Henry K. Beecher chose "The Powerful Placebo" as the title for his influential paper published in 1955 in the Journal of the American Medical Association (vol. 159, pp. 1602-1606). In it he recounted the effects of placebos in 15 clinical trials involving a variety of diseases. Improvement ranged from 21 to 58 percent, the average being 35 percent. This figure has since been foolishly transmuted into a magical percentage of improvement to be expected from a placebo in any disease.
Recently, in the Journal of Clinical Epidemiology (1997, vol. 50, pp. 1311-1318), Kienle and Kiene have criticized Beecher's paper, concluding that "contrary to his claim, no evidence was found of any placebo effect in any of the studies cited by him," and listing a number of factors that could account for the reported improvement that had nothing to do with psychological anticipation.
The Powerful Placebo discusses the placebo effect over the centuries, reminding the reader how complex the issue is, from the very definition of a placebo and the success of dubious or fraudulent remedies to the modern worship of placebos as controls in randomized clinical trials. The authors assert that "until recently, the history of medical treatment was essentially the history of placebo effect." (Cinchona is given credit as a legitimate ancient remedy, but one looks in vain for Withering and digitalis, and opium is grudgingly dismissed as a treatment whose "usefulness tends to be exaggerated.")
The authors are more negative than some of us about acupuncture and traditional folk remedies. Given the richness of active chemicals in the plant and animal kingdoms, surely some correct links have been identified by trial and error over the centuries. Indeed, some folk remedies have been studied with success in recent decades, and they add to the list of such useful effective plant derivatives as reserpine and curare.
There is little evidence that modern physicians (at least in the United States) dispense lactose pills or saline injections with any frequency; it is all too easy to prescribe a drug approved by the Food and Drug Administration as safe and effective that could conceivably alleviate the patient's symptoms. Penicillin might be termed a placebo if given to a patient with a viral sore throat, but not, I would submit, if given to a patient with a sore throat of uncertain cause.
Today in developed countries, placebos are important primarily as controls in clinical trials. For this purpose -- the equivalent of the chemist's "blank" -- it is not crucial whether improvement after placebo is due to spontaneous change or suggestibility; the "active" treatment has to provide superior performance. It is usually impossible to dissect the reasons for what is observed after placebo, because to do so requires a trial group in which randomized patients receive neither active drug nor placebo. I have conducted only one such trial, and in it the placebo "benefit" was explained entirely by spontaneous events. Without such information, one is not likely to be able to distinguish the characteristics -- experiential and psychological -- that predispose people to anticipate positive or negative effects after the demonstration by a health care provider of therapeutic intent.
The authors devote more than half their book to the research implications of the placebo. They conclude that currently "it is impossible to control nondrug factors without the use of... placebo controls." This position ignores the traditional regulatory approval of antibiotics on the basis of comparisons with a standard antimicrobial agent, the use of dose-response curves (without placebos), the growing tendency to question the ethicality of placebo controls in diseases for which adequate (if imperfect) remedies exist, and the Helsinki Declaration's insistence that all subjects -- including those in control groups -- must "be assured of the best proven" treatment, a requirement that, taken literally, would make all new therapeutic-product research impossible except for diseases for which no remedy exists. (Remedial language for the Helsinki Declaration has been drafted and is currently going through the World Medical Association process for change.)
Psychotherapy and psychoanalysis are subjected to lengthy and penetrating critiques by the authors, but one misses any reference to the work of Ader and others on Pavlovian conditioning with active treatments and placebos in both animals and humans, and the case reports of addiction to placebos (O. Vinar, "Dependence on a Placebo: A Case Report." British Journal of Psychiatry, 1969, vol. 115, pp. 1189-1190; I. Mintz, "A Note on the Addictive Personality: Addiction to Placebos." American Journal of Psychiatry, 1977, vol. 134, p. 327).
Reviewed by Louis Lasagna, M.D.
Copyright © 1998 Massachusetts Medical Society. All rights reserved. The New England Journal of Medicine is a registered trademark of the MMS.
