Authors’ response: we must distinguish symptoms caused by the medicine from other symptoms

Quirk and Sherr raise a number of important and controversial issues concerning our systematic review of Homeopathic Pathogenetic Trials (HPTs, provings) which we will attempt to address.[ ¹ ] Our ‘proposition’ that ‘the theory linking symptoms detected through HPTs in healthy volunteers to those treated in the sick is wrong’ is a hypothesis, a tentative explanation of the observed facts, and is not new. It is established that many symptoms in homeopathic materia medica derive from clinical observation, rather than HPTs. The non-systematic clinical verification of symptoms reported in HPTs is prone to several errors, such as focusing on the numerator and forgetting the denominator. The numerator may include hundreds of ‘documented cures’, if the medicine has also hundreds of symptoms, is frequently indicated for common symptoms/complaints and well represented in electronic homeopathic repertories. But what is the size of the denominator, how many cases were treated using the medicine without favourable results?[ ² ] Campbell showed, in the early 1980s, that, for instance, if we prescribed Lycopodium solely on the basis of its provings, we would use it for a few, rather non-descript, gastrointestinal symptoms.[ ³ ]

Quirk and Sherr assert that 1945–1995 is ‘well known as the weakest period of homoeopathic provings’. We disagree: how can this be ‘well-known’ when ours is the first published attempt to study systematically the methodological quality of HPTs in that or any other period? The Methodological Quality Index used in our study includes randomization and blinding, traditionally used in medical studies, to reduce the effects of suggestion and selection bias. Quirk and Sherr complain that we did not extend our review by a further decade, to 2005. Ideally we would have done so, but this large international collaborative project took nearly a decade to complete, its first results were published in 1998.[ ⁴ ] To extend the inclusion period to 2005 would have further delayed publication.

Our systematic review showed a high incidence of pathogenetic effects reported by volunteers in HPTs which could be attributable to design flaws: more effects per volunteer were noted when the methodological quality of the trial was low. In conducting HPTs researchers need to discriminate the signal (symptoms caused by the substance being tested) from the noise (confounding factors such as the myriad events, incidents and spontaneous changes of daily life, and the symptoms and sensations related to them). In analysing HPTs the critical task is to separate the signal (changes caused by the medicine) from everything else.

For us the point of an HPT is to identify true symptoms caused by a potential medicine in healthy volunteers. This can be expressed in a simple formula: Symptoms (with medicine) − Symptoms (without medicine) =Symptoms (due to medicine).

The same logic is applied in clinical trials of new treatments, but in an inverse way (suppression of symptoms). Time is a critical factor: one cannot take an action (for instance take an homeopathic medicine), observe what happens, then go back in time and observe what would have happened if you had not taken the medicine. The closest we can come to this is to observe what happens in different people at the same time (parallel group method) or the same person at different times (crossover), or combinations of these. The use of intra-individual placebo control (crossover design) helps but does not guarantee 100% precise results. Other innovative strategies are in development.[ ^5,6 ] Above all we need adequate mechanisms to cancel out ‘noise’ in HPTs to accept their results as valid and useful.

We do not advocate a methodological straitjacket to restrain innovation in HPTs but innovations must have a rational basis. Several biases, taken together, could contribute to the high percentage of volunteers presenting symptoms. Our results suggest that, in many cases, some or all of these symptoms would have occurred even if the volunteer had not taken the medicine. In terms of accepted knowledge, it is difficult to justify attribution of a symptom such as “got sunburn mainly on the left side of the back, although the whole of him was in the sun”, reported more than 3 months after the initial dose of Hydrogenium, as in Sherr's report.[ ⁷ ] Repertories boast of the number of new symptoms they have incorporated, but never of the number of redundant ones they have eliminated, although this is necessary.[ ⁸ ] Quirk and Sherr may wish to suggest other criteria by which to judge what is a good HPT, but these should be justified in terms of the model they propose.

We agree with Signorini's statement that ‘it remains very uncertain that HPTs yield valid results’: much of homeopathic knowledge in materia medica does not come from HPTs and needs a critical review. To claim that Signorini's statement amounts to saying that ‘none of our remedies work in practice’ is a gross overinterpretation.

There are other important aspects to be explored in future studies, such as the sensitivity of individual volunteers (those volunteers who react strongly to a particular medicine may be the ‘constitutional type’), the relationship between toxicity of the medicine and number of effects, the effects of different routes of administration, and, crucially, innovative methods to separate symptoms truly related to the medicine from those which would have occurred even if it had not been taken. We hope that all these aspects will be addressed in the future and that our systematic review will be of value to those researching these subjects.

• References

• [1] Dantas F, et al. A systematic review of the quality of homeopathic pathogenetic trials published from 1945 to 1995. Homeopathy 2007; 96: 4–16.
• 2 Stolper C.F., Rutten A.L.B., Lugten R.F. Barthels. Improving homeopathic prescribing by applying epidemiological techniques: the role of likelihood ratio. Homeopathy 2002; 91: 230-238.
• 3 Campbell A. Lycopodium from provings. Br Hom J 1981; 70: 94-99.
• 4 Dantas F, Fisher P. A systematic review of homoeopathic pathogenetic trials (‘provings’) published in the United Kingdom from 1945 to 1995. In: Ernst E, Hahn EG (eds). Homoeopathy: A Critical Appraisal. London: Butterworth-Heinemann, 1998, p. 69–97.
• 5 Dominici G., Bellavite P., di Stanislao C., Gulia P., Pitari G. Double-blind, placebo-controlled homeopathic pathogenetic trials: symptom collection and analysis. Homeopathy 2006; 95: 123-130.
• 6 Walach H., Sherr J., Schneider R., Shabi R., Bond A., Rieberer G. Homeopathic proving symptoms: result of a local, non-local, or placebo process? A blinded, placebo-controlled pilot study. Homeopathy 2004; 93: 179-185.
• 7 Sherr J. The Homoeopathic Proving of Hydrogen. Malvern: The Dynamis School of Homoeopathy; 1993, no ISBN.
• 8 Rutten A.L.B., Stolper C.F., Lugten R.F., Barthels R.W.J.M. Repertory and the symptom loquacity: some results from a pilot study on likelihood ratio. Homeopathy 2004; 93: 190-192.