or longer with no radiation M ajor trauma or systemic disease
Normal CS nerve function Other conflicting or ongoing treatments, except medication with uniform dosage for 1 mo
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
in the upper trapezius (MTrP diagnosed as tenderness and pain Anticoagulant medication
reproduction with palpation of a tight band) Pregnancy with threatened abortion
Problem that might interfere with pain/pain threshold assessment
Cognitive deficit
Needling treatment in past
Tekin et al39 M PS (local spontaneous pain, referred pain or sensory changes hysical therapy or local injection within prior 3 mo
P
from MTrP, palpable taut band, localized tenderness, reduced ROM) Fibromyalgia
At least 1 active MTrP Pregnancy
Symptom duration at least 6 mo Cervical nerve root irritation
Abnormal lab results
Thoracic outlet syndrome
Upper-limb entrapment syndromes
Abbreviations: CS, cervical spine; CVA, cerebrovascular accident; DN, dry needling; LTR, localized twitch response; MPS, myofascial pain syndrome; MTrP,
myofascial trigger point; ROM, range of motion; UE, upper extremity; VAS, visual analog scale.
by Hsieh et al,20 Chu,4 and Hong.17 As in- sham or control and assessed immedi- effect20 used the same subjects unin-
dicated in TABLE 4, all studies reported sig- ate effects on pain (FIGURE 3).20,22,39,43 The volved side as the control, and reported a
nificant decreases in pain in the groups overall effect size (standardized mean dif- raw between-group effect size of 4.0 VAS
receiving dry needling. In many cases, ference) of 1.06 (95% confidence interval points, which is clinically meaningful.
comparison groups also realized an im- [CI]: 0.05, 2.06) suggests a large effect7 The other 2 studies that favored dry nee-
provement in pain. favoring dry needling over sham or con- dling39,43 had large treatment effects (0.88
trol. Heterogeneity was high (I2 = 86.3%). and 0.75, respectively), but their raw be-
Meta-analysis: Dry Needling Compared Three of the 4 studies entered into this tween-group effect sizes (1.4 and 1.2 VAS
to Sham or Control, Immediate Effects meta-analysis favored dry needling. points, respectively) were of questionable
Four studies compared dry needling to The study with the largest treatment clinical meaningfulness.
624 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy
Irnich et a l22
36 51.9 36.7 mo al22 used nonlocalized acupuncture as the
Ilbuldu et al21 60 35.3 9.2 38.5 31.9 mo other treatment. Hong17 reported results
33.9 10.4
32.9 28.6 mo separately for subjects who had a local-
32.3 6.9
36.5 33.6 mo ized twitch response and those who did
DiLorenzo et al9 101 69.6 6.2 3.53 wk not, and these data were entered sepa-
67.4 9.1 rately into the meta-analysis because
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
Kamanli et al26 29 37.2 8.1 32.5 22.0 mo the results could not be combined. The
37.3 9.8 49.2 35.0 mo overall effect size (standardized mean dif-
38.3 5.3 50.7 19.9 mo ference) of 0.64 (95% CI: 1.21, 0.06)
Ga et al15 40 79.2 6.8 suggests a moderate effect7 favoring other
76.3 8.6 treatment over dry needling. Hetero-
Hsieh et al20 14 60.2 13.2 geneity was high (I2 = 90%). Although
Itoh et al23 40 62.3 10.1 2.9 2.7 y both studies entered into this meta-anal-
62.3 11.0 3.2 3.1 y ysis favored other treatment, the raw be-
65.0 10.5 3.3 3.9 y tween-group effect sizes (0.58-1.69 VAS
Journal of Orthopaedic & Sports Physical Therapy
65.0 10.5 2.3 1.5 y points for Hong17 and 1.01 VAS points for
Ay et al1 80 38.1 9.8 34.3 40.9 mo Irnich et al22) were of questionable clini-
37.2 10.1 30.6 37.2 mo cal meaningfulness.
