EXECUTIVE SUMMARY: PAIN

As shown in the bubble plot, a large number of studies have addressed the treatment of headaches with acupuncture; a 2008 review included 31 RCTs¹²⁴ and 5 independent systematic reviews have been published since 2005. A Cochrane review on tension-type headache by Linde et al., last updated in 2009, reported that 3 to 4 months after randomization, the pooled responder rate ratio was 1.24 (95% confidence interval [CI]: 1.05, 1.46) with 50% responders in the acupuncture groups compared to 41% in sham groups across 4 RCTs.¹²⁰ The review concluded that acupuncture could be a valuable non-pharmacological tool in patients with frequent episodic or chronic tension-type headaches. A 2012 individual patient data meta-analysis published by Vickers et al for the Acupuncture Trialists' Collaboration included data from 29 RCTs evaluating acupuncture for chronic pain.⁴⁷ The review reported that patients receiving acupuncture had less pain, with scores that were 0.23 (95% CI: 0.13, 0.33), 0.16 (95% CI: 0.07, 0.25) and 0.15 (95% CI: 0.07, 0.24) standard deviations lower than sham controls for back and neck pain, osteoarthritis, and chronic headache. The review concluded that acupuncture is effective for the treatment of chronic pain and is a reasonable referral option. However, the most recent available best evidence syntheses concentrating on back pain, neck pain, or osteoarthritis individually do not summarize the evidence as equally unrestrictedly positive, as outlined further below. Thus, the conclusion that acupuncture had evidence of effectiveness with high confidence for chronic pain patients is currently still limited by the lack of conclusive evidence syntheses for the individual conditions that make up 50-65% of chronic pain, namely back pain and neck pain. There is considerable research available for migraine prophylaxis; a 2009 Cochrane review by the same author group working on headaches included 22 acupuncture RCTs.¹²¹ The review reported sufficient detail for a reanalysis and a positive effect across all passive controlled RCTs as defined in this review of reviews was identified. However, it should be noted that effects were driven by RCTs comparing acupuncture to no acupuncture (relative risk [RR] 2.33; 95% CI: 2.02, 2.69), not RCTs comparing acupuncture and sham (RR 1.13; 95% CI: 0.95, 1.35). The review concluded that acupuncture should be considered a treatment option for patients willing to undergo this treatment. More than half of the 7 included RCTs on chronic headaches included in the chronic pain IPD meta-analysis⁴⁷ are also included in the Cochrane reviews on headaches and migraine.

Dysmenorrhea has also been addressed in a large number of primary studies; a 2010 systematic review on primary dysmenorrhea included 27 RCTs.⁹² A Cochrane review on dysmenorrhea last updated in 2012 reported an improvement in pain relief from acupuncture compared with placebo control (odds ratio [OR] 9.5, 95% CI: 21.17, 51.8) and concluded that acupuncture may reduce period pain but further well-designed RCTs are needed.⁷² Osteoarthritis has also been targeted in a large number of systematic reviews (we identified 6 recent reviews from independent author groups) and individual research studies; a 2012 Centre for Reviews and Dissemination (CRD) network meta-analysis on the relief of chronic pain due to osteoarthritis of the knee included 22 acupuncture RCTs.¹⁷⁴ The report, comparing different physical treatments, concluded that acupuncture is one of a number of physical treatments that produces a clinically-relevant effect in alleviating pain in the short-term, and although further research is needed to substantiate these conclusions, acupuncture should be considered as an evidence-based treatment option for relieving pain due to osteoarthritis of the knee. A 2010 Cochrane review on acupuncture for peripheral joint osteoarthritis reported positive results for acupuncture in comparison to sham and waiting list control but not as add-on treatment compared to exercise-based physiotherapy alone. The review concluded that benefits compared to shame were small, did not meet pre-defined thresholds for clinical relevance, and were probably due at least partially to placebo effects from incomplete blinding, while effects compared to waiting list were clinically relevant but could be associated with expectation or placebo effects.⁹⁷ A recent RCT²⁰⁶ not yet included in the existing systematic reviews and one of the largest available studies on acupuncture and osteoarthritis (N=527) reported no statistically significant differences between acupuncture and sham, but a reanalysis combining the largest review and this trial showed that the pooled treatment effect would remain positive if included in an updated meta-analysis. The IPD meta-analysis on chronic pain⁴⁷ included 9 osteoarthritis RCTs. Acupuncture for pain management regardless of the underlying conditions has been addressed in some of the identified reviews; the largest review on auriculotherapy for pain management included 17 RCTs.⁷⁸ The review reported auriculotherapy was superior to controls for studies evaluating pain intensity (standardized mean difference [SMD] 1.56, 95% CI: 0.85, 2.26) but concluded that a more accurate estimate of the effect requires further large, well-designed trials. A 2009 systematic review on acupuncture for pain treatment published in the BMJ concluded that a small analgesic effect of acupuncture was found, which seems to lack clinical relevance and cannot be clearly distinguished from bias.¹¹⁹ A 2013 review on acupuncture for ankle sprain included 17 RCTs.³⁰ The review found that significantly more participants in acupuncture groups reported global symptom improvement compared with no acupuncture (RR 0.56, 95% CI: 0.42, 0.77), but the review was primarily based on non-indexed publications, trial quality was poor, no sham controlled RCT was identified, and the review concluded that given methodological shortcomings and the small number of high-quality primary studies, the available evidence is insufficient to recommend acupuncture as an evidence-based treatment option. Cancer-associated pain has been addressed in 15 RCTs according to the largest recent review.⁵⁹ A 2012 Cochrane review identified one relevant RCT that showed statistically significant differences between the acupuncture and placebo groups but the review concluded there is insufficient evidence to judge whether acupuncture is effective in treating cancer pain in adults.⁷³ Labor pain has also been addressed in a number of primary studies; a 2010 review included 10 RCTs.⁸⁵ A 2011 Cochrane review reported less intense pain from acupuncture compared with no intervention (SMD -1.00, 95% CI: -1.33, -.067) and positive effects for other outcomes and comparators; however, all comparisons were based on one RCT each and the review concluded acupuncture may have a role in relieving pain during labor but more research is needed.⁶⁷

