Homeopathic remedies came under the scrutiny of the medical establishment again recently and the positive conclusions were met with a predictably grudging response.

The results of a large, scrupulous meta analysis of homeopathy trials has concluded that the positive effects of homeopathy cannot be completely due to the placebo effect (Lancet, 1997; 350: 834-43).

The accompanying commentaries in the Lancet are an exercise in blinding readers with science. Reasons are given for any and all possible positive results noted in the meta analysis.

Criticism is levelled at the researchers for being too eager to demonstrate the robustness of their findings as if being meticulous was a sign of a guilty conscience; at homeopathy trials in general for investigating something which lacks a "credible prior hypothesis"; and at the lack of scientific credibility of the "infinite dilutions" theory, which is at the heart of homeopathy.

The results of the analysis, however, speak for themselves. They showed that trials proving the efficacy of homeopathy outnumber those that do not by nearly two and a half times. The authors admit that because their analysis does not provide evidence for any single homeopathic approach to any single condition it has no major implications for clinical practice.

What it does do is show that a serious effort to design and conduct proper research into homeopathy is greatly overdue.

They also note that what is seen by some as the implausibility of homeopathy should not be a barrier to such research.