Subacromial Pain Syndrome
- Chiropractic Post
- Feb 18
- 24 min read
Dry Needling for Subacromial Pain Syndrome: A Systematic Review with Meta-Analysis Free
David Griswold, PT, DPT, PhD ,
Ken Learman, PT, PhD ,
Edmund Ickert, PT, DPT ,
Annie Tapp, PT, DPT, NCS ,
Omar Ross, PT, DPT, OCS
Pain Medicine, Volume 24, Issue 3, March 2023, Pages 285–299, https://doi.org/10.1093/pm/pnac131
Published:
26 August 2022
Article history
A correction has been published: Pain Medicine, Volume 24, Issue 7, July 2023, Pages 917–921, https://doi.org/10.1093/pm/pnad064
Views
Share Icon Share
Abstract
Objective
The authors of this systematic review with meta-analysis evaluated the evidence for the effectiveness of various applications of dry needling (DN) combined with other conservative treatments for subacromial pain syndrome (SAPS).
Methods
Six databases (PubMED, CINAHL, Biosis, Web of Science, SPORTDiscus, and Cochrane Central Register of Controlled Trials) were searched after the study had been registered in PROSPERO. The authors included randomized clinical trials investigating the clinical effects of DN in combination with other conservative interventions for SAPS. Outcomes included pain and disability.
Results
Eight studies were selected. All eight studies involving 10 comparisons were included in the analyses (N = 538). A random-effects model was used to analyze between-group effects. Dry needling performed in combination with other conservative interventions produced favorable outcomes at all time points for pain and disability. Standard mean differences ranged from –0.57 (moderate) to –1.29 (large) for pain and –0.69 (moderate) to –1.07 (large) for disability, favoring groups receiving DN in addition to conservative treatment. Four of the eight studies were rated as having unclear or high risk of bias.
Conclusion
The meta-analysis suggests that various applications of DN performed with other conservative interventions are more effective than conservative treatment alone for reducing pain and disability in patients with SAPS. Direct-comparison studies are needed to determine whether one application of DN is superior to another.
Dry Needling, Subacromial Pain Syndrome, Shoulder Impingement
Topic:
pain
shoulder region
pain disorder
disability
simplified acute physiology score
statistical analysis plan
dry needling
Issue Section:
Introduction
Subacromial pain syndrome (SAPS) is described as nontraumatic, unilateral shoulder pain around the acromion, which worsens during or after arm elevation [1]. Shoulder conditions such as bursitis, tendinosis, supraspinatus tendinopathy, partial rotator cuff tear, biceps tendinitis, and rotator cuff degeneration are subtypes of SAPS. The accepted diagnostic test cluster for SAPS involves combining the Hawkins-Kennedy test, the painful arc test, and the resisted infraspinatus muscle test [1]. Prevalence estimates of SAPS range from 7% to 26% for the general population [2], and it accounts for about half of all primary care consultations for shoulder dysfunction [3].
The exact etiology of SAPS is unknown. Rotator cuff muscle performance impairments [4], altered shoulder kinematics and capsular tightness [5], and scapular dyskinesis [6] have all been associated with SAPS. Suprascapular nerve dysfunction has also been hypothesized to generate symptoms related to the rotator cuff pathology [7, 8]. Myofascial pain or “myofascial trigger points” (MTrP) are hyperirritable spots in skeletal muscles accompanied by a palpable nodule that, when pressed, can produce local or referred pain [9]. Myofascial pain is complex, and its pathophysiology remains elusive. Recent evidence shows that specific MTrP biomarkers, systemic biomarkers, and sensitization of the nervous system could all play a role in the development of myofascial pain [10]. Altered blood flow associated with myofascial pain and surrounding soft tissue [11, 12] is another physiological mechanism for myofascial pain. Although the presence of MTrP is a common feature of SAPS [13], it remains unknown whether MTrP could cause SAPS, results from SAPS, or is a secondary clinical finding caused by spinal segmental sensitization [14] or central sensitization [15, 16].
Conservative management for SAPS includes exercise, manual therapy, injections, medication, electrotherapy, and cognitive behavioral therapy as recommended first-line treatment [1, 17]. An update of systematic reviews [18] recommends both manual therapy and exercise done as a multimodal intervention strategy. Needling therapies involving medicinal injections [19], percutaneous electrical nerve stimulation (PENS) [20], percutaneous electrolysis (PE) [21, 22], or dry needling (DN) [23] are also routinely used as nonsurgical treatments.
