We Should Use Placebos More Often, Especially In Surgery
You’re ill, probably scared, not sure of the future. Your doctor talks to you about your condition and says that, unfortunately, there are no effective treatments. However, she knows of a clinical trial for a new drug, and they’re currently recruiting participants.
The problem is that you’ll be “randomised” to one of two groups. One will receive the new treatment and the other a placebo. The odds of getting the new treatment are 50:50.
So, what are you going to do?
Placebo treatment is a dummy or inactive treatment. Clinical trials are designed so that you (and often your doctor) don’t know whether you’re getting the experimental treatment or the placebo.
This “blinded” trial design helps to avoid “bias”. If a patient knows they’re getting the real pill, they may be more likely to report feeling better.
Blinded trials let researchers compare outcomes from those that get the new treatment with those that don’t without any bias corrupting the results. In the case of drug treatment, a placebo will look like a real drug (a pill or infusion) but it won’t contain an active ingredient. In surgical research, the placebo may involve creating an incision in the skin and nothing more.
Why We Need Placebos
If you’re given an experimental treatment then there might be two ways in which you’ll benefit. The treatment could improve the underlying condition, as it’s designed to do.
Alternatively, even if it doesn’t work, the fact that you’re getting something that you believe (or hope) will make you better might in fact make you feel better. This is the placebo effect, and it can convince people that the treatment works even when it’s ineffective.
So it’s important that the researchers take this into account (“control for it” in medical parlance) when testing to see if a treatment is effective.
Placebo controls are particularly important when the outcome of the clinical trial is “soft” or subjective. For example, the amount of pain someone feels is dependent, in part, on the person’s state of mind.
People often report a reduction in pain when they are given a drug, even a placebo. “Harder”, more objective, outcomes (such as the shrinkage of a tumor) are less prone to being influenced by the beliefs of the patient and doctor.
Weighing Up Your Options
But even given the scientific need for placebos, why should you agree to take part? At the start of a study no one knows if the experimental treatment will benefit patients; if they did, there would be no need for the study.
The scientists who are designing the study might believe that it will help (after all they have worked hard to develop their treatment), but that’s very different from knowing.
Indeed, the majority of new treatments never make it to clinical practice. They also carry a risk – no drug or intervention is totally risk free.
So if you take part in the trial you risk either getting an experimental treatment that may not work and may cause harm, or having a placebo that is less harmful but is not going to treat your medical problem. Of course, you may get a new treatment that works. You will also be helping the development of new treatments for that condition.
Placebo controlled trials should be used more often in surgery to see if treatments work or not, as highlighted by recent Royal College of Surgeons guidelines. There are too few surgical studies that use placebo; only 53 had been published up to November 2013.
What’s astounding is that half of these studies showed no benefit of the surgical intervention over placebo. In many cases the operations are still routinely performed.
This means that many patients underwent the risk and inconvenience of a surgery for no benefit. The ethical failing is not that the studies used placebo controls, but that the studies weren’t done earlier.
The use of placebo is a vital, though often misunderstood, part of medical research. But the careful and proper use of placebo is not only good science, it is good ethics.
Author: Andrew George, Deputy Vice-Chancellor, Brunel University London