New Evidence In Favour Of Homeopathy
Two newly published studies (1,2) show that the conclusions of a review published in The Lancet about the effectiveness of homeopathy (3) were seriously flawed.
The review, which appeared in 2005, started with 110 placebo-controlled clinical trials of homeopathy matched with 110 of conventional medicine. Twenty-one homeopathy trials and 9 conventional medicine trials were of "higher quality". From that smaller set of trials, the authors then focused their attention on respectively 8 and 6 "large trials of higher quality" in several different medical conditions. Based on these 14 studies, the review concluded that there was "weak evidence for a specific effect of homoeopathic remedies, but strong evidence for specific effects of conventional interventions". The review was criticised at the time for its opacity, as it gave no indication of which trials were analysed nor the various assumptions made about the data. Sufficient detail to enable a reconstruction of the analysis was eventually published by the authors.
The new studies, which reconstruct the review's analysis, cast severe doubts on its negative conclusions, showing that it was based on a series of hidden judgments unfavourable to homeopathy. A positive conclusion for homeopathy is in fact obtained if all 21 "higher quality" trials are included in the analysis. If fewer than 21 trials are analysed, a positive or negative conclusion for homeopathy is crucially dependent on the exact number of trials selected. A firm positive conclusion is found, for example, merely by omitting four trials that showed Arnica is ineffective for muscle soreness after long-distance running.
Whether the conclusions of the analysis are positive or negative is therefore highly sensitive to the choice of which "higher quality" trials are included. The clear implication is that homeopathy for certain conditions is not placebo.