Dry needling is a minimally invasive procedure in which a monofilament or acupuncture needle is used to penetrate symptomatic soft tissue to reduce pain and disability [24]. MTrP DN involves inserting a solid needle into a local focus in the muscle and repeatedly manipulating the needle to elicit and exhaust the local twitch response [25]. Repeatedly eliciting the local twitch response can result in more posttreatment soreness [26], and no studies have demonstrated long-term benefits on pain correlating with the local twitch response [27]. In addition to targeting an MTrP, clinicians can also perform DN on peri-neural tissue [28], joint structures/periosteum [29, 30], and tendons [31]. Dry needling can also be applied by inserting multiple needles [32], needling away from the site of pain [28], rotating the needle [33], allowing longer needle retention time [34], or adding electrical stimulation to the needle [35].
There are several reasons DN might be beneficial for SAPS. Mechanisms of DN include various peripheral and central nervous system effects associated with pain modulation [36–38]. Dry needling improves microcirculation [39] by altering biochemical [40, 41] and neuroinflammatory mediators [42] and inducing mechanical disruption of connective tissue [43]. Dry needling also reduces neuromuscular end-plate activity [44, 45] and activates central pathways to alleviate pain [38]. Furthermore, DN improves sympathetic activity, thus reducing local and remote mechanosensitivity [46]. Dry needling of tendons reduces pain and disability for up to 6 months for various tendinopathies [31, 47]. Tendon needling increases blood flow through vasodilation [48] and the promotion of collagen proliferation [43]. Dry needling can also be used to address neuromuscular dysfunction [49, 50] and reduce peripherally sensitized nerves [30, 51] by improving microcirculation. Various applications of electrical stimulation can be incorporated to enhance the local blood flow changes [52] and central neurophysiological mechanisms [53, 54] for further pain reduction [35].
Percutaneous electrical nerve stimulation, otherwise referred to as “dry needling with electrical stimulation,” uses either a pulsed electrical [55] or continuous biphasic current to an inserted needle [35]. The use of PENS has been shown to be effective for various painful musculoskeletal conditions [29, 56, 57], including SAPS [20, 58]. PE is similar to PENS; however, PE uses a continuous galvanic current and is performed under ultrasound. PE has also been shown to be an effective treatment for musculoskeletal pain [35].
Several studies [20–22, 25, 59–62] have been published investigating the clinical effect of DN on SAPS, with varying results. A recent descriptive systematic review [23] reported that adding DN to other conservative interventions produces superior outcomes for improving pain, range of motion, and disability in SAPS, but no quantitative evaluation was completed. Another systematic review with meta-analysis investigated the clinical benefits of MTrP DN for nontraumatic shoulder pain, reporting small to moderate effect sizes for pain and large effect sizes for disability [63]. The review included studies involving only MTrP DN and studies involving other shoulder conditions besides SAPS. To date, no systematic review with meta-analysis has been published investigating the effects of adding various applications of DN to conservative interventions for SAPS. The purpose of the present review was to quantitatively evaluate the effectiveness of DN for SAPS.
Methods
Protocol and Registration
This study was prospectively registered with the International Prospective Register of Systematic Reviews (PROSPERO# CRD42022303063). We used the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines during the execution and reporting phases of this study [41].
Data Sources and Searches
An electronic search of six databases (PubMED, CINAHL, Biosis, Web of Science, SPORTDiscus, and Cochrane Central Register of Controlled Trials) was completed after the study had been registered in PROSPERO. Database searches included articles published from the time of database inception until the searches were completed on March 1, 2022. A health sciences librarian helped develop the search strategy, which used Medical Subject Heading (MeSH) terms, keywords, and text words related to DN and SAPS for articles written in English. Search strategies were altered as necessary to complete searches in all databases. Database-specific filters were used if available to restrict article searches to randomized controlled trials, controlled trials, or other trials with an appropriate control group. In addition, hand searches of reference lists from selected articles were performed. The complete search strategy for PubMed can be seen in Supplementary Data Appendix 1.
Inclusion Criteria
Population
We included studies that involved participants of any age with a diagnosis of SAPS. The search strategy had to contain at least one of the following key words: “trigger points” OR “myofascial trigger points” OR “trigger point” OR “subacromial pain” OR “subacromial pain syndrome” OR “shoulder impingement.”
Intervention
Interventions included DN in which a solid filiform or acupuncture-like needle was used with or without electrical stimulation. The search strategy had to contain at least one of the following key words: (Needles OR Needling OR “dry needling” OR “dry needle” OR “electro dry needling” OR “intramuscular electrical stimulation” OR “percutaneous electrical”).
Comparator
The comparison groups included conservative physical therapy interventions. The search strategy had to contain at least one of the following key words: (“physical therapy” OR “manual therapy”).
