Employees of several companies that conduct clinical trials for the pharmaceutical industry have just published a paper to test a tool intended to reduce the placebo effect in drug trials. It’s called the Placebo-Control Reminder Script (PCRS), and it was created by and licensed to several drug companies.
Their study’s goal was to demonstrate that the PCRS, when used in a clinical trial, can make the active drug look better by diminishing the improvement that people naturally experience in the placebo arm of the trial.
Elan A. Cohen led the study at the Hassman Research Institute (HRI). It included other employees of HRI as well as employees of APEX Innovative Services, Collaborative Neuroscience Network (CNN), and VeraSci. The owner of VeraSci is also a listed author. All of these firms are funded by the pharmaceutical industry to conduct clinical trials. The study was published in Neuropsychopharmacology.
Cohen et al. write that “The remarkably high and growing placebo response rates in clinical trials for CNS indications, such as depression and schizophrenia, constitute a major challenge for the drug development enterprise.”
Some researchers suggest that the strong placebo effect is a good thing, which could be harnessed to enable improvement without an active drug’s adverse effects. For instance, Beth Israel Deaconess Medical Center and Harvard Medical School have set up a Program in Placebo Studies & Therapeutic Encounter (PiPS) that aims to enhance the placebo effect.
Nonetheless, the placebo effect is quite troubling to the pharmaceutical industry, as their drugs have to be better than placebo (literally, better than nothing) to be approved for use—and they often struggle to meet that mark.
According to Cohen et al., “Drug development programs continue to face dauntingly low success rates in late phase randomized placebo-controlled trials (RCTs).”
The placebo effect—including natural improvement over time as well as an improvement due to expectations—is quite powerful when it comes to diagnoses like depression and schizophrenia. According to the authors, up to 50% of people taking a placebo find their depression improves, while for schizophrenia, that number is 41%.
In clinical trials of drugs intended to treat these disorders, the active drug is often no better or slightly better than the placebo. When unpublished trials are included, only half (51%) of antidepressant studies find that the drug is better than the placebo. Researchers have suggested, in fact, that any supposed antidepressant effect of SSRIs may be entirely due to the placebo effect.
There are several reasons for having a placebo-controlled study. First, the placebo group can indicate how many people would have improved anyway, even if they did not take the drug. This can include natural improvement over time and regression to the mean (in which the most severe cases improve significantly more). Second, the placebo group can indicate the effects of expectation—that is, people believe they are taking an active drug and so improve because they believe they will improve.
But while these reasons are helpful to the consumer, they are dangerous to the pharmaceutical industry since they make the active drug look less appealing. After all, if the same number of people improve on the drug or off it, then why should the consumer take the drug?
Cohen and the other consultants believe the PCRS can solve this problem.
The PCRS consists of a paragraph about the placebo effect that is first read aloud to the participant; then, the participant must repeat back the information to ensure they understand. The “information” includes advising the participant that they should not expect to feel better, that they should not expect the drug to work, and that they should know that they are in the placebo group if they experience no side effects. For instance: “Remember that the placebo is inactive and should do nothing to help your symptoms and should cause no side effects.”
In essence, the PCRS is a technical and purposeful way to break the blinding of a study—to ensure that the participants know whether they are receiving the active drug or not.
Researchers already argue that accidentally breaking the blind of a study is a severe problem that biases the study drug’s results. As Irving Kirsch writes in a peer-reviewed paper about antidepressant ineffectiveness, “The difference in improvement between drug and placebo is not clinically meaningful and may be due to breaking blind by both patients and clinicians.”
It is unclear how it could be considered ethical to purposefully create this bias in a clinical trial.
To test their hypothesis that the PCRS could stymie the placebo effect, Cohen et al. conducted a two-week single-blind trial. Single-blind means that the participants didn’t know which group they were in, but the researchers did. (Since the researchers could be biased in single-blind trials, double-blind trials are considered more likely to provide less-biased information.)
The consultants included 137 participants, all of whom met the criteria for current depression of at least moderate severity but no suicidal thoughts. Some had formal depression diagnoses, while others had diagnoses of schizophrenia or schizophreniform disorder.
The participants were told that they were in a clinical trial for a new antidepressant and that they would be split into two groups—one receiving the drug and one receiving the placebo. However, in reality, both groups received the placebo. The actual difference was that one group was also given the PCRS, while the other group was not.
Both groups improved over time, consistent with the natural course of “depression” and with the placebo effect. However, in line with Cohen et al.’s expectations, the group that received the PCRS showed less improvement than the group that didn’t receive the PCRS.
In the group that received the PCRS, 29.7% believed they had improved; in the group without the PCRS, 52.4% believed they had improved.
Average BDI scores dropped from about 30 to about 25 in the group that received the PCRS. However, in the “true placebo” group who did not receive the PCRS, average BDI scores dropped to below 20—meaning that they no longer even met the criteria for moderate depression.
That is, the “true placebo” group improved enough to no longer meet the criteria for depression severity that got them into the study in the first place. But those in the PCRS group improved only about half as much and still had moderate depression according to the BDI. So, the PCRS works as intended—it helps stop people from getting better on their own.
Cohen and the other consultants write that their results demonstrate that the PCRS is a useful tool for the pharmaceutical industry’s development of new drugs for psychiatric conditions.
“Briefly educating participants about placebo response factors can help mitigate the large placebo response rates that are increasingly seen in failed CNS drug development programs,” they write.
“By enhancing signal detection, such procedures may help drug developers progress compounds to faster approval and reach patients who are suffering sooner.”
Cohen, E. A., Hassman, H. H., Ereshefsky, L., Walling, D. P., Grindell, V. M., Keefe, R. S. E., . . . & Horan, W. P. (2021). Placebo response mitigation with a participant-focused psychoeducational procedure: a randomized, single-blind, all placebo study in major depressive and psychotic disorders. Neuropsychopharmacology, 46, 844–850. https://doi.org/10.1038/s41386-020-00911-5 (Link)