Can a sugar pill placebo cause more harm than good?

Science and Technology

A placebo is an inert medical treatment, often a sugar pill, that is administered during a clinical trial to differentiate between a response to active medication and a response with no treatment. They have no therapeutic value but can alter the patient’s perception of their symptoms.

A new review of 1271 randomised trials and 250,726 placebo-treated patients from the last 12 years shows that 5% of all participants withdrew from trials and placebo treatment due to serious side effects, while nearly half reported less serious side effects. Serious side effects included reports of abdominal pain, cancer, strokes and even death, while less serious side effects included chest pain and fatigue. Surprisingly, the review also found that more side effects were reported in the placebo group than in the untreated group.

Why do these side effects occur in placebo groups? The review identifies two main reasons for this: misattribution and negative expectations.

Symptoms that develop independently of the trial are often reported as side effects and are misattributed to the trial. For example, a participant could develop a migraine for any number of reasons, but by being a part of a trial, the patient is more likely to attribute the migraine to that trial and report it as a side effect. Misattributions are incredibly difficult to avoid and hard to minimise.

“The current model of informing participants about randomised trials involve discussing more harms than benefits of the trial…”

The effects of negative expectations are called ‘nocebos’, a negative placebo. This is when the patient anticipates a negative effect of the disease for which the placebo is administered and suffers that effect even though they have only taken a sugar pill. It is the opposite of the placebo effect.

The BMJ published a study comparing two different placebos for arm pain, highlighting the effects of negative expectations. In the study, one group received a sugar pill while the second group received a sham acupuncture with retractable needles. The first group reported that they felt so sluggish that they could not get out of bed while the second group reported swelling, redness and extreme pain. All of these symptoms had been described as potential side effects of the trial.

In another study patients being treated for unstable angina with aspirin or sulfinpyrazone were warned of gastrointestinal side effects. It was found that they were six times more likely to withdraw from the study due to reports of gastrointestinal problems.

Verbal suggestion, past experience and expectations can all contribute to negative expectations. As a result trial quality may decrease as participants drop out, and also the issue of trial ethics is brought into question due to the harm being caused. Jeremy Howick, an award-winning philosopher and medical researcher, believes that we can reduce harm by reducing negative expectations.

The current model of informing participants about randomised trials involve discussing more harms than benefits of the trial. Research is being done at the University of Oxford and University of Cardiff to find an optimal model for informing patients in a clinical trial such that there is consent which respects patient autonomy and does not introduce unnecessary harm. One way of doing this would be to tell participants, for example, that the new treatment is safe for 90% of all patients undergoing treatment, rather than saying that is it harmful to 10% of patients. Others want to ‘contextualise’ information, tailoring and adapting it for individual patients in order to eliminate natural history bias.

Co-author Professor Kerry Hood, Director of Cardiff Centre for Trials Research, believes that it is possible to balance the information about trial benefits and harms in a way that is fact-based and doesn’t cause unnecessary harm.

Image Credit: Mpelletier1 (CC BY-SA 3.0)


Sign up for the newsletter!

Want to contribute? Join our contributors’ group here or email us – click here for contact details