ORIGINAL ARTICLE

Periosteal Electrical Dry Needling as an Adjunct to Exercise

and Manual Therapy for Knee Osteoarthritis
A Multicenter Randomized Clinical Trial
James Dunning, DPT, MSc, FAAOMPT,*†
Raymond Butts, DPT, MSc, PhD,†‡ Ian Young, DSc, PT, Dip Osteopractic,†§
Firas Mourad, PT, OMT, Dip Osteopractic,*∥
Victoria Galante, DPT, FAAOMPT,¶ Paul Bliton, DPT, FAAOMPT,#
Michelle Tanner, DPT, FAAOMPT,**
and César Fernández-de-las-Peñas, DMSc, PT, PhD*

more patients (n = 91, 75%) within the dry needling group achieved a
Objectives: To compare the effects of adding electrical dry needling successful outcome compared with the MT and exercise group
into a manual therapy (MT) and exercise program on pain, (n = 22, 18%) at 3 months. Effect sizes were large (standardized mean
stiffness, function, and disability in individuals with painful knee differences > 0.82) for all outcome measures in favor of the electrical
osteoarthritis (OA). dry needling group at 3 months.
Materials and Methods: In total, 242 participants (n = 242) with Discussion: The inclusion of electrical dry needling into a MT and
painful knee OA were randomized to receive 6 weeks of electrical exercise program was more effective for improving pain, function,
dry needling, MT, and exercise (n = 121) or MT and exercise and related-disability than the application of MT and exercise alone
(n = 121). The primary outcome was related-disability as assessed in individuals with painful knee OA.
by the Western Ontario and McMaster Universities (WOMAC)
Osteoarthritis Index at 3 months. Level of Evidence: Level 1b—therapy. Prospectively registered
February 10, 2015 on www.clinicaltrials.gov (NCT02373631).
Results: Individuals receiving the combination of electrical dry nee-
dling, MT, and exercise experienced significantly greater improve- Key Words: knee osteoarthritis, dry needling, manual therapy,
ments in related-disability (WOMAC: F = 35.504; P < 0.001) than exercise, clinical trial
those receiving MT and exercise alone at 6 weeks and 3 months.
Patients receiving electrical dry needling were 1.7 times more likely to
(Clin J Pain 2018;34:1149–1158)
have completely stopped taking medication for their pain at 3 months
than individuals receiving MT and exercise (OR, 1.6; 95% confidence
interval, 1.24-2.01; P = 0.001). On the basis of the cutoff score of ≥5
on the global rating of change, significantly (χ2 = 14.887; P < 0.001) O steoarthritis (OA) of the knee affects up to 37% of
adults in the United States between 45 and 60 years of
age.1 A recent meta-analysis found that the crude prevalence
Received for publication March 30, 2018; revised May 21, of knee OA was 25% in patients aged above 20 years and 39%
2018; accepted May 22, 2018. in people aged above 30 years.2 In addition, hip and knee OA
From the *Department of Physical Therapy, Occupational Therapy, are ranked as the 11th highest contributors to global disability
Rehabilitation and Physical Medicine, University King Juan Carlos,
Alcorcón, Spain; †American Academy of Manipulative Therapy in patients with chronic pain.3 Physiological changes in OA
Fellowship in Orthopaedic Manual Physical Therapy, Montgomery, are characterized by degeneration of articular cartilage with
AL; ‡Research Physical Therapy Specialists, Columbia, SC; osteophyte formation, microfractures, subchondral sclerosis
§CORA Physical Therapy, Savannah, GA; **Benchmark Physical and plate thickening, and exposure of the articular end of the
Therapy, Atlanta, GA; ¶STI Physical Therapy and Rehabilitation,
Phoenix, AZ; #Orthopedic Rehab Specialists, Rockford, IL; and bone.4–6 The clinical manifestations of knee OA are joint pain,
∥Department of Physiotherapy, University of Rome Tor Vergata, stiffness in the morning or after rest, limited joint motion, night
Italy. pain, and/or joint deformity. The clinical diagnosis of knee OA
J.D., R.B., and C.F.-d.-l.-P.: participated in the conception, design, data is typically made using the American College of Rheumatol-
acquisition, statistical analyses, data interpretation, drafting, and
revision of the manuscript. I.Y.: was involved in the statistical ogy clinical criteria developed by Altman, which has been
analysis, data interpretation, drafting, and revision of the manu- found to be 89% sensitive and 88% specific.7,8 The patho-
script. F.M.: was involved in the revision of the manuscript. V.G., genesis and temporal relationship of anatomic lesions is largely
P.B., and M.T.: were involved in data collection and revision of the unknown, and there are currently no curative treatments for
manuscript.
The authors declare no conflict of interest. OA. Long-term use of oral nonsteroidal anti-inflammatory
Reprints: James Dunning, DPT, MSc, FAAOMPT, 1036 Old Breck- drugs has been discouraged, and many patients with chronic
enridge Ln Montgomery, AL 36117 (e-mail: j.dunning.2016@alumnos. pain seek for nonpharmacological management options.9,10
urjc.es). Exercise11–16 and acupuncture17–23 are 2 nonpharmacological
Copyright © 2018 The Author(s). Published by Wolters Kluwer Health,
Inc. This is an open-access article distributed under the terms of the interventions recommended for individuals with knee OA in
Creative Commons Attribution-Non Commercial-No Derivatives recent meta-analyses and international clinical guidelines.18,24
License 4.0 (CCBY-NC-ND), where it is permissible to download A Cochrane review found a statistically significant
and share the work provided it is properly cited. The work cannot be benefit, with moderate effect sizes for pain (standardized
changed in any way or used commercially without permission from
the journal. mean differences [SMD], 0.49) and physical function (SMD,
DOI: 10.1097/AJP.0000000000000634 0.52) immediately after treatment, and small effect sizes

Clin J Pain  Volume 34, Number 12, December 2018 www.clinicalpain.com | 1149
Dunning et al Clin J Pain  Volume 34, Number 12, December 2018

