Thirteen RCTs (n = 3,025) were included in the review. Eight trials reported allocation concealment, none reported blinding of clinicians and 10 reported blinding of patients. Sample sizes ranged from 30 to 1,039.
There was a statistically significant benefit for acupuncture over placebo acupuncture (SMD -0.17, 95% CI: -0.26, -0.08, p < 0.001). Significant statistical heterogeneity was detected, which was reduced following the exclusion of an outlier trial. The absolute difference between the groups was small, corresponding to 4 mm (95% CI: 2 mm, 6 mm) on a 100 mm visual analogue scale.
There was a statistically significant benefit for placebo acupuncture over no acupuncture (SMD -0.42, 95% CI: -0.60, -0.23, p < 0.001). Again, significant statistical heterogeneity was detected.
Results of sensitivity analyses were also reported but were not markedly different from those of the main analyses. Funnel plot analyses showed no evidence of publication bias.