Control Conditions in Randomized Trials of Psychedelics

Summary

A systematic review of randomized trials of psychedelics in humans found that most studies use inert placebos rather than active comparators, and very few assess whether blinding was successful. Of 86 unique studies, 61.2% used an inert placebo, 20.0% used active comparators, and only 17.3% included any assessment of blinding; when assessed, blinding success was generally poor. Only 3 of 21 therapeutic trials compared psychological support to a minimally supportive condition. The review concludes that randomized psychedelic trials underutilize blind assessment, active drug controls, and adequate control conditions for psychological support, and recommends improvements to trial methods.

Study at a glance

Abstract

To systematically review control conditions of all available randomized psychedelic trials. We searched PubMed, PsycINFO, and EMBASE for randomized trials of psychedelics in humans from 1940 through May 2020 with no language restrictions. PRISMA guidelines were followed. (PROSPERO registration number: PROSPERO-CRD42020205341.). All randomized trials of psychedelics in humans from 1940 through May 2020 were included. Two independent reviewers performed extraction. Extracted data included study design, demographics, blinding type, whether and how blind integrity was assessed, psychedelic used and dose, drug control condition and dose, type of non-drug control condition, number of dosing sessions, and recruitment source. Outcome data were not collected. In total, 126 articles were included, encompassing 86 unique studies. Of studies with a drug control condition (80), 49 (61.2%) used an inert placebo control, 16 (20.0%) used active comparators, 12 (15.0%) used both, and 3 (3.8%) used only different active psychedelic doses as a control. Only 3 of 21 therapeutic trials compared the use of psychological support to a minimally supportive condition. The majority (81/86; 94%) of studies were blinded, though only 14 (17.3%) included blind assessment; only 8 of these 14 studies assessed participants' blinding. Blinding success, assessed in highly varied ways, was generally poor. Randomized psychedelic trials underutilize elements that would improve quality or provide important information: blind assessment, active drug controls, and testing psychological support against minimal-support conditions. Several queried categories, including blind integrity assessment and details of non-drug control conditions, were insufficiently reported by many reviewed studies. Recommendations are provided to improve trial methods.