Researchers Push to End Placebo Run-in Periods in Antidepressant Studies

Meta-analysis finds that the placebo run-in methodology reduces the placebo effect and finds antidepressants to be less useful.


The placebo run-in (PRI) is a common practice in antidepressant research. At the beginning of a study, all participants are given the placebo for a period of time. Those who do well on the placebo are kicked out of the study. Then the remaining participants, who did not respond to the placebo, are randomly assigned to receive either the drug or placebo for the rest of the study. This practice is designed to make the active drug look better by diminishing the placebo effect.

But in a new meta-analysis in JAMA Psychiatry, researchers criticize this practice and call for an end to this method. Most importantly, they found that the PRI does not do a very good job at making antidepressants look better. They write that this is because much of the “antidepressant” effect of the active drug is attributable to the placebo effect—so eliminating placebo responders makes both groups look worse.

The research was led by Amelia J. Scott at Macquarie University, Australia. They write:

“This systematic review and meta-analysis provide evidence to suggest that the use of PRI periods has no scientific basis in trials of antidepressants. At the same time, PRI periods carry numerous costs and risks. These findings suggest that the use of PRI periods should not continue in RCTs of antidepressants.”

Scott and the other researchers conducted a meta-analysis of 347 studies on antidepressant efficacy. They found that 174 (50%) used a PRI period. However, only 25 of those studies provided a rationale for using a PRI. Of those, 22 admitted that their reason for a PRI was “to identify and exclude placebo responders.”

Each study excluded, on average, 9.5% of their participants because they responded to the placebo during the PRI period.

As expected, Scott and the other researchers found that the placebo response was greater in studies that did not use a PRI. However, this did not affect the difference between groups because the antidepressant group’s response was also greater in studies that did not use a PRI.

Why was this? The researchers write:

“The use and reasoning behind PRI periods highlight an ongoing misconception that placebo responses do not contribute additively to symptom improvement within active treatment groups.”

A PRI period doesn’t help antidepressants look better than the placebo because much of the “antidepressant” effect of the drug is also due to the placebo effect.

This finding is consistent with the New England Journal of Medicine study that found the drugs only beat the placebo in half of the antidepressant trials. Researchers have consistently found that most—if not all—of the “antidepressant” effect of the drugs is actually just the placebo effect.

Scott and the other researchers also suggest that using a PRI period further diminishes the generalizability of antidepressant trials—which, they argue, already do a poor job of representing the patients who are actually seen in the real world. In addition, they write that the overzealous exclusion criteria used in these studies, for instance, means that the population studied does not reflect real-life practice.

Thus, on ethical grounds—and because the PRI doesn’t even enhance the outcomes the pharmaceutical industry hoped it would—they call for its elimination.



Scott, A. J., Sharpe, L., Quinn, V., & Colagiuri, B. (2021). Association of single-blind placebo run-in periods with the placebo response in randomized clinical trials of antidepressants: A systematic review and meta-analysis. JAMA Psychiatry. Published online November 10, 2021. doi:10.1001/jamapsychiatry.2021.3204 (Link)


  1. Great article, Peter! This study provides additional, unique, and really interesting evidence that antidepressants are basically placebos (but with adverse effects). I’m not sure from your write-up if the researchers addressed what I see as the biggest problem with using a PRI in a clinical trial: it intentionally biases the study by attempting to stack the deck against the placebo condition. The fact that this biased design feature doesn’t work is poetic justice but for me, the best reason to oppose PRI is that it is basically at attempt at scientific fraud. Whether it works or not is beside the point. A related design feature that is also basically scientific fraud is a drug run-in phase in which all participants are first assigned to the drug condition, and those who do not respond during the first few weeks, or have negative responses, are excluded from the trial. It never ceases to amaze me that such transparently scientifically ridiculous practices are commonplace in psychiatric drug trials published in top medical journals. Just goes to show that drug trials are typically conducted to *demonstrate that the drugs work,* which is marketing and not science, and makes medical journals more like consumer product magazines than scientific publications.

    Report comment