(pain, SMD, 0.24; physical function, SMD, 0.15) at 2 to Electrical dry needling and the combination of MT and
6 months follow-up for various forms of exercise in individ- exercise, when applied separately, have been found to be
uals with moderate knee OA.11 Another recent systematic moderately effective for knee OA. Although 3 previous
review reported that exercise plus manual therapy (MT) for studies46–48 investigated the combined effects of acupuncture
joint mobilization showed a moderate effect size (SMD, 0.69) and exercise in patients with knee OA, they used manual
that was significantly higher than the effect sizes observed for acupuncture rather than electroacupuncture. No previous
exercise alone (SMD, 0.34).12 study has investigated the combination of the effectiveness of
Pain may be a potential barrier leading to underdosage of electrical dry needling in addition to MT and exercise in
strength training and aerobic exercise stimulus in individuals patients with knee OA. Therefore, the purpose of this multi-
with painful knee OA; therefore, needling therapies may be a center randomized clinical trial was to compare the effects of
reasonable nonpharmacologic adjunct intervention for the adding electrical dry needling, into a MT and exercise program
reduction of chronic pain in individuals participating in exer- on pain, stiffness, function, and disability in individuals with
cise programs for knee OA.17,18,20,21 Needling therapy refers to painful knee OA. We hypothesized that individuals receiving
the insertion of thin monofilament needles, as used in the electrical dry needling combined with MT and exercise would
practice of acupuncture, without the use of injectate.25–29 Dry exhibit greater improvements in pain, stiffness, function, and
needling is typically used to stimulate muscles, ligaments, disability than those receiving only MT and exercise.
tendons, subcutaneous fascia, scar tissue, or peripheral nerves
for the management of pain and disability associated with MATERIALS AND METHODS
neuromusculoskeletal disorders.25,28–30 Interestingly, the most
common term used to describe dry needling is “acupuncture,” Study Design
that is, “acu” literally translates to needle and “puncture” to This randomized, single-blinded, multicenter, parallel-
penetration.29 group trial compared 2 treatment protocols for the management
The terminology, theoretical constructs, and philosophies of knee OA: MT and exercise versus MT and exercise plus
may differ; however, dry needling and acupuncture overlap in electrical dry needling. The primary outcome was related-
terms of needling technique with the use of thin monofilament disability as assessed by the Western Ontario and McMaster
needles.31 Notably, several previous meta-analyses and literature Universities (WOMAC total score) Osteoarthritis Index at
reviews have chosen to consider “acupuncture and dry nee- 3 months. Secondary outcomes included knee pain intensity as
dling” as one category of interventions.32–36 Therefore, from a measured by the Numeric Pain Rating Scale (NPRS), all
procedural and technical perspective, and for the purpose of WOMAC subscales (pain: WOMAC-P; stiffness: WOMAC-S;
evaluating and comparing efficacy and effect sizes within the physical function: WOMAC-PF), medication intake, and the
broader literature on the use of needling without injectate in global rating of change (GROC). The current clinical trial was
patients with knee OA published by acupuncturists, western conducted following the Consolidated Standards of Reporting
medical physicians, and physical therapists alike, “electro- Trials (CONSORT) extension for pragmatic clinical trials.49
acupuncture” and “electrical dry needling” will be considered The study was approved by the ethics committee at Universidad
interchangeable terms, and in this context do not rely on diag- Rey Juan Carlos, Madrid, Spain (URJC-DPTO 31-2014) and
noses from oriental medicine (eg, bi syndrome, blood stagna- the trial was prospectively registered (ClinicalTrials.gov:
tion, or kidney yang deficiency37,38) or theoretical movement of NCT02373631).
qi along traditional Chinese acupuncture meridians.39,40
Importantly, none of the knee OA studies cited herein used Participants
injectate in conjunction with their needling procedure; therefore, Consecutive individuals with painful knee OA from 18
all studies fit within the strict definition of dry needling, acu- outpatient physical therapy clinics in 10 different states
puncture, or “noninjection needling” (as opposed to “injection (Arizona, Florida, Georgia, Illinois, New Hampshire,
needling” or “wet needling”), regardless of the differing termi- New York, North Carolina, Rhode Island, South Carolina,
nologies, theoretical constructs, or philosophies.25,29,31 Virginia) were screened for eligibility criteria and recruited
The current body of evidence seems to support the use of over a 24-month period (from February 2015 to 2017).
dry needling therapies without injectate, that is, acupuncture For patients to be eligible, they had to have met the
for treating the pain, stiffness, and related-disability associated American College of Rheumatology criteria for the diag-
with knee OA.17,21,23,29,41–43 Zhang et al44 cited a 69% con- nosis of knee OA7,8 and have had chronic pain in the knee
sensus following a Delphi study recommending the use of joint for > 3 months. Patients had to have at least 3 of the
acupuncture for the symptomatic treatment of OA and following criteria7,8,49 to be included in the study: (1) above
reported a moderate effect size for this needling modality (ie, 50 years of age; (2) <30 minutes of morning stiffness; (3)
acupuncture). The OARSI guidelines20 for hip and knee OA crepitus on active motion; (4) bony tenderness; (5) bony
reported acupuncture to have a moderate effect size for pain enlargement; and (6), no palpable warmth of synovium.7 In
(0.51), stiffness (0.41), and function (0.51). In addition, based addition, participants had to have a minimum knee pain
on the individual effect sizes of 11 trials reported by intensity score of 2 points and be older than 18 years of age.
Manheimer et al,45 Zhang et al44 concluded that acupuncture Patients were excluded if they exhibited: (1) a history of
was superior to usual care and wait list controls with a surgery to the painful knee; (2) a history of surgery to either of
pooled effect size of 0.58 for pain relief. Although it is not the lower extremities in the last 6 months; (3) any red flags to
always appropriate to compare effect sizes among various MT, dry needling, or exercise; (4) had received physical therapy,
treatments,20 to our knowledge, a pooled standard effect size acupuncture, massage therapy, chiropractic, or intra-articular
for pain relief of 0.58 for acupuncture in patients with knee OA injections for the painful knee in the last 3 months; (5) presented
is higher than most other conservative treatments applied to with ≥ 2 positive neurological signs; or (6) had involvement in
this pain population, including nonsteroidal anti-inflammatory litigation or worker’s compensation regarding their knee pain.
drugs (0.32), muscle strengthening exercises (0.32), and aerobic Patients were also excluded if they were pregnant. All partic-
exercises (0.52).20,44 ipants signed an informed consent before their participation in

1150 | www.clinicalpain.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Clin J Pain  Volume 34, Number 12, December 2018 Electrical Dry Needling for Knee Osteoarthritis

the study. All participants were naïve to the use of dry needling
procedures and had not previously experienced needling without
injectate for their knee pain.

Treating Therapists
In total, 18 physical therapists (mean age, 38.4 y; SD,
10.44) participated in the delivery of treatment for patients
in this study. They had an average of 12.5 (SD, 9.54) years
of clinical experience, an average of 4.3 (SD, 1.88) years
using dry needling, and all had completed a 54-hour post-
graduate certification program that included practical
training in electrical dry needling for knee OA. All partic-
ipating physical therapists were required to study a manual
of standard operating procedures and participate in a 6-hour
training session with the principal investigator.

