Lying about Placebos: Does the Placebo Effect Work If Patients Know What They Are Taking?
Justman, Stewart, Skeptic (Altadena, CA)
IN 1955 HENRY BEECHER PUBLISHED HIS LANDMARK paper, "The Powerful Placebo," demonstrating that approximately 35% of a given treatment group will respond to placebo as to an active medication (a figure that does not take into account competing explanations for reported improvement, such as spontaneous remission). (1) Beecher's interest in placebos went back to his experience as an army doctor in World War II, when, lacking morphine to treat the wounded, he was reduced to injecting them with water.
A decade later Beecher published a paper documenting and deploring the practice of experimentation on patients without their knowledge. (2) This paper too proved historic, as it helped inaugurate the era of informed consent in which we now live. But if much has changed in consequence of the bioethics revolution, some things remain the same. By general agreement, the most promising use of placebos is in pain management, but in order to arouse expectation and exert their effect placebos, it seems, need to be mistaken for active drugs by the user, much as Beecher's wounded mistook water for morphine. Necessarily, therefore, much experimentation with placebos still entails an element of deception. Though subjects now give their consent, most placebo research (as opposed to a clinical trial employing a placebo as control) is in fact so dependent on deception that it could not be translated into medical practice.
It is sometimes said that subjects do not really need to be deceived in order for placebos to work their wonders. In a notable 1980 paper in the Journal of Family Practice, Howard Brody and David Waters claim that "Even when patients are informed of the inert nature of the placebo, they may respond positively." (3) Their authority for this statement was a study conducted in 1965 which employed no control group and has not been replicated. (4) More recently David Jopling, investigating the possibility of open placebos, offered the guarded conclusion that "While this has not been the subject of much research, there is some clinical evidence to suggest that patients who are informed that they are receiving saline injections, sugar pills or other placebos sometimes continue to experience measurable objective symptom relief." (5) The evidence referred to turns out to be the same unreplicated 1965 study criticized for "a small patient sample, questionable symptom matches and comorbidity profiles between patients, an overly short treatment course, no wash-out period for potentially confounding psychoactive medications taken by patients, and no control groups (e.g. a no-treatment group)" by Jopling himself. (6)
Recently one of the foremost placebo researchers, Fabrizio Benedetti, has reported that athletes treated with morphine while in training but with placebo on the day of competition respond to the placebo as if to the drug, a carry-over that raises the possibility of achieving "drug-like effects without drugs" in real-world conditions. However, the experiment on which this finding rests hinges on a deception. (Significantly, the athletic competition itself seems to have taken place in a laboratory.) Far from receiving truthful information, placebo groups were told on the last day that they were getting morphine and should expect an increase in pain tolerance--a strong message indeed. (7) It does not seem credible that athletes in the real world, looking to circumvent doping regulations, would train with morphine the better to fool themselves into mistaking placebo for morphine on the day of competition. Elsewhere Benedetti himself cites an impressive body of research showing that expectation strongly influences placebo analgesia, which is to say that placebos engender analgesic effects in good part because we mistake them for--and so expect them to behave like--active drugs.
Placebos are known to score high in clinical trials of antidepressants. …