Placebo effect ethics in medical treatment

Originally published in MedPage Today

by John Gever, MedPage Today Senior Editor

The placebo effect is a real and therapeutic psychological phenomenon that, with more research, could be exploited more systematically in medical practice, said Australian researchers.

“There is not one placebo effect, but many,” wrote Damien Finniss, MSc, of the University of Sydney in Australia, and colleagues in the Feb. 20 issue of The Lancet, arguing that recent research shows that placebo effects can exist even in the absence of an actual placebo.

Much remains to be learned about the mechanisms underlying placebo effects as well as the ethics of providing placebo-based treatments, but it seems likely that eventually these effects can be harnessed to improve patient care, they concluded.

Finniss and colleagues reviewed the literature on the placebo effect, which suggests that it is far more complicated than many people think.

It is more than just taking a pill that patients think is, or could be, an active drug. The act of receiving a pill brings a whole “psychosocial context” into play, according to Finniss and colleagues: it includes not only the patient’s desires and expectations but also those of the clinician; their past experiences; and their interaction with each other and with their physical and social environment,” they wrote.

“The placebo is inert but the psychosocial context is not,” the researchers asserted.

As an example, they cited a randomized, single-blind, three-arm trial that enrolled 262 patients with irritable bowel syndrome, published in 2008 in BMJ. One arm was no treatment; the other two arms used a placebo treatment (a sham acupuncture device).

In one of the latter arms, for interactions with the patients the clinician followed a script that sought to convey attention, warmth, and confidence as well as, occasionally, thoughtful silence. In the other, the clinician simply delivered the placebo needles as quickly and with as little interaction with the patients as possible.

Response rates in the no-treatment, placebo-alone, and placebo plus supportive care groups were 28%, 44%, and 62%, respectively (P<0.001).

“This trial showed that placebo effects can be incrementally added in a manner resembling a graded dose escalation of component factors,” Finniss and colleagues wrote.

They also noted that certain patient characteristics — extroversion, agreeability, and openness to experiences — were associated with heightened responses when they received supportive care in addition to the placebo treatment, but not with the placebo alone. Outcomes also differed with the practitioners.

The Picture Gets More Complicated

Finniss and colleagues also noted several studies showing that different delivery methods can influence the placebo effect. One trial in patients with chronic pain found that placebo pills were significantly more effective in some outcomes than were placebo acupuncture treatments, whereas for other outcomes the reverse was true.

This study also demonstrated the “nocebo” effect, in which patients experience adverse effects from an inert treatment that they might expect to have adverse effects, Finniss and colleagues added. Patients receiving the pills were told they might experience drowsiness, while those subjected to the sham acupuncture were informed about the side effects of acupuncture.

Rates of reported side effects in the study were about 30% in both groups, but the nature of adverse effects reflected the instructions they were given. In fact, ability to sleep was the chief efficacy outcome for which the placebo pills were better than placebo acupuncture.

Still another demonstration of the placebo effect’s complexity, according to Finniss and colleagues, comes from trials following an “open-hidden” design. These trials have two arms: one in which a clinician injects a drug directly into the patient in a clinical setting, while the other involves a computerized pump delivery that infuses the drug silently, outside the normal clinical setting and without the patient’s awareness (although the patient knows that the drug will be delivered at some point).

Several studies like this have shown that drugs are less effective when given by hidden methods. Again, Finniss and colleagues explained, “the overall outcome of a treatment combines the specific pharmacological or physiological action of the treatment and the psychosocial context in which it is delivered.”

They argued that this effect may underlie some or all of the apparent effectiveness of real drugs. They pointed to a cholecystokinin antagonist called proglumide, that appeared to be more effective than placebo against pain in a randomized trial. But when tested in an open-hidden study, it turned out the drug had no effect when patients didn’t know it was coming.

The conclusion was that proglumide doesn’t act on pain pathways, but rather on different neural systems that mediate expectations. “The drug achieves a response by interacting with and enhancing placebo mechanisms,” Finniss and colleagues wrote.

“Placebo mechanisms can interact with drug treatments, even if no placebo is given, since every treatment is given in a therapeutic context that has potential to activate and modulate placebo mechanisms,” they contended.

But Is It Ethical?

These findings suggest that perhaps the placebo effect could be harnessed deliberately in patient care, but that raises thorny ethical issues, according to Finniss and colleagues.

First and foremost, placebos as treatments haven’t been studied with enough rigor at this point, the researchers indicated, calling for additional research in specific clinical settings.

“A second important ethical consideration relates to whether and how placebo effects can be promoted without deception,” they continued.

Providing a supportive environment that boosts treatment effectiveness is clearly ethical, but recommending a treatment that acts purely on patients’ expectations rather than their actual pathologies “is a more complicated and controversial question,” according to Finniss and colleagues.

They noted that practicing deception in clinical practice is even more problematic than in a research setting. “To recommend or give a placebo intervention deceptively as a treatment with specific efficacy for a patient’s condition violates informed consent and threatens the trust that is central to clinical practice,” they wrote.

Finniss and colleagues indicated that the practice of some physicians to prescribe drugs “with the main intent of promoting a placebo response … is deceptive or at least not sufficiently transparent.”

They also expressed skepticism about telling patients that a placebo “has been shown to be effective by altering pain transmission in ways similar to other treatments.”

But if the clinician also said, “this pill has no active drug and would be working through psychological mechanisms to promote self-healing,” that would eliminate the deception — but it’s currently unclear whether that full disclosure would undermine the placebo effect.

“Clinically focused research is needed to explore nondeceptive techniques for prescribing treatments aimed at promoting placebo effects,” Finniss and colleagues concluded.

Visit for more clinical trial news.