Randomization and Blinding

Following the baseline examination, patients were ran-
domly assigned to receive MT and exercise alone or in combi-
nation with electrical dry needling. Concealed allocation was
conducted using a computer-generated randomized table of
numbers created by a statistician who was not otherwise
involved in the trial and did not participate in analysis or
interpretation of the results. Individual and sequentially num-
bered index cards with the random assignment were prepared
for each of the 18 data collection sites. The index cards were
folded and placed in sealed opaque envelopes. Blinded to the
baseline examination, the treating therapist opened the envelope
and proceeded with treatment according to the group assign-
ment. The examining therapist remained blind to the patient’s
treatment group assignment at all times; however, based on the
nature of the interventions it was not possible to blind patients
or treating therapists. FIGURE 1. Standardized 9-point protocol of periosteal electrical
dry needling for knee osteoarthritis.
Interventions
All participants received between 8 and 10 treatment exercise program during all days that they did not receive
sessions at a frequency of 1 to 2 times per week over a supervised physical therapy in the clinic. Patients were asked
6-week period. Both groups received MT (passive joint to monitor their compliance with the home exercise program
mobilizations and muscle stretching) and exercise (riding a by maintaining a home exercise program logbook.
stationary bicycle, range of motion, and strengthening In addition to MT and exercise, patients allocated to
exercises to the lower extremity) on each session. In addi- the dry needling group also received 8 to 10 sessions of
tion, the dry needling group also received electrical dry periosteal electrical dry needling at a frequency of 1 to 2
needling using a standardized 9-point protocol for 20 to times per week over 6 weeks. Electric dry needling included
30 minutes on each treatment session. a 9-point standardized protocol as depicted in Figure 1.
Although specific recommendations cannot be made Each needle insertion site and anatomic target is summar-
regarding the type of exercise12 or the optimal exercise dosage in ized within Appendix 1. In addition to the obligatory
patients with knee OA,11 patients received the following inter- 9-point standardized protocol, clinicians were also permitted
ventions at all treatment sessions: 30 minutes of lower extremity to insert needles at up to 4 additional locations based on the
strengthening (weight bearing, non–weight-bearing, concentric, presence of the symptoms.
eccentric), range of motion (riding a stationary bicycle), Sterilized disposable stainless steel acupuncture needles were
stretching exercises (static muscle stretching), and passive used with 3 sizes: 0.25 mm×30 mm, 0.30 mm×40 mm, and
accessory and physiological joint mobilizations.50 The exercise 0.30 mm×50 mm. The depth of needle insertion ranged from 15
program was taught to the patient by an experienced physical to 45 mm and depended on the point selected (intramuscular,
therapist on the first session and supervised on subsequent ses- periosteal, joint line, intra/periarticular) and the patient’s physical
sions. Strengthening, range of motion, and stretching exercises constitution. Following topical skin cleansing with sterile alcohol
were gradually progressed according to tolerance of each indi- prep pads, all needles were inserted and then manipulated
vidual patient. That is, progression only occurred if patients bidirectionally to illicit a sensation of aching, tingling, deep
reported a decrease in symptoms and in the absence of excessive pressure, heaviness, or warmth.51,52 In addition, at least 3 of the 9
soreness. Details regarding the exercise and MT program have obligatory needles (ie, over the posteromedial aspect of the
previously been described by Deyle et al.50 medial tibial condyle, within the depression posterior to the
All patients in both groups were asked to complete a femoral epicondyle, and over the anterolateral crest of the tibia 1
daily home exercise program.50 The home exercise program fingerbreadth lateral to the tibial tuberosity) were repeatedly
consisted of the same strengthening, range of motion, and thrusted and tapped on to the respective bone using a “periosteal
stretching exercises that were prescribed and supervised in stimulation” technique.53 Notably, with the exception of the 2
the clinic.50 Patients were asked to complete the home obligatory needles inserted at the level of the tibiofemoral joint

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved. www.clinicalpain.com | 1151
Dunning et al Clin J Pain  Volume 34, Number 12, December 2018

margin within the medial or lateral infrapatellar sulcus, and Treatment Side Effects
depending on the patient’s physical constitution, the needle length Patients were asked to report adverse events that they
selected by the practitioner and the patient’s tolerance to such, the experienced during any part of the study. In the current study,
remaining obligatory needles were also advanced toward the an adverse event was defined as a sequelae of 1-week duration
underlying bone to facilitate direct mechanical and electrical with any symptom perceived as distressing and unacceptable to
“periosteal stimulation.”53 The needles were then left in situ for the patient that required further treatment.70 Particular atten-
20 to 30 minutes41–43,54–56 with electric stimulation (ES-160 tion was given to the presence of ecchymosis and postneedling
electrostimulator ITO co.) in pairs (crossing through the knee soreness within the group receiving electrical dry needling.
joint in a superior-inferior and diagonal orientation) using 4
channels to 8 of the needles using a low frequency (2 Hz), Sample Size Determination
moderate pulse duration (250 μs), biphasic continuous waveform The sample size calculations were based on detecting a
at a maximum tolerable intensity.55,56 In cases of bilateral knee between-groups moderate effect size of 0.4 at 3 months,
OA, both knees were treated, but only the most painful side at assuming a 2-tailed test, an alpha level (α) of 0.05 and a
baseline was recorded and analyzed throughout the study to desired power (β) of 90%. The estimated desired sample size
satisfy the assumption of independent data.57 was calculated to be at least 105 patients per group. A
dropout percentage of 15% was expected, so 120 patients
were included on each group.
Outcome Measures
Participants received a standardized physical exami- Statistical Analysis
nation during which the affected knees were examined for Statistical analysis was performed using SPSS software,
conditions other than OA; that is, referred pain from the hip version 24.0 (Chicago, IL) and it was conducted according to
joint or lumbopelvic region were ruled out. The physical intention-to-treat analysis. We performed Little’s Missing
examination included, but was not limited to, measurements Completely at Random (MCAR) test71 to determine whether
of passive and active knee range of motion. missing data points associated with dropouts were missing at
The primary outcome was related-disability as assessed random or missing for systematic reasons. Intention-to-treat
with the WOMAC total index score, whereas each WOMAC analysis was performed by using expectation-maximization
subscale (WOMAC-P, WOMAC-S, and WOMAC-PF) were whereby missing data were computed using regression equations.
considered as secondary outcomes. The WOMAC is a valid The effects of treatment on pain, stiffness, physical func-
and reliable instrument and has been used extensively to tion, and related-disability were each examined with a 2-by-4
evaluate 3 dimensions (pain, stiffness, and physical function) mixed-model analysis of covariance (ANCOVA) with treatment
in patients with hip or knee OA.58–60 In patients with OA of group as the between-subjects factor, time as the within-subjects
the lower extremities participating in rehabilitation programs,
the minimum clinically important difference (MCID) for the TABLE 1. Baseline Characteristics by Treatment Assignment
WOMAC has been calculated to range from 9% to 12% of
the baseline score.61–63 However, in our study, we used 36% n = 242
change in the WOMAC (ie, triple the value of the 12%
Manual Therapy
MCID) to represent a successful outcome.
Manual +Exercise+
Secondary outcomes included knee pain intensity, the 3
Therapy Electrical Dry
WOMAC subscales, medication intake and the GROC. A
Baseline Variables +Exercise Needling
NPRS measured knee pain intensity. Patients were asked to
indicate the average intensity of knee pain over the past Sex (male/female) 55/56 56/55
week using an 11-point scale ranging from 0 (no pain) to 10 Age (y) 58.1 ± 13.1 57.1 ± 13.2
(worst pain imaginable) at baseline, 2 weeks, 6 weeks, and Weight (kg) 83.8 ± 16.6 83.4 ± 15.6
Height (cm) 172.0 ± 8.9 172.1 ± 8.6
3 months following the initial treatment session.64 The
Years with knee pain 4.6 ± 5.1 4.5 ± 4.7
NPRS is a reliable and valid instrument to assess pain Medication intake (n [%])
intensity.65–67 The MCID for the NPRS has been shown to Not at all 39 (32) 36 (30)
be 1.74 in patients with chronic pain conditions67; however, Once a week 13 (11) 13 (11)
the MCID for knee-related pain has not yet been estab- Once every couple of 29 (24) 28 (23)
lished. Nevertheless, a change of 2 points or a 30% decrease days
in pain from baseline can be considered as a MCID in Once or twice a day 37 (31) 40 (33)
patients with chronic musculoskeletal pain.67,68 ≥ 3 times a day 3 (2) 4 (3)
Medication intake was measured as the number of No. treatment sessions 8.9 ± 1.9 8.7 ± 1.8
Mean intensity of knee 5.4 ± 1.8 5.7 ± 1.6
times the patient had taken prescription or over-the-coun-
pain (NPRS, 0-10)
ter analgesic or anti-inflammatory medication in the past WOMAC Pain Scale 8.0 ± 3.3 8.7 ± 3.2
week for their knee pain, with 5 options: (1) not at all, (2) (0-20)
once a week, (3) once every couple of days, (4) once or WOMAC Stiffness 3.8 ± 1.4 4.0 ± 1.6
twice a day, or (5) ≥ 3 times a day. Medication intake was Scale (0-8)
assessed at baseline and at 3 months after the first treat- WOMAC Physical 28.1 ± 11.1 28.9 ± 10.6
ment session. Function Scale
At 2 weeks, 6 weeks, and 3 months following the initial (0-68)
treatment session, patients completed a 15-point GROC ques- WOMAC total score 39.9 ± 14.6 41.6 ± 14.3
(0-96)
tion based on a scale described by Jaeschke et al69 to rate their
self-perceived improved function. The MCID for the GROC NPRS indicates Numeric Pain Rating Scale; WOMAC, Western Ontario
has not been specifically reported but scores of +4 and +5 have and McMaster Universities.
typically been indicative of moderate changes in patient status.69