Tsai et al43 35 46.4 12.2 7.5 3.9 mo
41.5 10.4 6.8 4.5 mo Meta-analysis: Dry Needling Compared
Tekin et al39 39 42.9 10.9 63.5 50.7 mo to Other Treatments at Approximately
42.0 12.0 57.9 48.3 mo 4 Weeks
*Values are mean SD where those data were provided by the authors. Six studies compared the effects of dry
Reported age and duration of symptoms based on occurrence of a localized twitch response; the sub-
group that experienced a localized twitch response is listed first. needling to other forms of treatment on
Dry-needling group. pain at 4 weeks (FIGURE 6).1,9,15,21,23,26 Two
Comparison group(s). of the studies included 2 other treat-
Reported age and duration of symptoms based on pain relief outcome; subgroup experiencing pain
relief listed first. ment groups, and the results from each
Control (placebo or sham) group. of these treatments were entered sepa-
rately into the meta-analysis, such that
8 data sets were entered. The overall ef-
Meta-analysis: Dry Needling Compared 95% CI crosses the line of no difference, fect size (standardized mean difference)
to Sham or Control at 4 Weeks suggesting that caution should be used of 0.07 (95% CI: 1.39, 1.26) suggests a
Three studies compared the effects of when making conclusions based on over- small overall effect favoring other treat-
dry needling to sham or control on pain all effect size. Heterogeneity was high (I2 ment, with the 95% CI crossing the line
at 4 weeks (FIGURE 4).21,23,39 The overall ef- = 84.2%). Two of the 3 studies23,39 in this of no difference. Heterogeneity was high
fect size (standardized mean difference) meta-analysis favored dry needling over (I2 = 95%). Two of the studies9,23 entered
of 1.07 (95% CI: 0.21, 2.35) suggests a the sham or control at 4 weeks, and both into this meta-analysis favored dry nee-
large effect favoring dry needling over had large effect sizes (1.95 and 1.55, re- dling over other treatment at 4 weeks,
sham treatment or control; however, the spectively). Both had raw between-group and both had large7 effect sizes (2.26
journal of orthopaedic & sports physical therapy | volume 43 | number 9 | september 2013 | 625
626 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy
Improvement in rating of general complaints in nonlocalized acupuncture group compared to DN group or sham laser group
Ilbuldu et al21 Improved CS flexion in DN group compared to laser group at 1 mo 36/1b
Improved CS extension and lateral flexion in laser group compared to DN group (P<.001 for both) or sham laser group (P<.001, P<.01,
respectively) at 1 mo
Decreased pain in laser group at rest (P<.05) and with activity (P<.001) compared to DN group or sham laser group at 1 mo
Improved pressure pain threshold in laser group compared to DN group or sham laser group (P<.001 for both) at 1 mo
Improved health profile scores in laser group compared to DN group or sham laser group (P<.05 for both) at 1 mo
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
DiLorenzo et al9 D ecreased shoulder pain in both DN and rehabilitation groups on day 9, 15, and 21 35/1b
Greater decrease in pain in DN group compared to rehabilitation group at day 9 and 21
Kamanli et al26 Improved pain score (all groups) (P<.05) 37/1b
Improved pressure pain threshold (all groups) (P<.05); greater decrease in lidocaine injection group (P<.016)
Improved fatigue and work disability in lidocaine injection and botulinum injection groups (P<.05)
Improved CS ROM (all groups) (P<.05)
Improved health profile score in lidocaine injection and botulinum toxin groups (P<.05)
Improved anxiety and depression scale scores in botulinum toxin group (P<.05)
Ga et al15 ecreased pain (both groups) at 28 d (P<.001)
D 39/1b
Improved pressure pain threshold (both groups) at 28 d (P<.001)
Improved depression scale score at 28 d in IMS group (P = .024)
Journal of Orthopaedic & Sports Physical Therapy
journal of orthopaedic & sports physical therapy | volume 43 | number 9 | september 2013 | 627
Item
Study 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Total
Hong17 2 2 1 2 1 1 1 2 1 2 0 1 2 0 2 1 2 1 1 1 1 1 0 2 30
Chu4 1 1 1 2 1 1 1 0 1 1 0 0 2 1 1 1 1 0 2 1 2 0 1 1 23
Irnich et al22 2 1 2 2 2 1 1 2 1 2 0 2 2 2 2 2 1 0 2 2 2 2 2 2 39
Ilbuldu et al21 2 2 2 2 1 1 1 2 1 2 0 2 1 0 2 2 2 2 2 1 1 2 2 1 36
DiLorenzo et al9 2 2 2 2 1 1 1 0 1 2 0 2 2 0 2 2 2 1 2 1 2 1 2 2 35
Kamanli et al26 2 2 2 2 1 1 1 0 1 2 0 2 2 1 2 2 2 1 2 2 1 2 2 2 37
Ga et al15 2 2 2 2 1 2 1 0 1 2 0 2 2 1 2 2 2 2 2 2 1 2 2 2 39
Downloaded from www.jospt.org at on September 4, 2014. For personal use only. No other uses without permission.
