A J Vickers1. 1. Royal London Homoeopathic Hospital, United Kingdom. Vickers@mskcc.org
Abstract
OBJECTIVE: Homeopathy is controversial, primarily because of the use of medicines diluted beyond the Avogadro limit. This article examines the scientific debate on whether homeopathy can have effects greater than placebo in humans. METHODS: Five rigorous English-language clinical studies published in high-impact journals that favored homeopathy were identified. Letters and other articles written in response to these articles were then retrieved and analyzed. RESULTS: Much of the content of responses to positive homeopathic research was rhetorical in nature and antagonistic in tone: homeopathic researchers were accused of bias and of being in the pocket of homeopathic manufacturers; arguments were raised that are demonstrably inaccurate or irrelevant to the interpretation of research; words such as "magic" or "potion" were used to paint homeopathic research as inherently unscientific. Other commentators argued that based on theoretical grounds homeopathy could not possibly be effective, any positive trial result must be due to bias. Surprisingly few responses raised substantive methodologic or statistical issues. Many of these are clearly inaccurate or irrelevant. The possibility of publication bias-selective publication of positive results-was raised by some commentators and remains an important criticism of the apparently positive nature of the clinical trial evidence. The most persuasive argument against accepting differences between homeopathy and placebo on the basis of current evidence is that homeopathy is scientifically implausible and so requires more evidence than normal. CONCLUSIONS: Investigators undertaking clinical research in homeopathy need to be responsive both to the dangers of publication bias and to the requirement for stronger than usual levels of evidence.
OBJECTIVE: Homeopathy is controversial, primarily because of the use of medicines diluted beyond the Avogadro limit. This article examines the scientific debate on whether homeopathy can have effects greater than placebo in humans. METHODS: Five rigorous English-language clinical studies published in high-impact journals that favored homeopathy were identified. Letters and other articles written in response to these articles were then retrieved and analyzed. RESULTS: Much of the content of responses to positive homeopathic research was rhetorical in nature and antagonistic in tone: homeopathic researchers were accused of bias and of being in the pocket of homeopathic manufacturers; arguments were raised that are demonstrably inaccurate or irrelevant to the interpretation of research; words such as "magic" or "potion" were used to paint homeopathic research as inherently unscientific. Other commentators argued that based on theoretical grounds homeopathy could not possibly be effective, any positive trial result must be due to bias. Surprisingly few responses raised substantive methodologic or statistical issues. Many of these are clearly inaccurate or irrelevant. The possibility of publication bias-selective publication of positive results-was raised by some commentators and remains an important criticism of the apparently positive nature of the clinical trial evidence. The most persuasive argument against accepting differences between homeopathy and placebo on the basis of current evidence is that homeopathy is scientifically implausible and so requires more evidence than normal. CONCLUSIONS: Investigators undertaking clinical research in homeopathy need to be responsive both to the dangers of publication bias and to the requirement for stronger than usual levels of evidence.