1152 | www.clinicalpain.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Clin J Pain  Volume 34, Number 12, December 2018 Electrical Dry Needling for Knee Osteoarthritis

FIGURE 2. Flow diagram of patient recruitment and retention. FU indicates follow-up; OA, osteoarthritis.

factor, and adjusted for baseline data. Separate ANCOVAs difference (P = 0.468) between the mean number of completed
were performed with each outcome as the dependent variable. treatment sessions for the MT, exercise plus electrical dry
For each ANCOVA, the main hypothesis of interest was the needling group (mean, 8.7 ± 1.8) and the MT and exercise group
2-way interaction (group by time) with a Bonferroni-corrected (mean, 8.9 ± 1.9). In total, 235 of the 242 patients completed all
α level of 0.0125 (4 timepoints). We used χ2 tests to compare outcome measures through 3 months (97% follow-up). Of the
self-perceived improvement with GROC and changes in medi- 7 patients that dropped out or failed to complete outcome
cation intake. To enable comparison of between-group effect
sizes, SMDs were calculated by dividing mean score differences
between groups by the pooled SD. Numbers needed to treat
(NNT) and 95% confidence intervals (CI) were also calculated
at the 3-month follow-up period using each definition for a
successful outcome.

RESULTS
Between February 2015 and 2017, 431 consecutive patients
with knee pain were screened for possible eligibility criteria. In
total, 242 (56.15%) satisfied all the inclusion criteria, agreed to
participate, and were randomly allocated into the MT and
exercise (n = 121) or MT and exercise plus electrical dry needling
(n = 121) group. Randomization resulted in similar baseline FIGURE 3. Evolution of the WOMAC Osteoarthritis Index
characteristics for all variables (Table 1). The reasons for ineli- throughout the course of the study stratified by randomized
gibility are found in Figure 2, which provides a flow diagram of treatment assignment. Data are means (SE). DN indicates dry
patient recruitment and retention. There was no significant needling; WOMAC, Western Ontario and McMaster Universities.

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved. www.clinicalpain.com | 1153
Dunning et al Clin J Pain  Volume 34, Number 12, December 2018

measures, 3 were from the electrical dry needling group and 4 (0.82 < SMD < 0.94) at 3 months after the first treatment ses-
were from the MT and exercise group. sion in favor of the dry needling group (Table 3). Within-
In total, 87 patients assigned to the MT and exercise group percentage change from baseline to 3 months for the
plus electrical dry needling group (71.9%) experienced primary outcome (WOMAC) was 67.0% and 32.9% for the
postneedling muscle soreness and 57 (47.1%) experienced electrical dry needling group and nondry needling group,
mild bruising (ecchymosis) that most commonly resolved respectively.
spontaneously within 48 hours and 2 to 4 days, respectively. The intention-to-treat analysis also revealed a sig-
In addition, 6 patients (4.9%) in the electrical dry needling nificant group×time interaction for knee pain (NPRS)
group experienced drowsiness, headache, or nausea, which intensity (F = 29.094; P < 0.001): individuals receiving elec-
spontaneously resolved within several hours. No other trical dry needling experienced significantly greater decrease
adverse events were reported. in knee pain at 6 weeks (Δ, −1.2; 95% CI, −1.7 to −0.7;
Adjusting for baseline outcomes, the mixed-model P < 0.001) and 3 months (Δ, −2.7; 95% CI, −3.4 to −2.0;
ANCOVA revealed a significant group×time interaction for P < 0.001) than those receiving MT and exercise alone
the primary outcome (WOMAC: F = 35.504; P < 0.001): (Fig. 4). For knee pain intensity (NPRS), between-groups
patients receiving electrical dry needling experienced sig- effect sizes were moderate (SMD, 0.60) at 6 weeks and large
nificantly greater improvements in related-disability at 6 (SMD, 0.96) at 3 months in favor of the dry needling group
weeks (Δ, −10.4; 95% CI, −13.7 to −7.1; P < 0.001) and (Table 3). Within-group percentage change from baseline to
3 months (Δ, −13.9; 95% CI, −17.4 to −10.4; P < 0.001) 3 months for knee pain intensity (NPRS) was 67.2% and
than those receiving MT and exercise alone (Fig. 3). Sim- 28.9% for the electrical dry needling group and nondry
ilarly, significant group×time interactions were also found needling group, respectively.
for all WOMAC subscales (WOMAC-P: F = 30.131, P < 0.001; Patients receiving electrical dry needling were 1.7 times
WOMAC-S: F = 29.665, P < 0.001; WOMAC-PF: F = 30.114, more likely to have completely stopped taking medication
P < 0.001) in favor of the dry needling group (Table 2). For for their pain at 3 months than individuals receiving MT
the WOMAC and all subscales, between-groups effect sizes and exercise alone (OR, 1.6; 95% CI, 1.24-2.01; P = 0.001).
were moderate (0.53 < SMD < 0.76) at 6 weeks and large On the basis of the cutoff score of ≥ +5 on the GROC,