Hsieh et al20 2 0 2 2 0 1 1 0 1 1 0 2 1 0 0 2 1 1 2 1 2 2 0 2 26
Itoh et al23 2 2 2 2 1 2 1 2 1 2 0 1 2 0 2 2 1 2 2 1 1 0 2 2 35
Ay et al1 2 2 2 2 2 1 1 0 1 2 0 2 2 0 2 1 1 2 2 2 2 2 0 1 34
Tsai et al43 2 2 1 2 2 2 1 2 1 2 0 2 2 0 2 2 2 0 2 2 1 2 1 2 37
Tekin et al39 2 2 2 2 2 2 1 2 1 2 1 1 2 2 2 2 2 1 2 2 1 1 1 2 40
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
TABLE 6 Summary of Key Methodological Issues and Outcomes by Study
and 1.48-2.15, respectively). In the study dling was compared to dry needling of 1.08, respectively). Ay et al1 also reported
by DiLorenzo et al, 9 in which dry nee- nontender points or to acupuncture. In a large effect favoring lidocaine injection
dling was compared to rehabilitation, the studies that favored the comparison over dry needling (3.30), but the raw
the raw between-group effect size at ap- (other) treatment, only Kamanli et al26 between-group effect size of 1.55 VAS
proximately 4 weeks approached clini- reported clinically meaningful raw be- points (at 4 weeks) was of questionable
cal meaningfulness (1.81 VAS points). tween-group effect sizes at 4 weeks (2.44 clinical meaningfulness.
The raw between-group effect size be- VAS points favoring botulinum toxin Ilbuldu et al21 reported statistical
tween groups at 4 weeks was clinically injection and 3.17 VAS points favoring significance and a moderate7 effect size
meaningful (2.73-3.98 VAS points) in lidocaine injection), with corresponding (0.71) favoring laser over dry needling
the study by Itoh et al,23 where dry nee- large7 treatment effect sizes (0.83 and at 4 weeks, but meta-analysis results
628 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy
Precision
3.0 2.5 differences in inclusion criteria. Hsieh
2.5
2.0 2.0 et al20 used the same subjects unin-
1.5 volved side as the control, Irnich et al22
1.5
1.0 used sham laser acupuncture, and Tsai
0.5 1.0
et al43 and Tekin et al39 used sham nee-
1 0 1 2 3 4 5 0.5 0.0 0.5 1.0 1.5 2.0 2.5
dling. Despite the high heterogeneity, 3
Effect Effect of the 4 studies provided evidence of a
large7 effect of dry needling compared to
C D sham or control. However, such results
Downloaded from www.jospt.org at on September 4, 2014. For personal use only. No other uses without permission.
