The review concluded that the majority of adverse reactions to acupuncture treatment reported in trials of osteoarthritis of the knee are unspecific and not all reported adverse events should be attributed to acupuncture treatment. Due to lack of reporting of review methods, lack of validity assessment and paucity of study details the reliability of the authors' conclusions are uncertain.

To determine the safety of acupuncture for osteoarthritis (OA) of the knee.

PubMed and Japana Centra Revuo Medicana databases were searched in November 2006. Search terms were reported.

Randomised controlled trials (RCTs) evaluating acupuncture for OA of knee in comparison with sham or no treatment, and which reported information on adverse events, were eligible for inclusion.

The included studies evaluated acupuncture in comparison with: sham (including minimal acupuncture and dummy electroacupuncture); acupuncture and medication; or no treatment. The outcomes in the included trials were the number of patients experiencing adverse events or number of adverse events.

The authors did not state how papers were selected for the review, or how many reviewers performed the selection.

The authors did not state that they assessed validity.

The number of adverse events and number of participants were extracted. Only adverse events classified by the primary study authors as having been caused by the intervention were extracted.

The authors did not state how data were extracted, or how many reviewers performed the data extraction.

Data were reported in a table and briefly discussed.

Seven RCTs (n=1,533) were included in the review.

No studies reported serious adverse events. One crossover RCT (n=32) reported three patients with worsening pain, nausea or dizziness, one patient with a large haematoma (no sequela) and one patient with vaso-vagal attack in the acupuncture group, with no reported adverse effects in the no treatment control group. One RCT (n=91) found joint swelling in one patient in the acupuncture group compared with two patients in the dummy electroacupuncture group. Local contusion was also reported for 21 patients in the acupuncture group. One RCT (n=97) reported three participants with bruising at acupuncture site compared to none in the Streitberger needle group. One RCT (n=224) reported 18 small haematoma/bleeding events in the acupuncture group (n=150) compared to 9 in the minimal acupuncture group (n=74), six other events were also reported in each group (details not specified). One RCT comparing acupuncture (n=326) with minimal acupuncture (n=365) reported: 16 and 15 events respectively in each group for arthralgia; 10 events in each group for bone pain; 12 and 18 events respectively in each group for haematoma; 8 and 11 events respectively in each group for back pain; and 9 and 8 events respectively in each group for joint lock. One RCT (n=379) found no statistically significant differences in adverse events between acupuncture and sham acupuncture (data not reported). One RCT (n=19) had no reports of side effects.

The results found that the majority of adverse reactions to acupuncture treatment reported in RCTs of OA of the knee are unspecific and that not all reported adverse events should be attributed to the mechanism of acupuncture.

Inclusion criteria were clearly defined in terms of study design, intervention, outcomes and participants. Only two databases were searched and this may have resulted in the omission of other relevant studies. No attempts were made to minimise publication bias, and it is unclear whether any attempts were made to minimise language bias. Methods used to select studies and extract data were not described and so it is not known whether efforts were made to reduce reviewer errors and bias. Only RCTs were included but the quality of the included trials was not assessed, therefore results from these studies and any synthesis may not be reliable. No details were presented on the demographics of participants in the included studies, thus it is not possible to assess the generalisability of the review findings. Due to lack of reporting of review methods, lack of validity assessment and paucity of study details, the reliability of the authors' conclusions are uncertain.

Practice: The authors did not state any implications for practice.

Research: The authors stated that future research should include multiple methodologies to assess the safety of acupuncture including retrospective case reports for detection of rare side effects, large-scale prospective surveys to calculate incidence and RCTs or systematic reviews of RCTs to estimate the proportion of specific reactions that are associated with the mechanism of acupuncture.

Tsukuba University of Technology; Japan Acupuncture & Moxibustion Association; Japanese Association of Acupuncture, Moxibustion and Massage; Foundation for Training and Licensure Examination in Anma-Massage-Acupressure, Acupuncture and Moxibustion.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.