TABLE 2. WOMAC Osteoarthritis Index at Baseline, 2 Weeks, 6 Weeks, and 3 Months After the First Treatment Sessions as Well as Within-
group and Between-groups Mean Scores by Randomized Treatment Assignment
Timeline Scores: Within-group Change Scores (n = 242)
(Mean ± SD [95% CI])
Between-group Differences
Outcomes MT+EX MT+EX+EDN (Mean [95% CI])
WOMAC-P: pain (0-20)
Baseline 8.0 ± 3.3 (7.4-8.6) 8.7 ± 3.2 (8.1-9.3)
2 wk 6.1 ± 3.0 (5.6-6.6) 5.4 ± 3.2 (4.8-6.0)
Change baseline → 2 wk −1.9 ± 2.5 (−1.5 to −2.3) −3.3 ± 2.6 (−2.8 to −3.8) −1.4 (−2.1 to −0.7)
6 wk 4.8 ± 2.8 (4.3-5.3) 3.4 ± 2.6 (2.9-3.9)
Change baseline → 6 wk −3.2 ± 3.1 (−3.8 to −2.6) −5.3 ± 3.0 (−5.9 to −4.7) −2.1 (−2.9 to −1.3)
3 mo 5.2 ± 3.2 (4.7-5.7) 2.8 ± 2.5 (2.3-3.3)
Change baseline → 3 mo −2.8 ± 3.2 (−3.4 to −2.2) −5.9 ± 3.3 (−6.5 to −5.3) −3.1 (−3.9 to −2.3)
WOMAC-S: stiffness (0-8)
Baseline 3.8 ± 1.4 (3.6-4.0) 4.0 ± 1.6 (3.7-4.3)
2 wk 3.0 ± 1.5 (2.7-3.3) 2.5 ± 1.4 (2.2-2.8)
Change baseline → 2 wk −0.8 ± 1.4 (−1.1 to −0.5) −1.5 ± 1.3 (−1.8 to −1.4) −0.7 (−1.0 to −0.4)
6 wk 2.4 ± 1.5 (2.1-2.7) 1.7 ± 1.4 (1.5-1.9)
Change baseline → 6 wk −1.4 ± 1.6 (−1.7 to −1.1) −2.3 ± 1.5 (−2.6 to −2.0) −0.7 (−1.0 to −0.4)
3 mo 2.4 ± 1.5 (2.2-2.6) 1.3 ± 1.3 (1.1-1.5)
Change baseline → 3 mo −1.4 ± 1.6 (−1.8 to −1.2) −2.7 ± 1.5 (−3.0 to −2.4) −1.3 (−1.6 to −0.9)
WOMAC-PF: physical function (0-68)
Baseline 28.1 ± 11.1 (26.1-30.1) 28.9 ± 10.6 (27.0-30.8)
2 wk 22.3 ± 11.6 (20.3-24.3) 17.1 ± 10.6 (15.1-19.1)
Change baseline → 2 wk −5.8 ± 8.7 (−7.0 to −4.6) −11.8 ± 9.6 (−13.6 to −10.0) −6.0 (−8.4 to −3.6)
6 wk 18.7 ± 10.9 (16.8-20.6) 12.1 ± 9.8 (10.2-14.0)
Change baseline → 6 wk −9.4 ± 9.0 (−11.0 to −7.8) −16.8 ± 10.2 (−18.7 to −14.9) −7.4 (−9.9 to −4.9)
3 mo 18.7 ± 11.7 (16.8-20.6) 10.1 ± 9.3 (8.2-12.0)
Change baseline → 3 mo −9.4 ± 9.8 (−11.1 to −7.7) −18.8 ± 10.6 (−20.7 to −16.9) −9.4 (−12.0 to −6.8)
WOMAC: Total Index (0-96)
Baseline 39.9 ± 14.6 (37.4-42.4) 41.6 ± 14.3 (39.0-44.2)
2 wk 31.4 ± 15.1 (28.8-34.0) 25.0 ± 14.3 (22.3-27.7)
Change baseline → 2 wk −8.5 ± 11.0 (−10.5 to −6.5) −16.6 ± 12.3 (−18.9 to −14.3) −8.1 (−11.1 to −5.1)
6 wk 25.9 ± 14.3 (23.5-28.3) 17.2 ± 13.1 (14.7-19.7)
Change baseline → 6 wk −14.0 ± 12.4 (−16.2 to −11.8) −24.4 ± 13.4 (−26.9 to −21.9) −10.4 (−13.7 to −7.1)
3 mo 26.4 ± 15.6 (23.9-28.9) 14.2 ± 12.5 (11.7-16.7)
Change baseline → 3 mo −13.5 ± 13.3 (−15.9 to −11.1) −27.4 ± 14.1 (−29.9 to −24.9) −13.9 (−17.4 to −10.4)
CI indicates confidence interval; EDN, electrical dry needling; EX, exercise; MT, manual therapy; WOMAC, Western Ontario and McMaster Universities.

1154 | www.clinicalpain.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Clin J Pain  Volume 34, Number 12, December 2018 Electrical Dry Needling for Knee Osteoarthritis

TABLE 3. Between-group Effect Sizes (Standardized Mean Difference) in Favor of the Dry Needling Group When Compared With the
Combination Manual Therapy and Exercise
Outcome WOMAC Total WOMAC Pain WOMAC Stiffness WOMAC Function NPRS Pain
6 wk 0.76 0.67 0.53 0.74 0.60
3 mo 0.94 0.90 0.82 0.87 0.96
Large between-group effect size: Cohen’s d ≥ 0.8. Medium effect size: Cohen’s d = ≥ 0.5. Small size: Cohen’s d ≥ 0.2. Effect size provides information about
the magnitude or strength of the difference between the 2 groups.
NPRS indicates Numeric Pain Rating Scale; WOMAC, Western Ontario and McMaster Universities.

significantly (χ2 = 14.887; P < 0.001) more patients (n = 91, Three previous studies found nonsuperior results when
75%) in the dry needling group achieved a successful out- adding acupuncture as an adjunct therapy to exercise-based
come compared with the nondry needling group (n = 23, physical therapy in knee OA.46–48 Notably, Foster et al47
19%) at 3 months (Table 4). Therefore, based on the cutoff reported no statistically significant between-groups difference
score of ≥ +5 on the GROC, the NNT was 1.78 (95% CI, in WOMAC pain subscale scores after adding a course of
1.50-2.18) in favor of the electrical dry needling group at acupuncture to exercise in knee OA. Nevertheless, in the
3-month follow-up. Likewise, based on the cutoff score of Foster et al47 trial, the acupuncture points were not
36% improvement (ie, triple the MCID) on the WOMAC, standardized but selected based on the “clinical opinion” of 67
the NNT was 2.37 (95% CI, 1.89-3.19) in favor of the different physiotherapists at different centers. Considering the
electrical dry needling group at 3-month follow-up. recent findings regarding the influence of acupuncture on
cartilage repair72 and the efficacy of periosteal stimulation53 in
knee OA, it is also possible that the needles in these previous
DISCUSSION studies (0.2 to 3.5 cm) were not inserted deep enough.46–48 In
To our knowledge, this study is the first randomized addition, a recent meta-analysis73 and a separate secondary
clinical trial comparing the effectiveness of MT and exercise analysis that pooled data from the Cochrane review19,74 con-
plus electrical dry needling to MT and exercise alone in cluded that electroacupuncture is superior to manual acu-
patients with painful knee OA. The results suggest that a puncture in knee OA; however, neither the Foster et al47 nor
mean of 9 sessions of MT and exercise plus electrical dry Chen et al46 trials used electrical stimulation with the needles.
needling, using a 9-point standardized protocol targeting the
knee locally at a frequency of 1 to 2 times per week over Mechanisms of Periosteal Electrical Dry Needling
6 weeks, resulted in greater improvements in pain, stiffness, The underlying mechanisms as to why the electrical dry
function, related-disability, and medication intake than MT needling group in the current study experienced greater
and exercise alone. For the primary outcome of related- improvements than the MT and exercise group remains to
disability (WOMAC), between-groups effect sizes were be elucidated. However, appropriate needle depth may be
moderate at 6 weeks and large at 3 months in favor of the an important component to consider when using dry nee-
dry needling group. The between-groups difference for dling therapies for joint OA. A number of studies have
change in related-disability, as measured by the WOMAC shown that periosteal needling, that is, getting the needle
(34.1%; 95% CI, 26.6-41.4) exceeded the reported MCID (ie, close to the bone, cartilage or joint line, or tapping the
12%61) at 3 months. In addition, for knee pain intensity, the
point estimate for the between-groups change (3.23 points; TABLE 4. Self-perceived Improvement With GROC in
95% CI, 2.4-4.0) also exceeded the reported MCID (ie, 1.74 Both Groups
points67,68) at 3 months. Finally, the NNT suggests for every
2 patients treated with electrical dry needling, rather than N = 242
MT and exercise alone, 1 additional patient with knee OA Manual Manual Therapy+
achieves clinically important reductions in related-disability GROC Therapy Exercise+Electrical Dry
at 3-month follow-up. (−7 to +7) +Exercise Needling
2 wk after first treatment session
Moderate changes 18 (14.9) 49 (40.5)
(+4/+5)
Large changes 2 (1.7) 11 (9.1)
(+6/+7)
6 wk after first treatment session
Moderate changes 39 (32.2) 59 (48.8)
(+4/+5)
Large changes 8 (6.6) 36 (29.8)
(+6/+7)
3 mo after first treatment session
Moderate changes 27 (22.3) 45 (37.2)
(+4/+5)
Large changes 10 (8.3) 57 (47.1)
(+6/+7)
FIGURE 4. Evolution of knee pain intensity (NPRS, 0 to 10)
throughout the course of the study stratified by randomized Values are represented as n (%).
treatment assignment. Data are means (SE). NPRS indicates GROC indicates global rating of change.
Numeric Pain Rating Scale.