Precision
3.0
2.5
tionable clinical meaningfulness.39,43 The
2.5
2.0
data by Chu4 were not included in the
2.0
meta-analysis because they could not be
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
1.5 1.5
extracted in a way conducive to inclusion
1.0 1.0
2.0 1.5 1.0 0.5 0.0 4 3 2 1 0 1 2 3
in the meta-analysis. Chu4 reported a
greater percentage of subjects with pain
Effect Effect relief for the dry-needling group com-
pared to the control group (P<.0001).
FIGURE 2. Funnel plots for meta-analyses showing (A) dry needling compared to sham or control immediately However, the internal validity of that
after treatment, (B) dry needling compared to sham or control at approximately 4 weeks, (C) dry needling
study was the weakest of the 12 stud-
compared to other treatment immediately after treatment, (D) dry needling compared to other treatment at
approximately 4 weeks. The diameter of the circles represents the standardized mean difference of each study, ies reviewed, with a score of 23 points
with larger diameters corresponding to larger standardized mean differences. on the MacDermid Quality Checklist.
Journal of Orthopaedic & Sports Physical Therapy
I
intramuscular stimulation. However, in- nterpretation of the collective or Control at 4 Weeks
tramuscular stimulation is very similar body of results of the studies reviewed At 4 weeks, 2 studies23,39 provided evi-
to dry needling, and therefore the lack of is complicated due to the variance in dence of a strong effect of dry needling
difference was expected. comparison groups, control conditions, compared to a sham or control, with clini-
dosage of intervention, outcomes, out- cally meaningful raw between-group ef-
Publication Bias come measurement tools, times to out- fect sizes. Although the overall effect was
Funnel plots (FIGURE 2) were created to comes, and internal validity (quality) of strong, it was confounded by a wide 95%
determine the risk of publication bias the studies. The studies that have been CI due to the equivocal findings of the
for the 4 separate meta-analyses. The published to date were conducive to the study by Ilbuldu et al.21 It was unclear if
funnel plots for dry needling compared 4 meta-analyses described, but the high the examiners in the Ilbuldu et al21 study
to sham or control for both immediate heterogeneity for all analyses performed were blinded, and a low number of sub-
effects and at 4 weeks, as well as the requires special consideration. jects (n = 40) without a priori power anal-
funnel plot for the immediate effects of ysis might have contributed to the finding
dry needling compared to other treat- Dry Needling Compared to Sham of a lack of difference between groups
ments, were asymmetrical, demonstrat- or Control, Immediate Effects (type II error). The high heterogeneity for
ing a risk for publication bias. The funnel In studies that compared dry needling this meta-analysis (84.2%) may, in part,
plot for dry needling compared to other to sham or control, high heterogeneity be explained by the small number of stud-
journal of orthopaedic & sports physical therapy | volume 43 | number 9 | september 2013 | 629
0 1 2 3 4
Favors Favors dry
control needling
*Values are pain scores immediately posttreatment. Outcome measure was pain rating on a 0-to-10 visual analog scale.
Downloaded from www.jospt.org at on September 4, 2014. For personal use only. No other uses without permission.
Values are standardized mean difference (95% confidence interval). In the plots, the squares represent point estimates of treatment effect; bigger squares
indicate larger samples; the diamond represents the pooled treatment effect; the horizontal lines are 95% confidence intervals; and the vertical line represents
no difference. Tests for heterogeneity: 2 = 0.855, df = 3.0 (P<.001), I2 = 86.3%.
Approximately 4 Weeks
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
0 1 2 3
Favors Favors dry
control needling
Journal of Orthopaedic & Sports Physical Therapy
*Values are pain scores immediately posttreatment. Outcome measure was pain rating on a 0-to-10 visual analog scale.