Designs
The trial designs were limited to randomized clinical trials. The search strategy included (“Randomized Clinical Trial”).
Outcomes
Primary outcomes were pain and disability. These constructs could have been measured with any commonly used and validated instrument. Pain is most commonly measured subjectively on several scales, such as the 100-mm visual analog scale [64] or the numeric pain rating scale [65]. Various disability outcomes were used for SAPS.
Data Extraction and Quality Assessment
Authors (DG, AT) independently screened titles, abstracts, and full-text articles in three steps and resolved disagreements through discussion. A third author (KL) was available to act as a tiebreaker if needed. The reliability of author agreement at each screening step was determined with an unweighted Cohen’s Kappa (k) [42]. Data were extracted independently by two authors (DG, EI) and entered into a standardized data extraction form. The accuracy of data extraction was confirmed by a third author (KL). Extracted data were information pertaining to 1) trial design, 2) sample size, 3) diagnosis, 4) subject demographics (subject inclusion/exclusion), 5) sample size, 6) intervention, 7) duration of care, 8) follow-up time frames, 9) outcome measures used, and 10) measures of central tendency and dispersion.
Two investigators (DG, OR) independently assessed the included trials using the Cochrane Risk-of-Bias (RoB) tool. Disagreements were resolved through discussion. A third investigator (KL) was available as a tiebreaker if needed. The Cochrane RoB tool assesses bias across six domains with seven items listed in order as column headings in Figure 1: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other biases. The seven items are rated as having “high,” “low,” or “unclear” risk [66]. We classified studies as having a high overall RoB if they were rated as having high or unclear RoB for allocation concealment, random allocation, or incomplete outcome assessment [67, 68]. Because of the difficulty of blinding participants and personnel from allocation to treatment assignment, high RoB in this domain did not automatically result in an overall assessment of high RoB. Interrater reliability was determined with an unweighted Cohen’s kappa [69].
Figure 1.
RoB for the included articles.
Data Synthesis and Analysis
Study data for pain and disability outcomes were pooled and analyzed to synthesize results across studies in Review Manager (RevMan) [Computer program]. Version 5.4, The Cochrane Collaboration, 2020. Data conversions were implemented when calculation was possible on the basis of the data reported. Standardized mean differences (SMDs) with 95% confidence intervals (CIs) were calculated, as outcomes were reported on different scales. The I2 statistic represented heterogeneity across studies to determine whether a fixed-effects or random-effects model was appropriate. Level of heterogeneity was categorized as follows: ≤25%= homogeneity, 50%= average homogeneity, and 75% or greater= heterogeneity [46]. The random-effects SMD values were used for overall effect magnitude interpretation because some analyses had heterogeneity. Cohen’s interpretation of SMD values was used, with an effect size of 0.2 being small, 0.5 being moderate, and 0.8 being large [47]. Clinical importance was determined by calculating the weighted mean difference at each time point and comparing with the minimally clinically important difference of the outcomes.
Comparisons for the effect of including DN on both pain and disability were made at three time points: 1–4 weeks, >4 weeks to 3 months, and >3 to 12 months. The cumulative total for each time point was used in the meta-analysis.
Results
Study Selection
The database search returned 2,711 trials, and once duplicates had been removed, 1,868 studies were screened. After title and abstract screening, 26 studies were selected for full-text review. Figure 2 provides a flow diagram of the selection process. Eight studies were included after full-text review.
Figure 2.
PRISMA Flow Diagram for included studies.
Study Characteristics
This systematic review included eight studies [20–22, 25, 59–62] meeting the inclusion criteria for full-text review. Eight studies involving 10 comparisons provided data that were analyzed. Seven studies [20–22, 25, 59, 61, 62] involving nine comparisons provided data for pain comparisons. One of those studies [60] provided data for disability only. The aggregate sample size of all eight studies included in the meta-analysis was 538. Two of the studies [61, 62] involved additional comparisons, as one of the studies [62] compared their control group (muscle energy) to DN or DN with muscle energy. The other study [61] compared two types of DN with a control group. These additional comparisons resulted in the control group data being included twice. The numbers of treatment sessions for the intervention groups were different from those for the control groups in two studies [59, 61]. All other studies had an equal number of sessions per group. Additional study characteristics can be found in Table 1.