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved. www.clinicalpain.com | 1155
Dunning et al Clin J Pain  Volume 34, Number 12, December 2018

needle repeatedly on to the bone, leads to significant and 10 mm to 45 mm and depended on the point (intramuscular,
clinically meaningful improvements in pain and disability in periosteal, joint line, intra/periarticular) and the patient’s con-
hip and knee OA.53,75,76 Zhang et al72 recently reported stitution (ie, size and bone depth, muscle and/or connective
significantly lower T2 values on magnetic resonance imaging tissue thickness). The following 9 needles were inserted:
at the anteromedial and anterolateral tibial subregions of 100
knees following 20 minute sessions over 4 weeks of 7-point, (1) Superolateral and anterior insertion within the popliteus,
low-frequency electroacupuncture; that is, electroacupuncture with periosteal stimulation over the posteromedial
seems to play a role in cartilage repair in individuals with aspect of the medial tibial condyle.
knee OA.72 Moreover, acupuncture has been shown to reduce (2) Inferolateral insertion angle within the distal adductor
interleukin-6 mRNA expression in bone marrow, thereby magnus, with periosteal stimulation within the depres-
limiting inflammation and inhibiting myelogenic osteoclast sion posterosuperior to the femoral epicondyle.
activity driving degeneration.77 (3) Perpendicular insertion within the tibialis anterior, with
Electroacupuncture to local points at the knee has been periosteal stimulation over the anterolateral crest of the
found to modulate knee joint microcirculation, significantly tibia one fingerbreadth lateral to the tibial tuberosity.
increase endogenous opioid levels, and significantly reduce (4) Perpendicular insertion within the quadriceps tendon, one
plasma cortisol levels.78,79 In addition, electroacupuncture has fingerbreadth proximal to the superior border of the patella.
been found to block the local release of inflammatory cytokines (5) Perpendicular insertion within the vastus lateralis, 3
(ie, interleukin-1 β and tumor necrosis factor-α) in the synovia fingerbreadths proximal to the superolateral border of
of osteoarthritic joints80 and the systemic release of inflamma- the patella.
tory factors in the periaqueductal gray of the brain stem.81 (6) Perpendicular insertion within the vastus medialis, 3
Acupuncture may also stimulate an increase in hyaluronic acid, fingerbreadths proximal to the superomedial border of
allowing the synovial fluid to better lubricate the joint.82 the patella.
(7) Perpendicular insertion at the level of the tibiofemoral
Strengths and Limitations joint margin within the medial infrapatellar sulcus.
Major strengths of the current study include the inclusion (8) Perpendicular insertion at the level of the tibiofemoral
of a large sample size with 18 treating physical therapists from joint margin within the lateral infrapatellar sulcus.
18 clinics in 10 different geographical states, and the use of the (9) Perpendicular insertion within the extensor digitorum
same standardized 9-point needling protocol and dosage longus, one thumb width distal and anterior to the fibula
parameters. However, we only assessed mid-term follow-up; head. Unlike the other 8 needles that were electrically
thus, we do not know if the significant between-groups dif- connected in pairs, and for the purpose of standardization,
ferences observed at 3 months would be sustained in the long the ninth needle was not paired with 1 of the 4 electrical
term. We also cannot be certain that the results are general- channels; nevertheless, it was manually manipulated and
izable to other dry needling protocols, dosages, techniques, or left in situ for the duration of the treatment (Fig. 1).
needle placements. In addition, we did not include a dry nee-
dling placebo group; which should be included in future
studies. Finally, therapist and patient treatment preferences
were not collected and could potentially affect the results. REFERENCES
1. Zhang Y, Jordan JM. Epidemiology of osteoarthritis. Clin
Geriatr Med. 2010;26:355–369.
CONCLUSIONS
2. Kobayashi S, Pappas E, Fransen M, et al. The prevalence of
The results of the current randomized clinical trial patellofemoral osteoarthritis: a systematic review and meta-
demonstrated that patients with painful knee OA who analysis. Osteoarthritis Cartilage. 2016;24:1697–1707.
received MT and exercise plus electrical dry needling expe- 3. Cross M, Smith E, Hoy D, et al. The global burden of hip and
rienced significantly greater improvements in pain intensity, knee osteoarthritis: estimates from the global burden of disease
stiffness, physical function, related-disability, and medi- 2010 study. Ann Rheum Dis. 2014;73:1323–1330.
cation intake as compared with the group that received MT 4. Martel-Pelletier J. Pathophysiology of osteoarthritis. Osteo-
and exercise alone. Future studies should examine the arthritis Cartilage. 1999;7:371–373.
effectiveness of different types and dosages of electrical dry 5. Carter DR, Beaupre GS, Wong M, et al. The mechanobiology
of articular cartilage development and degeneration. Clin
needling and include a long-term follow-up. Orthop Relat Res. 2004;427(suppl):S69–S77.
6. Ene R, Sinescu RD, Ene P, et al. Synovial inflammation in
ACKNOWLEDGMENTS patients with different stages of knee osteoarthritis. Rom J
The authors thank all the participants of the study. Morphol Embryol. 2015;56:169–173.
7. Altman R, Asch E, Bloch D, et al. Development of criteria for
the classification and reporting of osteoarthritis. Classification
APPENDIX 1: DESCRIPTION OF PERIOSTEAL of osteoarthritis of the knee. Diagnostic and Therapeutic
ELECTRICAL DRY NEEDLING INTERVENTION Criteria Committee of the American Rheumatism Association.
Arthritis Rheum. 1986;29:1039–1049.
Technique 8. Altman RD. Criteria for classification of clinical osteoarthritis.
9-point electrical dry needling protocol for knee OA. J Rheumatol Suppl. 1991;27:10–12.
9. Hsu C, Sherman KJ, Eaves ER, et al. New perspectives on
Technique Description patient expectations of treatment outcomes: results from
qualitative interviews with patients seeking complementary
The technique is performed with the patient supine with and alternative medicine treatments for chronic low back pain.
the treated knee slightly flexed over a towel roll. Sterilized BMC Complement Altern Med. 2014;14:276.
disposable stainless steel Seirin J-type acupuncture needles were 10. Pound P, Britten N, Morgan M, et al. Resisting medicines:
used with 3 sizes: 0.25 mm×30 mm, 0.30 mm×40 mm, and a synthesis of qualitative studies of medicine taking. Soc Sci
0.30 mm×50 mm. The depth of needle insertion ranged from Med. 2005;61:133–155.

