It states that the reason for the decision was the lack of evidence that homeopathy is clinically effective. "This follows a careful review of the scientific literature on clinical effectiveness and extensive discussions with local consultants and GPs."

The document then goes on to raise issues about the methodology used in research to determine the effectiveness of treatment. In particular it is concerned that placebo or non-specific factors, including the therapist's interpersonal skills, could have an effect on patient outcomes. "Non-specific factors are not reproducible", it states, "and the HA does not consider that purchasing a service which only produces non-specific effects is a good use of resources."

But, as Peter Fisher, Clinical Director of the Royal London Homoeopathic Hospital, points out, the main case for homeopathy is built precisely around the principle of separating the specific and non-specific factors through the use of double-blind, placebo-controlled trials. 200 of these trials of homeopathy have now been carried out.

In a section which purports to look at the methodological deficiencies of research on homeopathy, the first point raised is about the underlying principles of homeopathy, especially potentiation, which the document states "are unacceptable to most other scientific disciplines".

Dr Fisher says that it would be more accurate to say that the underlying principles "are not understood and are therefore rejected by most doctors".

"The information medicine hypothesis, which could explain how potentiation works, proposes that homeopathic medicines store information in a way similar to a floppy disk. It can then transmit this information into a pre-sensitised (sick) biosystem. Physicists are much less phased by this hypothesis than physicians. It is also getting considerable support from physical theory¹⁰,¹¹,¹²."

The second methodological point raised by the Lambeth, Southwark and Lewisham document is about a meta-analysis published in 1991 by Kleijnen and colleagues¹³, which, it claims, found "that the evidence for homeopathy is not sufficient to draw definite conclusions because most trials are of low methodological quality".

In fact, the document incorrectly quotes Kleijnen's conclusion, which stated: "The evidence is positive, but not sufficient...". By leaving out the first four words, the tone of the quote is completely altered, and misrepresents the author's findings.

The Kleijnen study has now been superseded by a high-quality meta-analysis by Klaus Linde and colleagues¹⁴ published in The Lancet in 1997. Linde's conclusion is carefully, if rather obliquely, worded: "The results of our meta-analysis are not compatible with the hypothesis that the clinical effects of homoeopathy are completely due to placebo". But a reading of the study shows a clearly positive result for homeopathy. 42% of the trials were clearly positive and a further 39% showed a positive trend. Unfortunately, this study, too, has been misrepresented and used against homoeopathy. (See the Bandolier case following.)

Two other meta-analyses published since the Kleijnen paper, one on homeopathy as a whole¹⁵ and the other on post-operative ileus¹⁶ have also provided positive results for homeopathy.

The final point on methodology raised by the document is concern about the lack of "high quality evidence that homeopathic treatment is better than, or equivalent to, standard medical treatment".

While it is true that there are not many trials comparing homeopathy with conventional treatment, Peter Fisher says the document is an example of the way opponents of homeopathy keep moving the goal posts.

"Until recently, it was claimed that there is no evidence that homeopathy has any specific effect at all, and that trials should therefore compare it with placebo. This document, however, contends that homeopathy does not have specific effects, but at the same time demands that it is compared with treatment known to be active."

Another problem with comparing homeopathic and conventional treatments is that it may not be comparing like with like.

"Reilly's work on hay fever¹⁷,¹⁸, for instance, shows that homeopathy has a long-term effect and can be safely used in a general practice setting," says Dr Fisher. "There is no current conventional therapy which has a similar profile in terms of duration of action and safety.

"Homeopathy is often intended to improve quality of life and reduce the number of exacerbations and drug requirements. Our research shows it is often effective in this, but then the effect being sought is not the same as in conventional treatment which is usually targeted at specific symptoms."