Abstract
Clinical trials of psychedelic compounds like psilocybin, lysergic acid diethylamide (LSD), and N,N-dimethyltrptamine (DMT) have forced a reconsideration of how nondrug factors, such as participant expectations, are measured and controlled in mental health research. As doses of these profoundly psychoactive substances increase, so does the difficulty in concealing the treatment condition in the classic double-blind, placebo-controlled trial design. As widespread public enthusiasm for the promise of psychedelic therapy grows, so do questions regarding whether and how much trial results are biased by positive expectancy. First, we review the key concepts related to expectancy and its measurement. Then, we review expectancy effects that have been reported in both micro- and macrodose psychedelic trials from the modern era. Finally, we consider expectancy as a discrete physiological process that can be independent of, or even interact with, the drug effect. Expectancy effects can be harnessed to improve treatment outcomes and can also be actively managed in controlled studies to enhance the rigor and generalizability of future psychedelic trials.
Section snippets
Placebo Response, Placebo Effect, and Expectancy
Expert consensus distinguishes the placebo response from the placebo effect (6,7). The placebo response includes all health changes that result after administration of an inactive treatment, which includes behavior related to being observed, i.e., the Hawthorne effect (8), regression to the mean (9), and other nonspecific effects of clinical care (10).
The placebo/nocebo effect refers to the changes that are specifically attributable to placebo mechanisms and can be defined as the difference
Unblinding and Its Relationship to Expectancy
Unblinding occurs when patients recognize their treatment allocation in a randomized controlled trial (RCT). Usually, unblinding is captured by asking patients at the end of the trial to guess their treatment. If this treatment guess/perceived treatment matches the actual treatment at a higher than chance rate, then the trial has weak blinding integrity.
Most psychiatric trials have not assessed blinding integrity; it has been estimated that in this domain, only 2% to 7% of trials report
Changing Expectancy
Expectancy is not a static trait; rather, it changes dynamically according to new information and experiences (27). The ViolEx model has been proposed to account for both how expectancies change and why they are sometimes maintained even when contradictory information is obtained (27). For example, consider the fact that many psychedelic treatments include more than 1 psychedelic drug session. It is reasonable to assume that patients change their expectations in-between drug sessions as a
Measuring Expectancy and Blinding Integrity
Studies that assess expectancy often use a self-constructed questionnaire rather than a standardized and validated measure, which hinders generalization. To help overcome this issue, here we list validated expectancy measures. The most popular one is the Credibility and Expectancy Scale, which has been shown to have good psychometric properties (34). Alternative measures are the Stanford Expectations of Treatment Scale (35), which focuses on perceived benefits and risks of treatments; the
Macrodose Trials
Blinding integrity has been found to be poor in psychedelic macrodose RCTs that measured it (42, 43, 44). For example, Holze et al. noted that only 1 patient of 20 mistook lysergic acid diethylamide (LSD) for a placebo (95% correct guess rate) (45). Similarly, Bogenschutz et al. reported that 90 of 95 treated patients correctly guessed their treatment allocation (94% correct guess rate), showing that blinding integrity was almost nonexistent in these trials (4).
Expectancy has only been measured
Separating Expectancy and Drug Effects
Attempts to conceal treatment allocation in psychedelic studies often involve incomplete disclosure of the study design. In some nontherapeutic studies, investigators inform participants that they may receive one among many psychoactive substances, although the actual study design may only involve a limited number of drugs and doses (61,62). For example, in an LSD microdose trial, de Wit et al. told patients that they might receive a hallucinogen/cannabinoid/opioid/stimulant/sedative or placebo
Drug-Placebo Interactions
Most trials consider the observed treatment effect to be the sum of the placebo and the drug effect. However, this simple arithmetic may not apply to all circumstances, for example where interactions exist between drug and placebo effects, i.e., the total treatment effect may be less or more than the sum of the drug and placebo effects (79). Such drug × placebo interactions can be demonstrated in balanced placebo designs; for example, Hammami et al. (80) used a crossover variant of the balanced
Lessons From the Pain Literature
Outsized placebo responses have been problematic for clinical trials of novel pain management drugs and interventions for decades and have contributed to the difficulty in developing novel analgesics despite an impressive array of novel molecular targets validated in animal models (82). Pain, as a sensory experience, can be quantified using various psychophysical measures, which allows for clear delineation of drug-induced relief. In contrast, depression, which is the subject of most
The Neurophysiological Basis of Expectancy
When placebo first appeared in medical dictionaries and texts from the 18th and 19th century, the term was generally defined as an inert substance whose primary value is to soothe, rather than benefit, the patient (61). Since the mid-20th century, most usages of the term preserved the distinction between true benefit and psychological benefit (62), perhaps reflecting the increasing emphasis across medicine and psychiatry that drug therapy could be explained primarily in terms of the drug’s
Conclusions
Expectancy effects are widely expected in psychedelic trials, but to date, only a few studies have assessed them. We distinguish between expectancy and unblinding and recommend treating these as separate but related concepts. A deeper understanding of these effects in psychedelic trials will require, first and foremost, a consistent effort across research groups to incorporate standardized measures of both expectancy and blinding integrity throughout the life cycle of these complex trials, as