Methodological Issues in Trials of Acupuncture

To the Editor: Dr Shen and colleagues¹ conclude that electroacupuncture helps control chemotherapy-induced emesis. I believe that the only legitimate conclusion of their study is that more intensive placebo is more effective than less intensive placebo. There are 2 flaws in their study design, both related to blinding. First, despite the authors' claims to the contrary, their subjects seem not to have been fully blinded to the procedure. Those receiving electroacupuncture had a more complex intervention, and this alone could have produced a greater placebo effect than that received by those with minimal needling. The group receiving electroacupuncture had more manipulation and experienced more sensation and discomfort during the procedure and was more likely to report having received excellent quality acupuncture. This suggests (although not statistically significantly, perhaps because of small numbers) that these patients were more likely to think they were getting "real" acupuncture and, perhaps, that they were not naive recipients of acupuncture and knew whether they were receiving "real" treatment. Second, the acupuncturists were not blinded to the procedure type they were administering; their attitudes could well have been conveyed to the participants. Acupuncturists could have been blinded by a variety of means, including training previously untrained individuals specifically for this process or not telling participating clinicians the study design or particular desired outcomes.