Table 1
Study characteristics
Study (author, date) | Study design | N: DN/Control | Diagnosis | Inclusion criteria | Exclusion criteria | Between-Group Interventions | Other treatment | Pain | Disability |
Imani, 202161 | Single blinded RC | Hong DN+PT: 22 DDN+PT: 22 PT: 22 | Subacromial Impingement Syndrome | Hawkins-Kennedy tests and Neer sign were positive in patients and infraspinatus muscle had at least one trigger point with referral pain pattern |
|
| NA | NPRS | SPADI |
Johansson, 200560 | Single blinded RCT | 44/41 | Subacromial Impingement Syndrome | Age: 30-65, Pain with arm elevation, (+) Neer Impingement, 2-month history. 3 of 4: (+) Hawkins/Kennedy, Jobe test, Neer Impingement, painful arc | Abnormal radiology, systemic inflammatory diagnosis, dislocation or instability history, frozen shoulder, Neck pathology, received either treatment, communication problem. | DN: 30 mm x.30 gauge rotated 3x (to create sensation) in predetermined locations locally around the shoulder and one point distally. Ultrasound: 10 minutes of 1 MHz and 1W/cm2 using a 4cm2 transducer. | Groups received 2-step home exercise program: without provoking pain. | NA | AL, UCLA, CM combined scores. |
Jalilipanah, 202162 | Single blinded RCT | DN: 13 DN+MET: 13. MET: 13 | Shoulder Impingement Syndrome | (+) Neer and Hawkins-Kennedy Tests, Presence of trigger points | Prior treatment, pregnant, postural disorder, allergy or fear or needles, contraindications to DN, medication use, tumor, myopathy, DVT, prior surgery, Fibromyalgia, Frozen Shoulder |
| NA | VAS | NA |
Arias-Buria, 201759 | Single blinded RC | 25/25 | Subacromial Pain Syndrome | Unilateral atraumatic shoulder pain of 3 months duration, pain level of at least 4/10. (+) painful arc test, and 2/3 (+) Hawkins-Kennedy test, drop arm test, or lift off test | Bilateral shoulder pain, <18 or >65 years old, hstory of fractures or dislocations, cervical radiculopathy, steroid interventions, Fibromyalgia, prior surgery of neck or shoulder, fear of needles, contraindications to treatment. | DN (2 sessions) +Exercise (1x/wk. for 4 sessions): MTrP DN using.32 mm 40 mm Exercise only (1x/wk. for 4 sessions): 3 x12 repetitions concentric phase after the eccentric phase of the exercise focusing on supraspinatus, infraspinatus, and scapular stabilizer musculature. | NA | NPRS | DASH |
Arias-Buría, 201521 | Single blinded RCT | 17/19 | Subacromial Pain Syndrome | Unilateral atraumatic shoulder pain of 3 months duration, 4/10 pain durng painful arc, at least one (+) Hawkins-Kennedy test, Neer, empty can, drop arm test, or lift off test, positive findings on MRI. | Bilateral shoulder pain, <18 or >65 years old, history of fractures or dislocations, cervical radiculopathy, steroid interventions, Fibromyalgia, prior surgery of neck or shoulder. | US-Guided DN (PE)+Eccentrics (1x/wk. for 4 weeks): 0.3x25mm needles directed towards the supraspinatus tendon. Intensity of PE :350 μA during 1.2min. Eccentrics (1x/week for 4 wks.): 3 eccentric exercises 3x10 reps: supraspinatus, infraspinatus, and scapular muscles. 2x/ay for 4 weeks. | NA | NPR | DASH |
de Miguel Valtierra22 | Single blinded RCT | 25/25 | Subacromial Pain Syndrome | Unilateral shoulder pain for 3 months, pain of 4/10 during arm elevation, positive painful arc and at last 1 of the following: (+) Hawkins-Kennedy, Neer sign, empty can test, drop arm, or lift off test. | Bilateral shoulder pain, <18 or >65 years old, history of fractures or dislocations, cervical radiculopathy, steroid interventions, Fibromyalgia, prior surgery of neck or shoulder, prior intervention to the neck or shoulder. | DN (PE)+Manual Therapy +Exercise (1x/wk. for 5 weeks): .30mmx 25mm needle inserted into the supraspinatus tendon and applied an intensity of 350mA for a total of 90 seconds. Manual Therapy + Exercise (1x/wk. for 5 weeks): Joint mobilizations to shoulder complex. 3 x12 concentric phase following eccentric phase for supraspinatus, infraspinatus, and scapular stabilizer musculature. Performed 2x/day for 5 weeks. | NA | NPRS | DASH/SPADI |
