Book Review: "Do You Believe in Magic?", by Paul A. Offit

Paul Offit’s new book, Do You Believe in Magic? The Sense and Nonsense of Alternative Medicine (HarperCollins, 2013) is a smoothly written, highly readable take on some important topics from the ever-changing complementary-and-alternative-medicine (CAM) territory. Offit’s primary theme is that treatments can harm as well as help—true of conventional medicine and its potential side effects, but even more true of unconventional treatments, with their untested effects. At several points in the book, Offit repeats the important dictum: there is really no such thing as “alternative” medicine, but only medicine that has been shown to work, effectively and safely,  and medicine that has not been shown to work or be safe. If it hasn’t been shown to be safe and effective, there’s no point in calling a treatment “medicine”.  

Offit begins the book with chapters considering two factors that help convince the public that an alternative treatment is a good idea: one is the belief that if a substance or procedure is part of an ancient tradition, it must be beneficial, and the other is the belief that whatever is “natural” is good for you.  (To these, I would add, by the way, the remnants of 19^th century Orientalism that suggest that whatever is done in Asia is a good idea.)

Obviously, these beliefs are unsupported by evidence. We read with horror of the devastation of the Black Death, of scurvy on sailing ships, of death and deformity from smallpox and polio. Strolling through a New England graveyard, we count infant deaths and note how many wives each man got through in those pre-divorce days. As for the goodness of the natural, we all must be aware of poisonous snakes and insects, of ergotism from contaminated grain, and of death cap mushrooms, deadly nightshade, oleander bushes, and the whole panoply of attractive toxins Ma Nature has put on the table. Yet, you can fool many of the people much of the time, and some of the people all of the time, if you just mention that shamans used to use a method or that a potion comes straight from nature’s bounty. Most curiously, a number of the foolees are quite convinced that they are cleverly avoiding the manipulations of the medical profession when they go for the old, the natural, the non-evidence-based. “Nobody can  fool me,” they think, and of course nobody needs to as long as they fool themselves.  

Most of Offit’s subsequent chapters recount sad stories of adults and children unhelped or even harmed by the treatments celebrities and “healers” sell to them. In the course of this discussion, Offit shows that snake oil proponents are not necessarily recycled used-car salesmen or preachers of the Elmer Gantry ilk. Stanislaw Burzynski, for example, is a Polish-trained physician who has been assistant professor at Baylor Medical College. He had a research Grant from the National Cancer Institute  and later FDA approval to test a cancer treatment, “antineoplastons”, extracted from that most natural substance, urine. The celebrity-expert, Suzanne Somers, says Burzynski is right about his treatment; unfortunately the evidence shows that despite his academic and research background, Burzynski’s claims are wrong.

Offit’s final chapter, “The Remarkable, Highly Underrated Placebo Response”, addresses the fact that people sometimes do get better when given alternative treatments. He refers briefly to several explanations of the placebo effect, including cognitive dissonance and regression to the mean. But what I’d hoped for from this chapter does not materialize. I would have liked to see some discussion of the effect of spending time with a patient and of personal interactions that may go a long way to explain the occasional effectiveness of implausible treatments.

When alternative practitioners say that their acupuncture or energy field methods have helped  people’s back pain (for example), it‘s not uncommon to hear conventional physicians say, “If I could spend an hour with each patient, I’d have more successes too”.  What happens during that hour of contact? It seems to me that this question, if answered carefully, could focus thinking on an important placebo factor, social and emotional interactions with another person.  

When researchers compare psychological interventions (alternative or conventional), they ideally design their studies so the targeted treatment is compared to another well-known treatment, rather than to no treatment at all. Unless the research question has to do with evaluating a brief method of treating a problem, as many features of the treatments as possible are matched, including the length of the treatment. Psychotherapies are aimed at making a person feel better, so all factors that could achieve this need to be taken into account. Although psychotherapists may employ different techniques, there are factors common to all these treatments that contribute a good deal to “feeling better”: the practitioner’s warmth, empathy, acceptance, and encouragement of risk taking. Presumably, the length of contact with each of these factors has an influence on patient changes.

Do those common factors also play a role in medical effectiveness? At one level, medical treatments  are aimed at creating physical changes that can be objectively measured--  lowering a fever, shrinking a tumor, bringing about weight gain or loss. But many medical treatments also deal with the “feeling better” aspect. If a person reports that he is still in pain or still hears ringing in his ears and does not “feel better”, presumably the entire medical task has not been accomplished.

Situations where the patient does not “feel better” in spite of treatment are situations that open the door to alternative practitioners. If the patient subsequently “feels better”, the alternative practitioner counts this as a success for his or her method and may report this or publish it in a journal dedicated to alternative medicine. But, in most comparisons of alternative and conventional treatments, the influential common factors go unconsidered. However difficult we might find it to assess warmth, empathy, and so on, it’s reasonable to say that less contact means less exposure to the common factors. Thus, research on alternative treatments, to be valid, needs not only to involve comparisons with conventional methods (rather than with “no treatment”), but to make sure that treatment plans equalize the contact time of patients and practitioners in the two groups.  Comparisons that indicate greater success for an unconventional treatment should not be accepted (by anyone) unless these rules are followed.  Otherwise, we are risking “believing in magic”.