1156 | www.clinicalpain.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
Clin J Pain  Volume 34, Number 12, December 2018 Electrical Dry Needling for Knee Osteoarthritis

11. Fransen M, McConnell S, Harmer AR, et al. Exercise for 33. Furlan AD, van Tulder MW, Cherkin DC, et al. Acupuncture
osteoarthritis of the knee. Cochrane Database Syst Rev. 2015;1: and dry-needling for low back pain. Cochrane Database Syst
CD004376. Rev. 2005;1:CD001351.
12. Jansen MJ, Viechtbauer W, Lenssen AF, et al. Strength 34. Tough EA, White A. Effectiveness of acupuncture/dry needling
training alone, exercise therapy alone, and exercise therapy for myofascial trigger point pain—a systematic review. Phys
with passive manual mobilisation each reduce pain and Ther Rev. 2011;16:147–154.
disability in people with knee osteoarthritis: a systematic 35. Tough EA, White AR, Cummings TM, et al. Acupuncture and
review. J Physiother. 2011;57:11–20. dry needling in the management of myofascial trigger point
13. Smidt N, de Vet HC, Bouter LM, et al. Effectiveness of exercise pain: a systematic review and meta-analysis of randomised
therapy: a best-evidence summary of systematic reviews. Aust J controlled trials. Eur J Pain. 2009;13:3–10.
Physiother. 2005;51:71–85. 36. Rathnayake T. Back pain (low): acupuncture and dry needling.
14. van Baar ME, Assendelft WJ, Dekker J, et al. Effectiveness of Evidence Summaries-Joanna Briggs Institute; 2009.
exercise therapy in patients with osteoarthritis of the hip or 37. Brinkhaus B, Witt C, Jena S, et al. Interventions and physican
knee: a systematic review of randomized clinical trials. Arthritis characteristics in a randomized multicenter trial of acupuncture
Rheum. 1999;42:1361–1369. in patients with low-back pain. J Altern Complement Med. 2006;
15. Thomas KS, Miller P, Doherty M, et al. Cost effectiveness of a 12:649–657.
two-year home exercise program for the treatment of knee pain. 38. Moffet H. Traditional acupuncture theories yield null out-
Arthritis Rheum. 2005;53:388–394. comes: a systematic review of clinical trials. J Clin Epidemiol.
16. Hay EM, Foster NE, Thomas E, et al. Effectiveness of 2008;61:741–747.
community physiotherapy and enhanced pharmacy review for 39. Deadman P, Al-Khafaji M, Baker K. A Manual of Acupuncture,
knee pain in people aged over 55 presenting to primary care: 2nd ed. Hove, East Sussex, UK: Journal of Chinese Medicine
pragmatic randomised trial. Bmj. 2006;333:995. Publications; 2011.
17. Corbett MS, Rice SJ, Madurasinghe V, et al. Acupuncture and 40. O’Conner J, Bensky D. Acupuncture: A Comprehensive Text.
other physical treatments for the relief of pain due to Seattle, WA: Eastland Press; 1981.
osteoarthritis of the knee: network meta-analysis. Osteoarthritis 41. Mavrommatis CI, Argyra E, Vadalouka A, et al. Acupuncture as
Cartilage. 2013;21:1290–1298. an adjunctive therapy to pharmacological treatment in patients
18. Hochberg MC, Altman RD, April KT, et al. American College with chronic pain due to osteoarthritis of the knee: a 3-armed,
of Rheumatology 2012 recommendations for the use of randomized, placebo-controlled trial. Pain. 2012;153:1720–1726.
nonpharmacologic and pharmacologic therapies in osteoarthri- 42. Vas J, Mendez C, Perea-Milla E, et al. Acupuncture as a
tis of the hand, hip, and knee. Arthritis Care Res (Hoboken). complementary therapy to the pharmacological treatment of
2012;64:465–474. osteoarthritis of the knee: randomised controlled trial. Bmj.
19. Manheimer E, Cheng K, Linde K, et al. Acupuncture for 2004;329:1216.
peripheral joint osteoarthritis. Cochrane Database Syst Rev. 43. Witt C, Brinkhaus B, Jena S, et al. Acupuncture in patients with
2010;1:CD001977. osteoarthritis of the knee: a randomised trial. Lancet. 2005;
20. Zhang W, Moskowitz RW, Nuki G, et al. OARSI recom- 366:136–143.
mendations for the management of hip and knee osteoarthritis, 44. Zhang W, Nuki G, Moskowitz RW, et al. OARSI recom-
Part II: OARSI evidence-based, expert consensus guidelines. mendations for the management of hip and knee osteoarthritis:
Osteoarthritis Cartilage. 2008;16:137–162. part III: changes in evidence following systematic cumulative
21. Lin X, Huang K, Zhu G, et al. The effects of acupuncture on update of research published through January 2009. Osteo-
chronic knee pain due to osteoarthritis: a meta-analysis. J Bone arthritis Cartilage. 2010;18:476–499.
Joint Surg Am. 2016;98:1578–1585. 45. Manheimer E, Linde K, Lao L, et al. Meta-analysis: acupuncture
22. Ezzo J, Hadhazy V, Birch S, et al. Acupuncture for osteo- for osteoarthritis of the knee. Ann Intern Med. 2007;146:868–877.
arthritis of the knee: a systematic review. Arthritis Rheum. 2001; 46. Chen LX, Mao JJ, Fernandes S, et al. Integrating acupuncture
44:819–825. with exercise-based physical therapy for knee osteoarthritis:
23. White A, Foster NE, Cummings M, et al. Acupuncture a randomized controlled trial. J Clin Rheumatol. 2013;19:308–316.
treatment for chronic knee pain: a systematic review. Rheuma- 47. Foster NE, Thomas E, Barlas P, et al. Acupuncture as an
tology (Oxford). 2007;46:384–390. adjunct to exercise based physiotherapy for osteoarthritis of the
24. Brosseau L, Taki J, Desjardins B, et al. The Ottawa panel knee: randomised controlled trial. Bmj. 2007;335:436.
clinical practice guidelines for the management of knee osteo- 48. Scharf HP, Mansmann U, Streitberger K, et al. Acupuncture
arthritis. Part two: strengthening exercise programs. Clin and knee osteoarthritis: a three-armed randomized trial. Ann
Rehabil. 2017;31:596–611. Intern Med. 2006;145:12–20.
25. Casanueva B, Rivas P, Rodero B, et al. Short-term improve- 49. Zwarenstein M, Treweek S, Gagnier JJ, et al. Improving the
ment following dry needle stimulation of tender points in reporting of pragmatic trials: an extension of the CONSORT
fibromyalgia. Rheumatol Int. 2014;34:861–866. statement. Bmj. 2008;337:a2390.
26. Manheimer E, White A, Berman B, et al. Meta-analysis: 50. Deyle GD, Allison SC, Matekel RL, et al. Physical therapy
acupuncture for low back pain. Ann Intern Med. 2005;142:651–663. treatment effectiveness for osteoarthritis of the knee: a randomized
27. Gunn CC, Milbrandt WE, Little AS, et al. Dry needling of muscle comparison of supervised clinical exercise and manual therapy
motor points for chronic low-back pain: a randomized clinical trial procedures versus a home exercise program. Phys Ther. 2005;
with long-term follow-up. Spine (Phila Pa 1976). 1980;5:279–291. 85:1301–1317.
28. Lewit K. The needle effect in the relief of myofascial pain. Pain. 51. Kong J, Gollub R, Huang T, et al. Acupuncture de qi, from
1979;6:83–90. qualitative history to quantitative measurement. J Altern Comple-
29. Dunning J, Butts R, Mourad F, et al. Dry needling: a literature ment Med. 2007;13:1059–1070.
review with implications for clinical practice guidelines. Phys 52. Zhou W, Benharash P. Significance of “Deqi” response in
Ther Rev. 2014;19:252–265. acupuncture treatment: myth or reality. J Acupunct Meridian Stud.
30. Neal BS, Longbottom J. Is there a role for acupuncture in the 2014;7:186–189.
treatment of tendinopathy? Acupunct Med. 2012;30:346–349. 53. Weiner DK, Moore CG, Morone NE, et al. Efficacy of
31. Zhou K, Ma Y, Brogan MS. Dry needling versus acupuncture: periosteal stimulation for chronic pain associated with
the ongoing debate. Acupunct Med. 2015;33:485–490. advanced knee osteoarthritis: a randomized, controlled clinical
32. Furlan AD, van Tulder M, Cherkin D, et al. Acupuncture and trial. Clin Ther. 2013;35:1703.e1705–1720.e1705.
dry-needling for low back pain: an updated systematic review 54. Zhang Q, Yue J, Lu Y. Acupuncture treatment for chronic knee
within the framework of the Cochrane Collaboration. Spine pain: study by Hinman et al underestimates acupuncture
(Phila Pa 1976). 2005;30:944–963. efficacy. Acupunct Med. 2015;33:170.