Values are standardized mean difference (95% confidence interval). In the plots, the squares represent point estimates of treatment effect; bigger squares
indicate larger samples; the diamond represents the pooled treatment effect; the horizontal lines are 95% confidence intervals; and the vertical line represents
no difference. Tests for heterogeneity: 2 = 1.042, df = 2.0 (P = .002), I2 = 84.2%.
ies and the variance in sham or control provided evidence that a lidocaine in- were some differences in the subject in-
conditions (eg, Ilbuldu et al21 used sham jection had a greater effect on pain, ap- clusion criteria between these studies.
laser, Itoh et al23 used sham acupuncture, proaching clinical meaningfulness, when
and Tekin et al39 used sham needling). In the treatments did not induce a localized Dry Needling Compared to Other
addition, there were differences in the twitch response. When a localized twitch Treatments at Approximately 4 Weeks
inclusion criteria of these studies. More response was associated with the treat- Based on 6 studies, dry needling is not
high-quality randomized controlled tri- ments, the difference between lidocaine superior, in general, to the other treat-
als are needed to further elucidate the injection and dry needling was neither ments studied to reduce pain at 4 weeks.
effects of dry needling compared to sham significant nor clinically meaningful. However, the overall small7 effect (0.07,
or placebo on pain at 4 weeks and other This finding supports the theory that a favoring other treatment) must be viewed
clinically relevant time points. localized twitch response is an important with caution because of the high hetero-
component of effective dry needling. The geneity (95%) attributable to the vari-
Dry Needling Compared to Other high heterogeneity (90%) in this meta- ety of other treatments, dosages of dry
Treatments, Immediate Effects analysis is partly explained by the small needling, and diagnoses of the subjects.
Based on 2 studies,17,22 dry needling is number of studies and the variety in Two studies1,26 provided evidence that a
not superior to lidocaine injection or comparison treatments: Hong17 used li- lidocaine injection or botulinum toxin
nonlocal acupuncture to decrease pain docaine injection and Irnich et al22 used injection had a greater effect than dry
immediately after treatment. One study17 nonlocal acupuncture. In addition, there needling on reducing pain, with raw be-
630 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy
indicate larger samples; the diamond represents the pooled treatment effect; the horizontal lines are 95% confidence intervals; and the vertical line represents
no difference. Tests for heterogeneity: 2 = 1.633, df = 3.0 (P<.001), I2 = 90.0%.
Approximately 4 Weeks
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
Ay et al1 (lidocaine) 40 3.82 0.47 40 2.27 0.98 11.2% 3.30 (4.15, 2.45)
Total 168 164 100% 0.07 (1.39, 1.26)
4 3 2 1 0 1 2 3
Favors other Favors dry
treatment needling
*Values are pain scores immediately posttreatment. Outcome measure was pain rating on a 0-to-10 visual analog scale.
Values are standardized mean difference (95% confidence interval). In the plots, the squares represent point estimates of treatment effect; bigger squares
indicate larger samples; the diamond represents the pooled treatment effect; the horizontal lines are 95% confidence intervals; and the vertical line represents
no difference. Tests for heterogeneity: 2 = 3.417, df = 7.0 (P<.001), I2 = 95.0%.
tween-group effect sizes that were clini- for pain scores that was clinically mean- tions might have influenced the results
cally meaningful. When dry needling was ingful. Despite the high heterogeneity of regarding the relative contribution of dry
compared to standard rehabilitation in this meta-analysis, the mixed results, and needling (or other interventions) to treat-
subjects with shoulder pain following lack of overall effect, close inspection of ment effects. For example, in the studies
a cerebrovascular accident,9 dry nee- the design of individual studies suggests by Ay et al1 and Ilbuldu et al,21 subjects
dling was favored (with a strong effect) that dry needling may be superior to oth- in all groups performed stretching exer-
over rehabilitation, with a raw between- er treatments, depending on the other cises. In these studies, it is possible that
group effect size that approached clini- treatment and patient diagnoses. How- the stretching exercises contributed to
cal meaningfulness. In another study of ever, when dry needling is compared to li- the treatment effects.
