Letter to the editor

The authors deserve recognition for their effort to do a welldesigned trial in the contested field of homoeopathy. However, for the following reasons we doubt that the effectivity of homeopathic treatment is supported by this study: 1. The exclusion of a quarter of all eligible study probands after they failed to show any improvement in the initial open trial phase does not correspond to usual practice and limits the generalisability of the results of the trial. 2. The worsening of symptoms in the double-blind trial phase under verum in arm A contradicts the hypothesis of the effectivity of homeopathic treatment. This lack of effect cannot be ascribed to the expectations of parents, as the authors argue. 3. The slight improvement of symptoms under verum in arm B in the second phase is with 1.67 negligible compared to the improvement of nearly 10 points seen in the open pre-trial phase (CGI scale). In our opinion, this points to the ineffectiveness of the treatment. This is all the more true as the p-value is only around .05, from the testing of a rather large number of scales. The sizeable improvement in the open after-trial phase points to a large effect of the doctors' encouragement. In our view, the overall picture emerging from this study is most simply interpreted assuming no intrinsic homeopathy effect, but rather a pronounced effect of the medical treatment procedure on the family system. Eur J Pediatr (2007) 166:509 DOI 10.1007/s00431-006-0262-5