Conclusions
Homeopathy is an alternative therapeutic system based on the "Principle of Similars" and the use of "minimum" doses. Homeopathy was a prominent component of 19th-century health care and recently has undergone a revival in the United States and around the world. Despite skepticism about the plausibility of homeopathy, some randomized, placebo-controlled trials and laboratory research report unexpected effects of homeopathic medicines. However, the evidence on the effectiveness of homeopathy for specific clinical conditions is scant, is of uneven quality, and is generally poorer quality than research done in allopathic medicine (61). More and better research is needed, unobstructed by belief or disbelief in the system (62). Until homeopathy is better understood, it is important that physicians be open-minded about homeopathy’s possible value and maintain communication with patients who use it. As in all of medicine, physicians must know how to prevent patients from abandoning effective therapy for serious diseases and when to permit safe therapies even if only for their nonspecific value.
Author and Article Information
From Samueli Institute for Information Biology and Uniformed Services University of the Health Sciences, Bethesda, Maryland; Harvard Medical School, Boston, Massachusetts; and Centre for Complementary Medicine Research, Technische Universität, München, Germany.
LETTER
A Critical Overview of Homeopathy
Giovanni Federspil, MD; Fabio Presotto, MD; and Roberto Vettor, MD
21 October 2003 | Volume 139 Issue 8 | Page W-75
TO THE EDITOR:
Jonas and colleagues (1) sought to critically appraise homeopathy. This issue, along with the spread of other alternative medical therapies in western countries, is very important because it places all of scientific medicine in doubt (2). Jonas and colleagues’ paper only partially describes homeopathy—it does not define the historical origin of this doctrine and discusses only its pharmacologic and therapeutic aspects.
Homeopathy was born in the 18th and 19th centuries, when "medical systems," such as the doctrine of John Brown in England, dominated Europe. According to "medical systems," each pathologic phenomenon consists of 2 opposing concepts (for example, Brown believed that all diseases were "hyposthenic" or "hypersthenic") (3). The doctrine developed by Hahnemann was a medical system in which homeopathic and allopathic remedies were the opposing approaches. This distinction between allopathy and homeopathy now makes no sense because most drugs are given not to suppress symptoms but to break down etiopathogenic sequences. Hahnemann’s theory claims that all chronic diseases can be divided into 3 pathologic forms—psora, lues, and sycosis (which do not correspond to the current nosography)—that arise not from material causes but from a perturbation of the "vital spirit." Accordingly, drugs act not because of their material structure but because of their power to influence the vital force of the living organism. Many orthodox homeopathic physicians still support this theory.
Every critical analysis of homeopathy must consider that homeopathy is not a scientific theory but a metaphysical doctrine based on concepts that cannot be defined or proven experimentally (4). Clinical trials of the properties of homeopathic preparations cannot demonstrate the truth of homeopathy or the likelihood of its effectiveness since the doctrine contains metaphysical concepts (5). Thus, Jonas and colleagues’ paper cannot supply any evidence in favor of homeopathy.
LETTER
A Critical Overview of Homeopathy
Renan M.V.R. Almeida, PhD
21 October 2003 | Volume 139 Issue 8 | Page W-74
TO THE EDITOR:
Jonas and colleagues (1) forgot some important "symptoms" that a homeopath has to look for when considering a Pulsatilla prescription: "flatulence, no two stools alike, averse to fat, drink and warm food" (2) and "morbid dread of the opposite sex, religious melancholy, given to extremes of pleasure and pain" (2), together with dozens of other unrelated signs, supposedly generated in healthy people by this "preeminently female remedy, especially for mild, gentle, yielding disposition...crying readily; weeps when talking" (2). Thus, the clear inverse correlation between the quality of a homeopathic study and its positive result comes as no surprise. When independent replication is demanded, it is even less surprising to see the positive results drop to zero.
For instance, consider 1 of Jonas and colleagues’ examples, in which the findings of 1 study (3) were nonindependently replicated by a second (4). Homeopathy was prescribed for diarrhea; some of the "symptoms" involved were head sweats during sleep and the presence of 1 red cheek and 1 pale cheek. Other investigators (5) have detailed the many inconsistencies and errors in the first of these 2 studies. In fact, the investigators of the original study seem to have manipulated variables and end points in search of a significant P value (which they found: P = 0.048). For instance, in the original study (3), a "significant difference in the average number of stools by day 3" is considered a positive outcome. However, in the second study (4), the difference appears by day 5 (in the first study, this had yielded a negative result). Yet Jonas and colleagues consider these to be high-quality studies!
Finally, 1 of the authors had a $50 million budget solely for complementary and alternative medicine research. The authors’ statement that "more and better research is needed, unobstructed by belief or disbelief" (1) sounds like a cry for mercy for a pseudoscience disguised as an "alternative" therapy.
LETTER
A Critical Overview of Homeopathy
Flávio Dantas, MD, PhD; Peter Fisher, FRCP; and Hagen Rampes, MBChB, MRCPsych
21 October 2003 | Volume 139 Issue 8 | Page W-73
TO THE EDITOR:
Homeopathy has become increasingly popular with consumers throughout the world. It is also highly controversial. In this context, we welcome Jonas and colleagues’ critical overview (1). Such an overview is not an easy task because of the different operator-dependent techniques used in homeopathic practice.
We wish to point out some shortcomings in this overview. A comprehensive review of homeopathy should consider adverse effects. A systematic review done by 2 of us (2) showed that the mean incidence of adverse effects was greater with homeopathic medicines than with placebo in 11 controlled clinical trials (relative risk, 1.8) but that the effects were minor, transient, and similar in type in both study groups. A single trial of influenza prophylaxis heavily biased the results. If this trial were excluded, the relative risk would be 1.3. The main risks associated with homeopathy seem to be indirect, relating to the prescriber rather than the medicine. To our knowledge, no studies have adequately investigated this, and government regulation of homeopathic practitioners varies widely among different countries and legislatures.
Homeopathic pathogenetic trials or "provings" are said to be the basis of the knowledge of homeopathic medicines, but Jonas and colleagues do not mention that many of these volunteer studies were done in the recent past. A systematic review by 2 of us (3) concluded that there is a strong negative correlation between quality and the number of pathogenetic effects (that is, low-quality studies yielded more symptoms). In other words, more subtle symptoms, mainly psychological, could be false and thus could bias criteria for homeopathic prescriptions.
Finally, we highlight the question of sample size. Trials of homeopathy and other forms of complementary medicine are frequently relatively small and statistically underpowered. Unfortunately, perhaps because most scientists view homeopathy as having low "prior probability," they often interpret "absence of evidence" as "evidence of absence" of effect. For instance, in a recent study of the effect of homeopathic arnica, postoperative patients receiving placebo required 45% more analgesia than those receiving 1 of 2 homeopathic treatments (4). The study was interpreted as negative. The investigators did not report 95% confidence intervals, which is regrettable since the study was clearly underpowered to detect what was certainly a clinically relevant difference. The answer, of course, is larger, better-quality trials. But these also need larger budgets!