patients with neck pain,23 dry needling docaine injection in patients with MTrPs
was favored (with a large7 effect) over dry in the neck, upper back, or shoulder,26 Importance of the Localized Twitch
needling of nontender points or acupunc- lidocaine injection may be superior. Response in Dry Needling
ture, with a raw between-group effect size In some cases, combined interven- Many descriptions of dry-needling
journal of orthopaedic & sports physical therapy | volume 43 | number 9 | september 2013 | 631
comparing dry needling with lidocaine did not have an effect on our conclusions forms of therapy, such as exercise and
injection, Hong17 noted that a lack of lo- or recommendations. Of great concern postural correction. Several studies in
calized twitch response in either group was the high heterogeneity in each of the this review reported statistical superi-
was associated with little change in pain, 4 meta-analyses we performed. In gen- ority of dry needling compared to sham
tenderness, or range of motion. Ga et al15 eral, such high heterogeneity may bring or other outcomes, including pain pres-
compared dry needling with intramus- into question whether it is even appropri- sure threshold,17,43 range of motion,17,22,43
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
cular stimulation, a variation of dry nee- ate to perform a meta-analysis. However, self-reported disability,23 and number of
dling that involves grasping and winding our discussion of likely reasons for this tender MTrPs.4 A limitation of this sys-
up of the muscle (by the needle) and a high heterogeneity and our consideration tematic review was that it did not provide
stronger stimulation response. Local- of findings of individual studies provide analyses of these secondary variables.
ized twitch response rates were not dif- a rationale to pursue the meta-analyses. All studies reviewed had methodologi-
ferent between the groups, with nearly Another limitation of this review is cal limitations, which were extensive in
all participants demonstrating localized the evidence of publication bias in the some cases. Key methodological limita-
twitch responses during treatment. Both asymmetrical funnel plots (FIGURE 2) for tions of the studies are summarized in
groups had decreased pain and improved dry needling compared to sham or con- TABLE 6. Only 1 study22 provided a cursory
Journal of Orthopaedic & Sports Physical Therapy
pain pressure threshold at 4 weeks. Fur- trol for both immediate effects and at 4 interpretation of pain reduction from
ther research is needed to clarify wheth- weeks, as well as dry needling compared the perspective of minimal clinically im-
er a localized twitch response is a valid to other treatments for immediate effects. portant difference. The parameters of
predictor of success or a necessary com- Publication bias may result from a lower dry-needling treatment technique var-
ponent of dry-needling treatment in pa- publication rate of negative results, exclu- ied across studies. The studies by Chu4
tients with upper-quarter MPS. However, sion of publications in foreign languages, and Ga et al15 referred to intramuscular
it does appear that provocation of a local- or an inability to access work not submit- stimulation as a consideration in dry nee-
ized twitch response is common with the ted for publication.6 The authors did not dling, with Ga et al15 actually using intra-
dry-needling technique. attempt to locate unpublished research or muscular stimulation as a comparison
research in foreign languages examining group. Times to outcomes varied across
Limitations the impact of dry needling on patients studies, with 4 reporting only immediate
The limitations of this review include the with upper-quarter MPS. However, fun- effects.17,20,22,43 The immediate effects on
use of only 1 search term (dry needling). nel-plot asymmetry can be influenced by pain are of interest, but longer-term ef-
However, based on the hand search of the heterogeneity of studies included in fects on a comprehensive group of func-
references from 2 other systematic re- a meta-analysis40 and can be challenging tional and clinically relevant measures
views,8,41 it is unlikely that any relevant to interpret when the number of studies should be considered when designing
articles were overlooked. Our methods included is small.6 Thus, the asymmetri- future studies. In general, future stud-
did not permit us to calculate concor- cal funnel plots in this study cannot be ies should be carefully designed to avoid
dance statistics for data extraction. The interpreted conclusively due to the small many of the methodological limitations
authors recognize the value of this infor- number of studies included (range, 3-4) found in the studies published to date.
mation in retrospect but cannot adjust as well as the heterogeneity of those stud- The external validity of several of the
for this aspect of the methodology. ies (range, 84.2%-90%). studies is limited due to the age ranges
Other tools, such as the PEDro scale,29 Because most studies of longer-term and gender bias of the sample. Four stud-
are available to rate the internal valid- effects described outcomes at approxi- ies9,15,20,23 focused on an older sample,
632 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy
B
ased on the studies published favoring dry needling or the alternate apmr.2001.24023
9. DiLorenzo L, Traballesi M, Morelli D, et al. Hemi-
to date, we recommend (grade A)34 intervention.