Positive effects were also reported for other clinical indications; however the evidence base was considerably smaller. A review on prostatitis / chronic pelvic pain syndrome included 9 acupuncture RCTs in total and reported a positive effect of acupuncture compared to sham (RR 1.56, 95% CI: 1.09, 2.24); however this result was based on one RCT only.⁷¹ The largest of 4 recent reviews on temporomandibular joint disorders included 7 RCTs in total.¹⁹⁵ The review reported significant improvements in pain intensity for a visual analogue scale (weighted mean difference [WMD] -12.6, 95% CI: -21.2, -6.1) but concluded that further rigorous studies are required to establish beyond doubt whether acupuncture has therapeutic value for this indication. A review on acupuncture for plantar heel pain included 5 RCTs and the passive controlled RCTs reported statistically significant positive results for pain outcomes, but only 2 RCTs were classified as high quality and no pooled result was presented to determine the size of the treatment effect.⁴⁴ A review specific to pregnancy-associated pelvic and back pain included 3 RCTs in total and both acupuncture as add-on treatment RCTs reported statistically significant results. However, no pooled effect was presented to estimate the size of the treatment effect.¹⁴³

Acupuncture for the treatment of back pain has received a great deal of research attention but the evidence base regarding the effectiveness of acupuncture remains unclear judging from the available systematic reviews. We identified 10 recent systematic reviews on acupuncture for back pain and the largest review, a review on the efficacy, cost-effectiveness, and safety of selected complementary and alternative medicine published by Furlan et al in 2012, included 33 acupuncture RCTs.⁶² The review showed a positive effect of acupuncture compared to no treatment but noted that sham-acupuncture controlled trials tended towards statistically nonsignificant results. A 2005 evidence synthesis on low back pain within the framework of the Cochrane Collaboration concluded that the data do not allow firm conclusions regarding the effectiveness of acupuncture for acute low back pain.¹⁷⁰ It is noteworthy that the IPD meta-analysis (see above) on chronic pain which concluded that acupuncture is effective for treating chronic pain also included 10 back pain studies. The largest review on neck pain is the review by Furlan et al. ⁶² published in 2012; it includes 24 acupuncture RCTs. The review came to the same conclusion as for back pain while 2 smaller reviews reported favorable results for acupuncture. The IPD meta-analysis by Vickers at al included some back pain and neck pain studies but was limited to chronic pain (defined as the current episode of pain being of at least 4 weeks' duration), a pooled result was only given for a combined back and neck pain analysis, and indication-specific effects or the individual size of the treatment effect are not known. Acupuncture effects on analgesia during surgery were reviewed by Lee and Ernst in 2005; the review included 19 RCTs and the evidence was judged to be inconclusive. Two systematic reviews on postoperative pain were published in 2008. The study selection was not identical across reviews and there were inconsistent results across included studies; one of the review concluded the evidence that auricular acupuncture reduces postoperative pain is promising but not compelling.¹²³ Fibromyalgia has been addressed in 12 RCTs according to one of 3 recent systematic reviews; effectiveness results are inconsistent within and across reviews. Results regarding shoulder pain are also inconclusive. A 2012 review on shoulder pain after stroke included 3 relevant acupuncture RCTs but did not report a pooled treatment effect estimate; a Cochrane review on shoulder pain, last updated in 2005, identified 9 RCTs with varying results and concluded that due to a small number of clinical and methodological diverse trials, little can be concluded from the review.^167,207 The evidence base for rheumatoid arthritis is also unclear and insufficient data were reported to determine the effectiveness across reviews and included trials. A Cochrane review last updated in 2005 highlighted that conclusions are limited by methodological considerations such as the type of acupuncture, the site of intervention, the small number of clinical trials, and the small sample size of the included studies.¹⁶⁵

A single review on carpal tunnel syndrome was identified that included 6 RCTs. The review did not find statistically significant differences in 2 sham controlled RCTs and conflicting results across outcomes for acupuncture as an add-on treatment in a further RCT.

In addition, a small number of reviews were identified that could not be incorporated in the bubble plot. They addressed primarily the comparative effectiveness of acupuncture in relation to other treatments. The reviews reported that acupuncture was more effective than conventional pharmacological therapies in the treatment of gouty arthritis³⁵ and neurovascular headache (although this is based on a very limited number of studies),⁶⁹ more effective than Chinese herbal medicine for endometriosis,⁶⁵ but no more effective than pharmacological sedation for egg retrieval during assisted reproductive therapy¹⁷³ and of similar efficacy as carbamazepine for trigeminal neuralgia in the existing low-quality studies.¹⁸¹ One systematic review on myofascial trigger point pain reported positive results. However, the number of traditional acupuncture trials, rather than trials on dry needling inserted directly into the trigger points, supporting the result was not reported. A systematic review on acupuncture or acupoint injection for management of burning mouth syndrome¹⁸⁰ found injections to be superior to laser acupuncture; no passive controlled acupuncture RCTs were identified.