Medical Examiner

It May Be Fake, but Trust Me—It’ll Work

When is it kosher for doctors to prescribe placebos?

Can placebos possibly work?

Pediatrician Adrian Sandler works at a busy referral center in Asheville, N.C., where he often treats kids with ADHD. Many parents worry about the long-term effects of stimulants, the standard treatment for the disorder, but when kids quit or cut back, their symptoms often spiral out of control. So a few years ago, Sandler decided to replace some of the medicine in each of the doses he gave out with a green-and-white placebo capsule that contained no medication. By pairing the placebo with a stimulant, he hoped to spark a conditioned response, just as Russian Nobel Laureate Ivan Pavlov had done with his famous dogs. Where Pavlov had trained his dogs to salivate at the sound of a buzzer, Sandler hoped to make kids with ADHD calm down in response to a fake pill.

Plenty of doctors already give placebos as medicine. In a survey of 679 physicians published in 2008, more than 46 percent of respondents said they regularly prescribe placebos, usually in the form of a relatively harmless drug that they know has no effectiveness for the patient’s complaint. Bioethicists generally frown on such deceptions, out of concern that these lies will erode the patient-physician relationship. No one goes to the doctor for a kiss on her boo-boo; people go to doctors with the trust that they’ll offer an effective treatment. If you find out that the “drug” you were prescribed was actually a sugar pill, this trust is undermined, and that’s why American Medical Association guidelines forbid giving placebos without a patient’s consent.

But Sandler figured he could get around the ethics problem by telling the parents and kids up front that they were getting a placebo, with the promise that this placebo-drug combo had the potential to control their symptoms just as well as the full dose, but with fewer side effects. He tested the idea in 99 children, randomly assigned to one of three treatments. The first group continued to take their regular dose of medication, another took half their optimal dose, and the third was instructed to take a half dose of their meds plus a placebo pill that was described as a “dose extender.” Before the study began, researchers explained to the parents and the kids, aged 6 to 12, that the dose extender contained no active ingredient. After eight weeks, the symptoms of ADHD had grown more severe in kids who took only a half dose, but they remained stable in the groups that received either the full dose or the half dose plus placebo.

The ADHD study, which was published last summer, isn’t the only one to show a benefit from giving placebos without pretending that they’re real drugs. A trial that came out in PLoS ONE last December found that patients with irritable bowel syndrome who were treated with placebo pills for three weeks and told that the pills were “made of an inert substance, like sugar” achieved greater improvements in their symptoms than those who received no treatment. In fact, the placebo’s effectiveness in this study rivaled that of other drugs used to treat the condition. With ethical problems solved, placebos might thus serve as part of a “wait and watch” approach to treating irritable bowel syndrome and other problems. Doctors could hold off on doling out potent drugs while still offering their patients a tangible treatment. And placebos don’t just have fewer side effects than standard pharmaceuticals, they presumably cost less, too.

While Sandler and Ted Kaptchuk, lead author of the irritable bowel syndrome study, say that while their so-called “open placebos” don’t hinge on deception, they do employ some sleight of hand. A doctor who uses an open placebo is like a magician. The trick is performed with full disclosure that it is, in fact, a trick, but it still requires a subtle form of deception to execute. For the placebo to work, the patient must suspend disbelief at the doctor’s urging. Kids in the ADHD study were told that, “the mind and body work together in interesting ways and placebos are known to work sometimes but no one knows why,” while researchers told patients in the IBS trial that placebos “have been shown in clinical studies to produce significant improvement in IBS symptoms through mind-body self-healing processes.” While some might argue that it’s not lying to tell patients the placebo might help if that’s what the doctor really believes—and there is ample evidence that placebos often work—even “truthful” manipulations like this raise ethical questions. Deception is verboten in medicine because it violates a patient’s trust, and shaping a patient’s expectations exploits that trust.

Yet doctors do this every day, often to the benefit of their patients. Randy Gollub, a psychiatrist at Massachusetts General Hospital in Boston tells of a colleague who treated a man for gout. The patient had tried numerous drugs without relief, and the sole remaining drug had a 50 percent chance of helping. Instead of saying, “This drug is your last resort, but it only works about half the time,” he told the patient, “This drug doesn’t help everybody, but it helps some people and I think it could help you.” Both statements are true, she says, but the second description was more likely to elicit a beneficial response.

Such framing sounds innocuous, but how far can it go before it becomes unethical? The ADHD and irritable-bowel data show that you can increase a patient’s chances of being helped by telling him that he’s likely to respond to a placebo pill. But you can also increase the benefit by asking him to pay some extra money, since studies show that expensive placebos work better than cheaper ones. And more invasive placebos—sham injections, say—are known to have more of an effect than inert pills or ointments. A doctor who really wants to help a patient might even offer a fake surgical procedure, in which an anesthetic is given and a cut is made, but nothing else. These procedures have rivaled the effectiveness of real surgeries in at least one clinical trial, but they’re far more costly than a sugar pill and, given their reliance on knives and anesthetics, they impose their own risks.

Judging a placebo treatment’s ethics solely by its use of deception—and viewing the limits of deception in rather legalistic terms (well, we didn’t exactly lie to the patient …)—overlooks important nuances. If you told a patient to expect some kind of mind-body benefit from a sham surgery, that would be less deceptive than a sugar pill given surreptitiously. But judge these treatments by their effect on the patient’s welfare, and the pill might come out ahead. If deception is the critical factor from the perspective of bioethics, how do you know you’ve truly eliminated it? Do people who receive a placebo under a doctor’s authority, with the expectation that this doctor is administering care, truly understand that the placebo has no active ingredient? Placebos may have a place in modern medicine, but many questions need answering before they become standard treatment.