[Proving the effectiveness of complementary therapy. Analysis of the literature exemplified by acupuncture]

At present, acupuncture may be considered one of the most popular forms of complementary medicine worldwide. However, in relation to the number of reviews on the subject, comparatively few controlled clinical trials have been reported so far. An analysis of all the controlled clinical trials listed in MEDLINE between 1987 and March 1994 (n = 39) that met certain basic requirements revealed that they addressed a wide variety of diseases and/or symptoms with no major focus (apart from the symptom of pain, which of course is highly complex in nature). In agreement with the findings of other meta-analyses, most of the more recent papers have been found to be still of indifferent quality. Besides the inherent problem that the term acupuncture subsumes within itself a substantial number of different techniques (and even philosophies), an obvious methodological deficit can be observed. Many groups seem to attach too little importance to choosing an appropriate control model, although seminal papers addressing this problem were already published in the early eighties. Similar remarks apply to inadequacies in study design, which should be at least single-blind. In summary, these findings may well help to explain why the effectiveness of acupuncture has still not been definitively demonstrated.