paretic shoulder pain syndrome treated with
dry needling, compared to sham IMPLICATIONS: We recommend (grade A)34 deep dry needling during early rehabilitation: a
or placebo treatment, for immediate re- dry needling for immediate reduction prospective, open-label, randomized investiga-
duction of pain in patients with upper- of pain in patients with upper-quarter tion. J Musculoskelet Pain. 2004;12:25-34.
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
MPS, based on results of 2 individual performed here. Variance in study fac- Myofascial pain syndrome of the head and
randomized controlled trials23,39 included tors, such as control conditions and neck: a review of clinical characteristics of
in a meta-analysis of 3 studies. However, comparison treatments, contributed to 164 patients. Oral Surg Oral Med Oral Pathol.
1985;60:615-623.
it must be noted that the overall effect high heterogeneity in the results of the
14. Furlan AD, van Tulder M, Cherkin D, et al. Acu-
of the 3 studies combined is ambiguous meta-analyses. puncture and dry-needling for low back pain: an
due to a large CI of the otherwise strong updated systematic review within the framework
effect size. Future studies should be criti- of the Cochrane Collaboration. Spine (Phila Pa
cally reviewed to inform the evolution REFERENCES 1976). 2005;30:944-963.
15. Ga H, Choi JH, Park CH, Yoon HJ. Dry needling
of these recommendations. Additional of trigger points with and without paraspinal
1. A y S, Evcik D, Tur BS. Comparison of injec-
research with high-quality study design tion methods in myofascial pain syndrome: a needling in myofascial pain syndromes in
and appropriate choices of comparative elderly patients. J Altern Complement Med.
randomized controlled trial. Clin Rheumatol.
2007;13:617-624. http://dx.doi.org/10.1089/
treatments will aid in developing more 2010;29:19-23. http://dx.doi.org/10.1007/
acm.2006.6371
conclusive evidence for dry needling. s10067-009-1307-8
16. Hong CZ. Considerations and recommendations
2. Baldry PE. Myofascial Pain and Fibromyalgia
More evidence is needed to establish ef- regarding myofascial trigger point injection. J
Syndromes: A Clinical Guide to Diagnosis and
ficacy of dry needling compared to other Musculoskelet Pain. 1994;2:29-59.
Management. Edinburgh, UK: Churchill Living- 17. Hong CZ. Lidocaine injection versus dry nee-
interventions for upper-quarter MPS. stone; 2001. dling to myofascial trigger point. The importance
However, it appears that injection with 3. Brudvig TJ, Kulkarni H, Shah S. The effect of of the local twitch response. Am J Phys Med
lidocaine may be superior to dry needling therapeutic exercise and mobilization on pa- Rehabil. 1994;73:256-263.
tients with shoulder dysfunction: a systematic 18. Hong CZ. Myofascial trigger point injection. Crit
for pain reduction both immediately after
treatment and at 4 weeks. t
review with meta-analysis. J Orthop Sports Rev Phys Rehabil Med. 1993;5:203-217.
Phys Ther. 2011;41:734-748. http://dx.doi. 19. Hong CZ, Simons DG. Pathophysiologic and
org/10.2519/jospt.2011.3440 electrophysiologic mechanisms of myofas-
KEY POINTS 4. Chu J. Does EMG (dry needling) reduce myo- cial trigger points. Arch Phys Med Rehabil.