Ekici, 202125 | Single blinded RCT | 21/19 | Subacromial Pain Syndrome | Unilateral shoulder pain for 3 months, between ages of 18-65, can consent and communicate, night pain, and have trigger points. | Bilateral shoulder pain, contraindicated to treatments, shoulder fracture or dislocation history, cervical radiculopathy, Fibromyalgia, neurological signs/symptoms, surgery to neck or shoulder, treatment to the shoulder in past 6 months, and pregnant. | DN (6 sessions over 4 wks.): .25-25 mm, .25-30 mm, using a fast in/out technique to elicit and exhaust the LTR. Treated shoulder muscles with identified trigger points. Deep Friction Massage (2x/wk. for 3 wks.): Transverse massage in tissue fiber direction until pain reduction occurred. | After 6 sessions of manual treatment, subjects were given a HEP of AROM and shoulder exercises. | VAS | NA |
Dunning, 202120 | Single Blinded RCT | 73/72 | Subacromial Pain Syndrome | Chief complaint of anterolateral shoulder pain, longer than 6 wks., (+) Neer Impingement and/or Hawkins-Kennedy test. 1 or more of the following: panful arc, pain with resisted external rotation, pain with resisted supraspinatus test. | Steroid treatment, prior surgery, contraindications to treatments, shoulder dislocation, adhesive capsulitis, whiplash in past 6 weeks, full thickness supraspinatus tear, breast cancer on involved side, ACJ pathology, Cervical spine referral, pregnant, pending litigation, prior treatment to the shoulder in last 3 months | DN (PENS)+Manipulation (2x/wk. x 6wks): 20 minutes of electrical DN to 8 points; local (intramuscular, musculotendinous, peri-articular, anterolateral/posterolateral acromion) and distal (scapular, lateral brachium), Optional: upper thoracic, peri-scapular, and glenohumeral areas. Cervical and thoracic spinal manipulation. Joint mobilization +exercise (2x/wk. x 6 wks.): Joint mobilizations to the shoulder complex, soft tissue mobilization, Interferential Current 15-20 min. Shoulder stretching and strengthening exercises. | NA | NPRS | SPADI |
Abbreviations in alphabetical order: ACJ: Acromioclavicular joint; AL: Adolfsson-Lysholm Shoulder Score; AROM: Active range of motion; CM: Constant-Murley Shoulder Assessment; DASH: Disabilities of the Arm, Shoulder, and Hand questionnaire; DN: Dry needling; DDN: Deep dry needling; DVT: Deep vein thrombosis; HEP: Home exercise program; G: gauge; LTR: Local twitch response; MET: Muscle energy technique; NPRS: Numeric pain rating scale; PE: Percutaneous electrolysis; PENS: ; RCT: Randomized control trial; SPADI: Shoulder pain and disability index; MTrP: Myofascial Trigger point; UCLA: University of California at Los Angeles End-Result Score; US: Ultrasound; VAS: Visual analogue scale; DN: Dry Needling; Wks: Weeks
Treatment Diagnoses
Five studies identified the diagnosis of interest as SAPS [20–22, 25, 59, 61], and three reported subacromial impingement syndrome [60–62], which are diagnostically synonymous terms [70]. Diagnostic criteria varied, but seven studies’ inclusion criteria required a positive Neer’s sign or Hawkins-Kennedy test [20–22, 59–62]. Those positive clinical tests were clustered with other clinical tests (empty can test, drop arm test, lift-off test, Jobe test, painful arc test, resisted external rotation test, or pain with resisted empty can test) in five studies [20–22, 59, 60]. Three studies included the identification of MTrP [25, 61, 62]. Only one study required positive imaging findings [21].
Intervention Details
Two studies [21, 22] used ultrasound-guided PE, targeting the supraspinatus tendon. Both study protocols included electrical stimulation parameters of 350 µA (microamperes) for 90 seconds [22] or 1.2 minute [21] for four sessions. In one of these studies, subjects received manual therapy and performed exercises, but PE was also done [22]. In the other study, subjects performed supervised eccentric exercises, but those who were randomized into the experimental group also received PE [21].
One study [61] investigated the effects of two types of DN. All subjects enrolled in the study received similar physical therapy interventions, including interferential current, hot pack, and exercises. Subjects were randomized to receive either Hong DN or deep DN (DDN) targeting MTrP of the infraspinatus during sessions three, five, and seven. The Hong DN method [71] was described as the repeated manipulation of the needle into the MTrP, followed by immediate removal once a localized response occurred. The DDN method involved the insertion of the needle into the targeted MTrP, with the needle left in for 10 minutes after a local reaction. After 10 minutes, the needle was rotated and manipulated, followed by another 10 minutes of the needle left in situ. Then, the needle was removed. All groups received 10 sessions of physical therapy.
