Objective: This study sought a fair sample of published acupuncture clinical trials to assess whether they provided scientific rationales for their interventions, clearly differentiated their exposures or observed differences in outcomes.
Study design and setting: A systematic review of all controlled, clinical trials of acupuncture published in English in 2006 and indexed in PubMed.
Results: Seventy-eight acupuncture clinical trials met the screening criteria; 36 had some rationale. Twenty-two studies compared acupuncture to usual care alone or other non-acupuncture controls; 18 (82% of 22) had statistically significant differences in outcomes, but failed to control for placebo effects. Sixteen studies used placebo controls; 11 (69% of 16) had statistically significant differences in outcomes. Forty studies compared acupuncture interventions mainly differentiated according to traditional acupuncture theories; 19 (48% of 40) had statistically significant differences in outcomes.
Conclusion: Clinical trials demonstrate that acupuncture can affect outcomes and is distinguishable from a placebo. However, trials that compare acupuncture exposures often observe no statistically significant differences in outcomes. Traditional acupuncture theories, for selecting points, locating them on the body and choosing appropriate stimulation, appear to be unreliable for creating distinct exposures.