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved. www.clinicalpain.com | 1157
Dunning et al Clin J Pain  Volume 34, Number 12, December 2018

55. Mata J, Cabrera S, Sanchis P, et al. Electro-acupuncture for measured on a numerical rating scale. Eur J Pain. 2004;
treatment of knee pain from osteoarthritis and the possible 8:283–291.
endocrinology changes: a study protocol for a randomized 69. Jaeschke R, Singer J, Guyatt GH. Measurement of health
controlled trial. Trials. 2015;16:248. status. Ascertaining the minimal clinically important difference.
56. Sangdee C, Teekachunhatean S, Sananpanich K, et al. Electro- Control Clin Trials. 1989;10:407–415.
acupuncture versus diclofenac in symptomatic treatment of 70. Carlesso LC, Macdermid JC, Santaguida LP. Standardization
osteoarthritis of the knee: a randomized controlled trial. BMC of adverse event terminology and reporting in orthopaedic
Complement Altern Med. 2002;2:3. physical therapy: application to the cervical spine. J Orthop
57. Menz HB. Analysis of paired data in physical therapy research: time Sports Phys Ther. 2010;40:455–463.
to stop double-dipping? J Orthop Sports Phys Ther. 2005;35:477–478. 71. Rubin LH, Witkiewitz K, Andre JS, et al. Methods for
58. Sun Y, Sturmer T, Gunther KP, et al. Reliability and validity of handling missing data in the behavioral neurosciences: don’t
clinical outcome measurements of osteoarthritis of the hip and knee throw the baby rat out with the bath water. J Undergrad
—a review of the literature. Clin Rheumatol. 1997;16:185–198. Neurosci Educ. 2007;5:A71–A77.
59. Bellamy N, Buchanan WW, Goldsmith CH, et al. Validation 72. Zhang Y, Bao F, Wang Y, et al. Influence of acupuncture in
study of WOMAC: a health status instrument for measuring treatment of knee osteoarthritis and cartilage repairing. Am J
clinically important patient relevant outcomes to antirheumatic Transl Res. 2016;8:3995–4002.
drug therapy in patients with osteoarthritis of the hip or knee. 73. Chen N, Wang J, Mucelli A, et al. Electro-acupuncture is
J Rheumatol. 1988;15:1833–1840. beneficial for knee osteoarthritis: the evidence from meta-analysis
60. McConnell S, Kolopack P, Davis AM. The Western Ontario of randomized controlled trials. Am J Chin Med. 2017;45:965–985.
and McMaster Universities Osteoarthritis Index (WOMAC): 74. Langevin HM, Schnyer R, MacPherson H, et al. Manual and
a review of its utility and measurement properties. Arthritis electrical needle stimulation in acupuncture research: pitfalls
Rheum. 2001;45:453–461. and challenges of heterogeneity. J Altern Complement Med.
61. Angst F, Aeschlimann A, Stucki G. Smallest detectable and 2015;21:113–128.
minimal clinically important differences of rehabilitation 75. Weiner DK, Rudy TE, Morone N, et al. Efficacy of periosteal
intervention with their implications for required sample sizes stimulation therapy for the treatment of osteoarthritis-associ-
using WOMAC and SF-36 quality of life measurement instru- ated chronic knee pain: an initial controlled clinical trial. J Am
ments in patients with osteoarthritis of the lower extremities. Geriatr Soc. 2007;55:1541–1547.
Arthritis Rheum. 2001;45:384–391. 76. Mcindoe A. A comparison of acupunture with intra-articular
62. Williams VJ, Piva SR, Irrgang JJ, et al. Comparison of steroids injection as analgesia for osteoarthritis of the hip.
reliability and responsiveness of patient-reported clinical out- Acupunct Med. 1995;13:67–70.
come measures in knee osteoarthritis rehabilitation. J Orthop 77. Liu X, Shen L, Wu M, et al. Effects of acupuncture on
Sports Phys Ther. 2012;42:716–723. myelogenic osteoclastogenesis and IL-6 mRNA expression.
63. Abbott JH, Hobbs C, Moa Trial T. The ShortMAC: minimum J Tradit Chin Med. 2004;24:144–148.
important change of a reduced version of the WOMAC 78. Ahsin S, Saleem S, Bhatti AM, et al. Clinical and endocrino-
Osteoarthritis Index. J Orthop Sports Phys Ther. 2017:1–16. logical changes after electro-acupuncture treatment in patients
64. Jensen MP, Karoly P, Braver S. The measurement of clinical with osteoarthritis of the knee. Pain. 2009;147:60–66.
pain intensity: a comparison of six methods. Pain. 1986;27: 79. Loaiza LA, Yamaguchi S, Ito M, et al. Electro-acupuncture
117–126. stimulation to muscle afferents in anesthetized rats
65. Cleland JA, Childs JD, Whitman JM. Psychometric properties modulates the blood flow to the knee joint through autonomic
of the Neck Disability Index and Numeric Pain Rating Scale in reflexes and nitric oxide. Auton Neurosci. 2002;97:103–109.
patients with mechanical neck pain. Arch Phys Med Rehabil. 80. Huang J, Zhuo LS, Wang YY, et al. Effects of electro-
2008;89:69–74. acupuncture on synovia IL-1beta and TNF-alpha contents in
66. Young IA, Cleland JA, Michener LA, et al. Reliability, construct the rabbit with knee osteoarthritis. Zhen Ci Yan Jiu. 2007;32:
validity, and responsiveness of the Neck Disability Index, patient- 115–118.
specific functional scale, and numeric pain rating scale in patients with 81. Zhang R, Lao L, Ren K, et al. Mechanisms of acupuncture-
cervical radiculopathy. Am J Phys Med Rehabil. 2010;89:831–839. electroacupuncture on persistent pain. Anesthesiology. 2014;120:
67. Farrar JT, Young JP Jr, LaMoreaux L, et al. Clinical 482–503.
importance of changes in chronic pain intensity measured on 82. Li ZD, Cao LH, Wang SC. Effect of moxibustion in treating
an 11-point numerical pain rating scale. Pain. 2001;94:149–158. knee joint osteoarthritis and its relation with contents of
68. Salaffi F, Stancati A, Silvestri CA, et al. Minimal clinically hyaluronic acid in serum and synovial fluid. Zhongguo Zhong
important changes in chronic musculoskeletal pain intensity Xi Yi Jie He Za Zhi. 2009;29:883–885.

1158 | www.clinicalpain.com Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.