Three studies [25, 59, 62] involved MTrP DN targeting symptomatic muscles of the shoulder. One of these studies [62] compared DN with or was done in combination with muscle energy techniques. A second study directly compared the effects of DN with those of deep friction massage [25]. In the third study [59], groups received an exercise program targeting the supraspinatus, infraspinatus, and scapular muscles, but one group also received MTrP DN.
Two studies [20, 60] involved a protocol of DN targeting various soft tissue structures. One of these studies [20] involved 20 minutes of PENS to eight selected points involving various muscles, musculotendinous, and peri-articular structures. Additionally, if indicated, DN could be performed in the thoracic, scapular, and glenohumeral area. Dry needling was combined with cervical, thoracic, and rib manipulation based on identified impairments. The comparison group received impairment-based joint mobilizations and exercise. In the other study [60], physical therapists used a standardized protocol of acupoints around the shoulder and one point distally in the hand. Three rounds of needle rotations were completed. Acupoints are similar to acupuncture points in Traditional Chinese Medicine and are hypothesized to be derived by rich concentrations of sensory nerve endings that become sensitized with injury [72]. Although DN and acupuncture share similarities in the needles used and the physiological mechanisms known to occur that facilitate pain relief [36, 73, 74], there are differences. Acupuncture and DN vary in theoretical constructions and educational requirements and have a different philosophical background [73]. The control group received 10 minutes of ultrasound. Both groups received a two-step home exercise program.
Outcomes
Seven studies included pain as an outcome variable [20–22, 25, 59, 61, 62]. Five studies [20–22, 59, 61] used the numeric pain rating scale and two [25, 62] the visual analog scale. Six studies [20–22, 59–61] included outcome instruments for disability. Three [20, 22, 61] of those studies collected the Shoulder Pain and Disability Index (SPADI) and three [21, 22, 59] of them collected the Disability of Arm, Shoulder, and Hand (DASH) questionnaire. One study collected both the DASH and SPADI [22]. One study [60] used the aggregate scores calculated from the Constant-Murley Shoulder Assessment, the Adolfsson-Lysholm Shoulder Score, and the University of California at Los Angeles End-Result Score. The authors [60] analyzed data from both the subgroup adherent to the study protocol and the full study group, using an intention-to-treat model. We elected to use the data for those who completed the trial, as attrition numbers were similar between groups, the purpose of the present review was to establish treatment effect, and the method used in the intention-to-treat was last observed value carried forward and not multiple imputations [75].
Risk of Bias
Four trials [25, 59, 61, 62] were rated as having unclear or high RoB and four [20–22, 60] as having low RoB. See Figure 1. Trials were considered to have a high RoB because of high or unclear RoB in random allocation (zero trials), allocation concealment (two trials), or incomplete outcome assessment (two trials). Interrater reliability for RoB on the seven criteria was κ = 0.817 (95% CI: 0.80 to 0.95; absolute agreement, 94%). Cohen’s kappa for the final assessment of high or low RoB was κ = 0.83 (95% CI: 0.89 to 1.02; absolute agreement, 92%).
Pain Outcomes
Pain outcomes at 1 to 4 weeks. Seven studies [20–22, 25, 59, 61, 62] reported on pain during the time interval from 1 to 4 weeks. Three of those studies [59, 61, 62] reported nonsignificant between-group differences for pain between groups that received DN and the comparator. One of those studies [61] reported nonsignificant findings between the Hong DN group and its control group but reported significant between-group effects between the DDN group and its control group. All other studies [20–22, 25, 61] and comparisons independently reported significant between-group differences on pain intensity favoring DN. All seven studies involving nine comparisons were included in the meta-analysis, resulting in a moderate and significant SMD of –0.57 (95% CI: –0.75 to –0.038), P < 0.001, favoring DN for pain reduction. The I2 demonstrated average homogeneity. See Figure 3A.
Figure 3.
Forest plots for pain illustrating the treatment effects of including dry needling to conservative intervention versus conservative intervention alone (control) at 1–4 weeks (A), >4 weeks to 3 months (B), and >3 to 12 months (C). Treatment effects ranged from moderate to large. Within the plots, squares represent estimates of treatment effects with larger square representing larger sample sizes. Diamond represents pooled treatment effect. Horizontal line represents 95% confidence intervals. Vertical line at 0 represents no difference.
