Clinical Research: Phase 1 - Phase 4

The Placebo Problem, Part 11: Trial-Independent Placebo Contributors

This is the eleventh installment of our look at the increasingly high placebo response that is plaguing clinical trials in analgesia and psychiatry. Read the rest of the posts in the series here.

As we discussed last week, there are so many trial-related details that can influence the magnitude of the placebo response. But there are also two extrinsic, trial-independent factors that also contribute to a subject’s improvement in the absence of a pharmacologically active treatment: the natural tendency of diseases to get better over time and “regression to the mean,” a statistical phenomenon in which outliers in a dataset tend to move toward the mean value over time.

Headed toward better
Many illnesses, including pain and depression, tend to improve over time. Coupled with the fact that subjects have a tendency to enroll in a clinical trial when their symptoms are more severe and therefore bothersome, it is not surprising that a substantial proportion of subjects get better over time without any treatment. This spontaneous improvement is thus part of the response seen in all clinical trial groups. One analysis of antidepressant trials that utilized a “wait list” control group to track natural improvement over time found that participants in this group experienced about 40 percent of the symptom improvement of those in the placebo group.

Outliers move in
A second extrinsic factor that can influence the placebo response is regression to the mean. When repeated measures are made on a subject over time, random error results in variability. Many of the values are higher or lower than the mean. If the initial measurement is not close to the mean, subsequent measurements are likely to be closer to the mean than the original. So, if in the initial measurement a person’s symptoms were worse than their true mean, they appear to improve as the trial progresses. When this happens on a group level, there appears to be more improvement in the group than there really is. However, there is no corresponding worsening toward the mean to offset this effect, because trials have minimum symptom severity cutoffs. Therefore, subjects whose initial value showed that they were less sick than reality are excluded from the trial.

Unlike the trial-related details we discussed last week, these extrinsic factors are largely outside of our control. No treatment and waitlist groups are potential ways to tease apart the relative contributions of these extrinsic factors compared to other placebo contributors in research on the placebo effect. However, these groups aren’t practical or really even necessary in placebo-controlled clinical trials evaluating novel therapeutics, since the placebo response incorporates the effects of these factors.

Next week, we’ll be shifting gears to discuss the placebo’s opposite, the nocebo effect, when negative expectations about treatment can lead to a worsening of symptoms. We’ll take a detailed look at what the nocebo is, the underlying mechanisms, and how it can impact clinical trials.

If you’re new to our Placebo Problem series, why not start from the beginning, with our part one overview. And, for more industry insights, don’t forget to sign up.