More on how to remove a Chinese IUD

Madam I am somewhat concerned that in the same issue of the Journal as the statistical paper entitled ‘Size does matter’1 there was an Overview on ‘Homeopathy treatments for premenstrual symptoms’2 which may be considered to have depended on evidence that contradicted the statistical paper. I would be grateful for your opinion as to the reliability of the one clinical trial by Yakir et al.3, which was cited as evidence in favour of homeopathy. I would like to make clear that my motivation is not to discredit homeopathy. 1. The Overview refers to the ‘recent’ trial by Yakir et al. It was actually carried out in 1992–1994 and not published until 2001. 2. The Overview refers to there being 20 women in the study but actually only 19 completed the study. 3. The Yakir et al. study claims 90% of the homeopathy group had improvement – actually it was 91%, i.e. even more favourable to homeopathy than was claimed. 4. Yakir et al. set an arbitrary improvement level of 30% of the menstrual distress questionnaire to determine effectiveness of therapy resulting in three of the placebo group improving. However, Figure 2 shows that five of the placebo group had improved. 5. While claiming homeopathy effective they admit under Outcomes that, ‘the between group difference ... fell just short of statistical significance’. 6. The study certainly suggests the possibility that homeopathy is effective so why no larger-scale trial 9 years later? My main concern is that our Journal has given credibility to the effectiveness of homeopathy in the management of premenstrual syndrome on the basis of inadequate evidence. What do you think?