The Mysterious Placebo (Skeptical Inquirer)

John E. Dodes
One of the most significant but widely misunderstood phenomena is the placebo effect. Research shows that the placebo effect can be greater and is far more ubiquitous than commonly thought.
If it was so, it might be; and if it were so, it would be; but as it isn't, it ain't.
-Tweedledee, in Lewis Carroll's Through the Looking Glass

One of the questions that skeptics are asked most persistently is to explain how acupuncture, homeopathy, faith healing, Qigong, and other treatments work. Skeptics often use the placebo effect-a response to the act of being treated, not to the treatment itself-as an answer, but usually to no avail. I believe that's because most people, both logical and fuzzy thinkers, don't truly understand what the placebo effect is.
Spontaneous remission and the placebo effect, which are known as nonspecific effects, are significant phenomena that have great impact on consumers and health-care professionals. Recovery from illness, whether it follows self-medication, legitimate treatment, or avant-garde therapies, may lead one to conclude that the treatment received was the cause of the return to good health.
A common saying is that if you treat a cold it will last a week, but if you leave it alone it will be gone in seven days. Even serious diseases have periods of exacerbation and remission; arthritis and multiple sclerosis are prime examples. There are even cases of cancers inexplicably disappearing. The major logical error in plotting disease progress is: post hoc, ergo propter hoc ("after it, therefore, because of it"). This common fallacy credits improvement to a specific treatment just because the improvement followed the treatment.
H. K. Beecher's seminal paper "The Powerful Placebo" (Beecher 1955) is among the most frequently cited and was undoubtedly responsible for the double-blind study design having been adopted as the universal standard. Beecher reported on twenty-six studies and arrived at an average placebo response rate of 32.5 percent. From this figure comes the often cited statement that a fixed fraction (one-third) of the population responds to placebos. But this is a myth. A recent paper (Roberts et al. 1993) concluded that "under conditions of heightened expectations, the power of nonspecific effects (placebos) far exceeds that commonly reported in the literature."
The paper, "The Power of Nonspecific Effects in Healing," is fascinating. The authors analyzed data for treatments that met the following criteria:
(a) strong positive reports by at least two groups of investigators;
(b) at least one well-controlled negative report;
(c) the treatment had been abandoned as ineffective; and
(d) data from a major portion of the positive studies were presented in a manner that permitted reliable classification into categories of excellent, good, or poor outcomes.
The treatments that were studied included: glomectomy (the surgical removal of a small, normal mass of tissue called the carotid body that is found on the carotid artery) for the treatment of asthma; levamisole (an immunomodulatory drug) for the treatment of Herpes Simplex Virus (HSV); organic solvents (ethyl ether and chloroform), also for HSV; and gastric freezing for duodenal ulcers. In all the cases the doctors and the patients expected the treatments to work. The results for all the positive studies combined showed an excellent or good outcome in 69.8 percent of the almost seven thousand cases studied. These results led to the conclusion that a treatment outcome "is always due to some interactive combination of specific and nonspecific effects."
Research also illustrates how difficult it is to separate valid treatments from apparently valid ones. In other words, we're never without some level of nonspecific effects.
A number of other myths are associated with placebos. Try to answer the following questions:
Does a positive response to a placebo mean the patient's problem is imaginary?
Does a patient have to believe in the therapy for a placebo effect to occur?
Are placebos harmless?
The answer to all three questions is no. Placebo responses can occur in patients with real disorders; the subjective symptoms can resolve while the objective ones remain. Belief in the treatment only appears to explain a portion of the placebo effect (Jarvis 1990). It appears that belief, operant conditioning, and suggestibility all play important roles. In an interesting experiment, a man experienced pain and exhibited marked depression of a specific part of his heartbeat while being monitored by an electrocardiogram (ECG) machine during a treadmill diagnostic test. This occurred at a treadmill setting of 44. For a second test, when the treadmill number was miscounted so that the patient exercised less, he exhibited the same pain and ECG depression as at the setting of 44 (Jarvis 1990). This dramatically demonstrates the power of suggestion.

The Mysterious Placebo (Skeptical Inquirer January 1997)