Homeopathy — a response

I hope I may be allowed to reply to the several letters1^–4 commenting on my deliberately provocative personal column on homeopathy.5 All four authors assert their belief that homeopathy ‘works’, two of them making the claim that the fact that it works in babies and animals proves that this is more than a placebo effect. None of them cites any objective source of evidence for their beliefs, nor do they address the main point of my piece, which was to try to lay out the extraordinary, and to me still literally incredible, rationale that lies at the heart of homeopathic practice.

I agree wholeheartedly with Peter Hanrath and Andrew Hillam regarding the direction of much of our current target- and contract-driven practice, as I hope my more recent piece on statins illustrates.6 I have no quarrel with the use of complementary therapies per se, but I do think that such therapies should be subject to the same scientific scrutiny as is now expected of conventional therapies. As a novice in acupuncture I am well aware that much of its benefit is likely to be due to non-specific effects, and I don't agonise too much over its probable additional specific, neurologically-mediated mechanisms — but I welcome research that explores both these areas. It is my firm belief that the scientific approach can be brought to bear on the still mystifying power of such factors as suggestion, the personality of both doctor and patient, the nature of the relationship between them and so on.

Nigel Williams (in his letter)1 mentions a meta-analysis7 in the Lancet in 1997. More recently Bandolier8 published a ‘systematic review of systematic reviews’ of homeopathy and concluded as follows: ‘Much of the argument about homeopathy ends up being about trivial differences of little or no clinical relevance. Until large well-conducted trials tell us differently, the conclusion is that homeopathy does not work …’. A search of the Bandolier website9 leads to a number of commentaries on trials of homeopathy, not one of them showing any clear evidence of benefit. Clinical Evidence10 contains only one reference, that being a negative report of two RCTs for homeopathic treatment of warts. The Cochrane collaboration11 adds nothing further.

Finally — and hot off the press — Shang et al from Switzerland12 report on a comparison of 110 homeopathy trials and 110 matched trials of conventional treatments. They found insignificant evidence for a specific effect of homeopathic remedies, and strong evidence for specific effects of conventional interventions. They conclude that, ‘This finding is compatible with the notion that the effects of homeopathy are placebo effects’.

It is this absence of evidence for any specific effect of homeopathic remedies that led me to use the word ‘deception’ in the title of my column; it is the absence of harm, and the apparent non-specific beneficial effects of the homeopathic approach that made me qualify it as ‘benign’. None of your correspondents has convinced me that this is an unfair description.

• © British Journal of General Practice, 2005.