FINDINGS: A large immediate effect of dry fascial pain symptoms due to cervical nerve 1998;79:863-872.
root irritation? Electromyogr Clin Neurophysiol. 20. Hsieh YL, Kao MJ, Kuan TS, Chen SM, Chen
needling compared to sham or placebo
journal of orthopaedic & sports physical therapy | volume 43 | number 9 | september 2013 | 633
domised trial of trigger point acupuncture pain in diagnosis and treatment of shoulder SH, Campbell JL. Acupuncture and dry needling
compared with other acupuncture for treatment pain. A randomized control trial. BMC Mus- in the management of myofascial trigger point
of chronic neck pain. Complement Ther Med. culoskelet Disord. 2009;10:92. http://dx.doi. pain: a systematic review and meta-analysis
2007;15:172-179. http://dx.doi.org/10.1016/j. org/10.1186/1471-2474-10-92 of randomised controlled trials. Eur J Pain.
ctim.2006.05.003 33. Rosen NB. The myofascial pain syndrome. Phys 2009;13:3-10. http://dx.doi.org/10.1016/j.
24. Iversen MD, Choudhary VR, Patel SC. Thera- Med Rehabil Clin N Am. 1993;4:41-63. ejpain.2008.02.006
peutic exercise and manual therapy for persons 34. Sackett DL, Straus SE, Richardson WS. 42. Travell JG, Simons DG. Myofascial Pain and
Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.
with lumbar spinal stenosis. Int J Clin Rheuma- Evidence-Based Medicine: How to Practice and Dysfunction: The Trigger Point Manual Volume
tol. 2010;5:425-437. http://dx.doi.org/10.2217/ Teach EBM. 2nd ed. Edinburgh, UK: Churchill 2: The Lower Extremities. Media, PA: Williams &
ijr.10.29 Livingstone; 2000. Wilkins; 1992.
25. Kalichman L, Vulfsons S. Dry needling in the 35. Salaffi F, Stancati A, Silvestri CA, Ciapetti A, 43. Tsai CT, Hsieh LF, Kuan TS, Kao MJ, Chou LW,
management of musculoskeletal pain. J Am Grassi W. Minimal clinically important changes Hong CZ. Remote effects of dry needling on the
Board Fam Med. 2010;23:640-646. http:// in chronic musculoskeletal pain intensity mea- irritability of the myofascial trigger point in the
dx.doi.org/10.3122/jabfm.2010.05.090296 sured on a numerical rating scale. Eur J Pain. upper trapezius muscle. Am J Phys Med Reha-
26. Kamanli A, Kaya A, Ardicoglu O, Ozgocmen S, 2004;8:283-291. http://dx.doi.org/10.1016/j. bil. 2010;89:133-140. http://dx.doi.org/10.1097/
Zengin FO, Bayik Y. Comparison of lidocaine ejpain.2003.09.004 PHM.0b013e3181a5b1bc
injection, botulinum toxin injection, and dry 36. Simons DG, Travell JG, Simons LS. Travell & 44. Venncio RA, Alencar FGP, Jr., Zamperini C. Dif-
needling to trigger points in myofascial pain Simons Myofascial Pain and Dysfunction: The ferent substances and dry-needling injections in
Journal of Orthopaedic & Sports Physical Therapy
syndrome. Rheumatol Int. 2005;25:604-611. Trigger Point Manual Volume 1: Upper Half of patients with myofascial pain and headaches.
http://dx.doi.org/10.1007/s00296-004-0485-6 Body. 2nd ed. Philadelphia, PA: Lippincott Wil- Cranio. 2008;26:96-103.
27. Lavelle ED, Lavelle W, Smith HS. Myofas- liams & Wilkins; 1998.
cial trigger points. Med Clin North Am. 37. Skootsky SA, Jaeger B, Oye RK. Prevalence of
@ MORE INFORMATION
2007;91:229-239. http://dx.doi.org/10.1016/j. myofascial pain in general internal medicine
mcna.2006.12.004 practice. West J Med. 1989;151:157-160.
28. MacDermid JC. An introduction to evidence- 38. Spadoni GF, Stratford PW, Solomon PE, Wishart WWW.JOSPT.ORG
634 | september 2013 | volume 43 | number 9 | journal of orthopaedic & sports physical therapy