Americans spend $34 billion a year on so-called alternative medicine—botanical pills, acupuncture, energy healing, and the like—despite the fact that few of these techniques are backed by any science. Study after study has rejected the ability of such treatments to cure. But the same studies routinely find that treated patients do wind up feeling better. For example, a randomized, controlled trial of Chinese herbs on women with ovarian cancer found no effective difference between the herbs and a dummy pill—because there was some improvement with both. A double-blind trial of saw-palmetto pills for men with enlarged prostates produced similar results. What gives?

The obvious answer is the placebo effect. We’ve known for decades that when sick people are given a treatment, even if it’s just a sugar pill, their condition often improves. But that can’t be the whole story, if only because the size of the effect varies wildly from one study to the next. One clue to a better answer is found in research led by Ted Kaptchuk at Harvard Medical School: Patients with irritable bowel syndrome were told they’d be participating in a study of the benefits of acupuncture—and one group, which received the treatment from a warm, friendly researcher who asked detailed questions about their lives, did report a marked reduction in symptoms, equivalent to what might result from any drug on the market. Unbeknownst to them, the researchers used trick needles that didn’t pierce the skin. 

Now here’s the interesting part: The same sham treatment was given to another group of subjects—but performed brusquely, without conversation. The benefits largely disappeared. It was the empathetic exchange between practitioner and patient, Kaptchuk concluded, that made the difference.

What Kaptchuk demonstrated is what some medical thinkers have begun to call the “care effect”—the idea that the opportunity for patients to feel heard and cared for can improve their health. Scientific or no, alternative practitioners tend to express empathy, to allow for unhurried silences, and to ask what meaning patients make of their pain. Kaptchuk’s study was a breakthrough: It showed that randomized, controlled trials could measure the effect of caring. But there was already abundant evidence from nursing science to suggest a healing power in the interaction between practitioner and patient. A study in Turkey found that empathetic nurses improved the symptoms of patients with hypertension. Midwestern cancer patients who received massages slept better and had less pain. …

Back in 2002, The New England Journal of Medicine published a paper showing that arthroscopic knee surgery for arthritis worked no better than a placebo, though 650,000 such operations were being performed a year. Knee doctors protested: They knew from personal experience that their patients felt much better after this procedure. These patients had come to them with their pain; the surgeons provided care in the form they knew best—surgery—and the patients’ pain decreased. But as with the saw palmetto and Chinese herbs, the surgery itself wasn’t the thing that helped. In the years since, other trials have confirmed the study’s findings and silenced the protests. It’s reasonable to think that the act of caring may be what led to the improvements.…

—Nathanael Johnson