They are common, diverse and powerful and they raise an interesting ethical question - can doctors justifiably prescribe placebos to their patients? The standard answer is no. Doing so patronises the patient, undermines their trust, and violates the principles of informed consent. It compromises the relationship between doctor and patient. At worst, it could do harm.
But many of these arguments are based on the idea that placebo effects depend on belief; people must expect that treatments will work in order to experience any benefits. For a doctor to prescribe a placebo, they'd need to deceive. But according to Ted Kaptchuk from Harvard Medical School, deception may not be necessary. In a clinical trial, he found that patients with irritable bowel syndrome (IBS) felt that their symptoms improved when they took placebo pills, even if they were told that the pills were inactive.
Fabrizio Benedetti, a placebo researcher at Turin Medical School who wasn't involved in the study, says, "Although several studies suggested that placebos can be equally effective without deception, this is the first rigorous study that provides scientific evidence for this."
Referring to an earlier study published in the British Medical Journal, he says, "We did the study because we knew that physicians were giving placebo to patients secretly without informed consent. Our study was designed to test whether placebo effects could be harnessed without this secret deception."
He recruited 80 patients with IBS and randomly split them into two groups: one who took a placebo pill twice a day, and another who didn't take any treatments. Throughout the study, Kaptchuk's group were honest and open about the nature of the trial. Every patient had a detailed consultation with one of the team, who explained that placebo pills were "inert or inactive pills, like sugar pills, without any medication in it."
They told the patients that "placebo pills, something like sugar pills, have been shown in rigorous clinical testing to produce significant mind-body self-healing processes." And they explained: that "the placebo effect is powerful; the body can automatically respond to taking placebo pills like Pavlov's dogs who salivated when they heard a bell; a positive attitude helps but is not necessary; and taking the pills faithfully is critical."
Before the trial, the patients answered several questionnaires about the severity of their symptoms, whether they had recently improved or worsened, whether they experienced any relief from their symptoms, and their quality of life. After 21 days, they answered the same questions again. Kaptchuk found that those who had been taking placebo pills felt better than those who didn't take anything. They reported more relief and larger improvements in their symptoms, which had become less severe. Only quality of life was not significantly different between the two groups.
This isn't entirely new. In 1965, Lee Park and Lino Covi asked 15 neurotic patients at a psychiatric clinic if they wanted to try a sugar pill that could help them, even though it had no actual medicine. The patients agreed and the pills helped to reduce their symptoms. Kaptchuk's trials extends upon that historical study by adding a control group.
"We did this study on a shoe-string," says Kaptchuk. "No one would fund a study that was going to tell patients that they were going to get placebo." In the end, funding came from the National Center for Complementary and Alternative Medicine. While they have a somewhat murky track record in terms of the science they fund, no one I spoke to criticised the design of this trial.
Edzard Ernst, a professor of complementary medicine at the University of Exeter, says, "This is an elegant study which suggests that the ritual of giving a patient a remedy is clinically effective, even if that patient has been told that the remedy is a placebo." Kaptchuk himself says, "I suspect that just performing "the ritual of medicine" could have activated or primed self-healing mechanisms." And Amir Raz, a neuroscientist who studies placebos at McGill University, adds, "Scientific reports make it clear, even if strange and counterintuitive, that receiving - rather than the actual content of - medical treatment can trigger and propel a healing process."
But did the patients actually expect the placebos to 'work'? Here's where the paper gets frustrating. After the trial, Kaptchuk used a survey to check that the patients understood the experiment. A few tidbits are published: for example, the placebo group mostly thought that their pills contained 'sugar', with some suggesting 'nothing', 'flour' or 'calcium'. That seems strange - it would be fascinating to know if the patients experienced greater benefits depending on their expectations.
But the surveys also asked what the placebo group thought about the idea of taking placebo, and whether they were sceptical or expected it to work. These results "will be reported elsewhere." Kaptchuk said, "We haven't written them up yet. Combining a quantitative study with a qualitative study is difficult and editors don't give you enough room… Medical journals aren't good with qualitative data." (I wasn't aware that PLoS ONE had such restrictions, but perhaps an editor can comment)
Regardless, Kaptchuk thinks that people don't need to believe that the placebos will work. "My hypothesis is that much of what we did was just below conscious expectation," he says. "We even told patients they didn't have to believe it would work (although we also gave them good reasons for expecting that it would.)"
Limits and Implications
Kaptchuk acknowledges a few other weaknesses. It only looked at a small number of patients and it was quite short. Indeed, it was only meant as a "proof-of-principle" study, which will now need to be checked with larger, longer trials.
More importantly, the study wasn't a fully "double-blind" trial, where neither patient nor scientist knows who has been assigned to which treatment. Kaptchuk's team certainly had no idea which group the patients were in. But there was no way of ensuring that the patients were similarly ignorant, when some were clearly taking pills and others weren't. The apparent benefits of the placebo pills depended on the patients reliably and comparably reporting what they felt, for a condition that naturally waxes and wanes. Those who took placebos might have felt pressure to feel better, and this problem can't really be overcome in a trial that compares placebos to no treatment at all.
This might not matter, given that the study measured subjective experiences anyway. If a patient says that they feel better, that's arguably a good result. Kraptchuk realises this and only suggests that placebos could be used transparently for "illnesses primarily diagnosed by subjective symptoms and introspective self-appraisal". Placebo pills aren't going to be shrinking tumours any time soon, but they may prove useful for conditions like depression, anxiety or chronic pain.
Ernst isn't convinced. He says, "The effect size is probably too small to be clinically relevant. [It] is unlikely to be of practical use." To him, the results are interesting "mostly from a theoretical point of view." But Kaptchuk thinks that the effect he saw is "clinically meaningful,"
and comparable to some drugs being tested for IBS such as alosetron.
All in all, Kaptchuk thinks that the prescription of placebos doesn't have to violate the ethics of modern medicine. If delivered with "a plausible rationale… without deception or concealment," they can produce improvements in a patient's symptoms (indeed, Janet Stemwedel, ethicist and science blogger, proposed a similar situation in which it might be ethical to prescribe a placebo) At the very least, he says that "open-label placebo may have potential as a "wait and watch" strategy before prescriptions drugs are prescribed.
Does the new trial add another dimension to the debates? Not yet. Raz says, "These results are not new per se but they are interesting and replication is of utmost importance in science, especially when controversy rules. The use of placebos is ethically tenuous and will probably remain so for a while longer, but this effort is a step in the right direction. It shows that we can use placebos non-deceptively - a genuine placebo medication."
Benedetti adds, "There is not a single placebo effect but many, with different mechanisms (expectancy, conditioning, reward, anxiety modulation, genetic variants, and the like) and in different systems and conditions. The challenge for the future is to understand all these complex interactions and exploit them ethically at the bedside."
By Ed Yong
Reprinted with permission from Discover