Pain outcomes at >4 weeks to 3 months. Four studies [20, 22, 59, 61] involving five comparisons reported on pain during the interval from >4 weeks to 3 months. Two of those studies [59, 61] reported nonsignificant between-group differences for pain. One of those studies [61] reported nonsignificant effects for the group who received Hong DN compared with control; however, between-group differences were reported for those who received DDN. Three of the studies [20, 22, 61] reported statistically significant between-group differences for pain favoring DN. Four of the studies involving five comparisons were included in the meta-analysis, resulting in a large and significant SMD of –1.29 (95% CI: –0.86 to –1.72), P<0.001, favoring DN for pain reduction. The I2 demonstrated significant heterogeneity. See Figure 3B.
Pain outcomes at >3 to 12 months. Three studies [22, 25, 59] reported effects on pain during the interval from >3 to 12 months. One of those studies [59] reported no between-group differences for pain, whereas the other two studies [22, 25] reported differences. All three studies were included in the meta-analysis, resulting in a moderate and significant SMD of –0.53 (–0.19 to –0.88), P<0.003, favoring DN for pain reduction. The I2 demonstrated average homogeneity. See Figure 3C.
Disability Outcomes
Disability outcomes at 1 to 4 weeks. Five studies [20–22, 59, 61] involving six comparisons reported on disability outcomes during the interval from 1 to 4 weeks. One of the comparisons from the trial by Imani et al. [61] reported no significant difference in disability for those who received the Hong DN technique compared with control. The other comparison in that study [61], involving DDN compared with the control and Hong DN, did demonstrate between-group differences. The remaining four studies [20–22, 59] also independently reported between-group differences for disability. Five studies [21, 22, 59, 61] involving six comparisons were included in the meta-analysis, resulting in a moderate SMD of –0.69 (95% CI: –0.48 to –0.91), P<0.001, favoring DN for disability. The I2 demonstrated homogeneity. See Figure 4A.
Figure 4.
Forest plots for disability illustrating treatment effects of including dry needling to conservative intervention versus conservative intervention alone (control) at 1–4 weeks (A), >4 weeks to 3 months (B), and >3 to 12 months (C). Treatment effects ranged from moderate to large. Within the plots, squares represent estimates of treatment effects with larger square representing larger sample sizes. Diamond represents pooled treatment effect. Horizontal line represents 95% confidence intervals. Vertical line at 0 represents no difference.
Disability outcomes at >4 weeks to 3 months. Five studies [20, 22, 59–61] involving six comparisons reported on disability during the time frame of >4 weeks to 3 months. Two of those studies [60, 61] individually reported nonsignificant between-group differences for disability. The remaining four studies [20, 22, 59, 61] reported statistically significant between-group effects favoring DN. All five studies [20, 22, 59–61] were included in the meta-analysis, resulting in a large and significant SMD of –0.82 (–0.62 to –1.03), P<0.001, favoring DN. The I2 demonstrated significant heterogeneity. See Figure 4B.
Disability outcomes at >3 to 12 months. Three studies [22, 59, 60] reported on disability during the time frame of >3 to 12 months. One of those studies [60] individually reported nonsignificant between-group differences for disability. The remaining two studies [22, 59] individually reported statistically significant between-group effects favoring DN. All three studies [22, 59, 60] were included in the meta-analysis, resulting in a moderate and significant SMD of –1.07 (–0.73 to –1.41), P<0.001, favoring DN. The I2 demonstrated heterogeneity. See Figure 4C.
Discussion
The authors of the present systematic review aimed to investigate the effectiveness of DN for SAPS. The meta-analysis included eight studies and 10 comparisons for pain and disability outcomes. The results indicate that various applications of DN, when added to other conservative interventions, provide superior outcomes for pain and disability compared with conservative interventions alone. On the basis of our criteria, four of the trials included in this review were rated as having high RoB and four were rated as having low RoB. The I2 indicated heterogeneity for three of the six analyses.
Most studies included in this review followed the diagnostic recommendations of clustering orthopedic tests to determine the likelihood of patients presenting with SAPS [20–22, 59–62]. It was not clear in one study [25] how the determination of SAPS was made, as the diagnosis was made before enrollment. The Hawkins-Kennedy and Neer Impingement tests were the most commonly used clinical tests, performed in seven of the trials [20–22, 59–62]. Three of the studies required the presence of trigger points [25, 61, 62]. One of the studies required imaging findings [21].
This was the first quantitative evaluation of the effectiveness of DN for patients with SAPS. Our findings were consistent with a recent descriptive systematic review that reported positive improvements in pain, disability, and range of motion when DN was added to other conservative interventions for SAPS. The earlier review [23] included nine articles, of which five were also included in the present study. The differences in selection were due to the differences in methodological criteria. We included three additional studies that were of low RoB and contributed substantially to the sample size. However, Blanco-Diaz et al. [23] did not perform a meta-analysis against which to compare our results.
