Commentary on rampes et al ‘Does electroacupuncture reduce craving for alcohol? A randomized controlled study’

Abstract

Dr Rampes and co-workers cite our criteria-based meta-analysis on the effects of (electro)acupuncture that was published 6 years ago. 1 It showed, especially for studies on smoking cessation, that good study design was associated with negative study outcome? In this critical commentary, we show that a negative study outcome (compare groups 1 and 2 in Rampes et al's study 1) does not always indicate good study design. Table 1 reproduces the criteria for good methodology that we used in 1990, showing the maximum score and the actual score for the study by Rampes et al? It is clear from the total score (at most 55 out of 100) that there would have been room for improvements. The study illustrates that the designation of study outcomes as positive or negative on the basis of statistical significance (P < cQ can sometimes be arbitrary in more than one way. First, even if we focus only on the outcomes for craving at week 8, the study is negative if we compare groups 1 and 2, but it is positive if we compare group 1 (or 2) with group 3. Second, let us not forget that the (orthodox) choice of an o~-level of 5% is arbitrary. Recently, one of us (AGHK) designed a list of criteria that enables one to judge the usefulness and the quality of clinical trials in a much more consistent way than the one we used in 1990. This list is presented in Table 2, which also illustrates how we (re)assessed the study by Rampes et al. 1 We shall mollify our somewhat reductionist approach by commenting briefly on the rationales behind our judgments. Suppose we are 'evidence-based' practitioners contemplating the use of auricular electroacupuncture (AEA) as an adjunct to conventional treatment for