Another recent systematic review with meta-analysis [63] that analyzed the effects of MTrP DN on nontraumatic shoulder pain reported a small treatment effect for pain and a large treatment effect for improving disability in the short term. The effects on disability, however, were based on low-quality evidence. Consistent with their findings, we also observed large effect sizes for disability in both the short and long term. Our findings, in contrast, showed a moderate to large effect size for pain reduction in the short and longer terms. Differences in the effect sizes for pain may be due to the various shoulder conditions included among the reviews and differences in the DN applications. Although four of the trials [25, 59, 61, 62] included in the present review used MTrP DN, some trials involved longer needle retentions [61], needle rotations [60, 61], PENS [20], and PE [21, 22]. Thus, our findings support that in addition to MTrP DN of the involved shoulder musculature, it might be advantageous to apply different DN techniques locally into the tendon or subacromial space and to consider the use of PENS or PE.
Changes observed in the present meta-analysis were also clinically meaningful. The weighted mean differences for pain were 1.2 (95% CI: 0.83 to 1.64) points at 1 to 4 weeks, 1.1 (95% CI: 0.68 to 1.54) points at >4 weeks to 3 months, and 1.2 (95% CI: 0.55 to 1.94) points at >3 months to 1 year. The weighted mean difference values met the minimally clinically important difference for the numeric pain rating scale of 1.1 points for patients with shoulder pain [76]. It should be noted, however, that the lower-bound confidence levels were not clinically meaningful. For disability, the weighted mean differences were 10.2 (95% CI: 6.4 to 14.0) points at 1 to 4 weeks, 11.8 (95% CI: 9.5 to 14.1) points at >4 weeks to 3 months, and 10.4 (95% CI: 7.3 to 13.3) points at >3 to 12 months, which meets the minimally clinically important difference for the DASH of 10.2 points [77] and falls within the range of 8–13 points for the SPADI [78].
Studies included in the present review combined DN with other interventions. In the present review, treatments combined with DN included general exercise programming [25, 60, 61], spinal manipulation [20], manual therapy and exercises [22], eccentric exercises [21, 59], and muscle energy [62]. Although the optimal exercise type and dosage parameters are unknown, exercise is considered the primary intervention for SAPS. Superior outcomes have been reported when exercise is combined with manual therapy techniques [18]. The total treatment effect observed in the meta-analysis includes other interventions combined with DN.
Limitations
There were limitations to this review. The RoB was rated as unclear or high for four of the eight studies. Also, three of the six analyses showed heterogeneity. Therefore, the results should be interpreted with caution. The heterogeneity might be explained by the different types of DN techniques, applications, and dosages applied in the studies. Additionally, some comparison groups in the control received different interventions from those given to the DN group, which could explain some of the variance in the between-group treatment effects. We did not establish an individual treatment effect from DN, as it was combined with other interventions. However, it is unclear whether it is advantageous to attempt to isolate the effect of individual treatments such as DN when multimodal care is the accepted method for managing musculoskeletal disorders [18]. Needling procedures in the present review were performed by physical therapists in seven [20, 22, 25, 59–61] of the eight studies (88%), and in one [21] of those studies, the investigator was referred to as a clinician. This could reduce the generalizability of the findings to other health care professionals legally permitted to perform DN. Lastly, we included only studies that had been published in English.
Conclusion
Our results suggest that various applications of DN, combined with other conservative interventions, produce moderate to large treatment effects for pain and disability in patients with SAPS, in both the short and long terms. Observations were both statistically significant and clinically meaningful. Four studies in this review were rated as having unclear to high RoB and four rated as having low RoB. Future studies should investigate which DN techniques and approaches are most effective for treating SAPS through direct comparisons.
Authors’ Contributions
The authors contributed in the following areas: conceptual/idea/research design—D. Griswold and K. Learman; writing—D. Griswold, K. Learman, O. Ross, A. Tapp, and E. Ickert; data collection—D. Griswold, O. Ross, A. Tapp, and E. Ickert; data analysis—D. Griswold, E. Ickert, and K. Learman; and project management: D. Griswold.
Funding sources: None.
Conflicts of interest: David Griswold teaches Professional Seminars in Dry Needling. No other conflicts of interest exist for any of the other authors.
Study registration: PROSPERO (CRD42022303063).
Acknowledgment
We thank Colleen Duchon, Reference Librarian, Health and Human Services at Youngstown State University.
Supplementary Data
Supplementary Data may be found online at http://painmedicine.oxfordjournals.org.
Comments