Reporting of side-effects in clinical trials of psilocybin-assisted psychotherapy for psychiatric conditions: systematic review

Side-effects reporting in clinical trials of psilocybin-assisted psychotherapy (PAP) for psychiatric conditions is inconsistent but improving over time. A systematic review of 24 trials published between 2005 and 2024 found that only six had high-quality side-effects reporting, while nine were low and five were very low. All nine randomized controlled trials showed high risk of bias for side-effects outcomes. There was no evidence of systematic underreporting in published articles compared with trial registers, but variability in reporting hindered comparisons. The authors conclude that existing evidence has a high risk of bias and that future trials should follow best-practice guidelines, and discussions with patients should emphasize current uncertainty about PAP side-effects.

Study at a glance

Abstract

Background Psilocybin-assisted psychotherapy (PAP) has gained attention as a promising intervention for conditions including depression, anxiety and post-traumatic stress disorder, but understanding of its side-effects is limited. This review evaluates the quality of side-effects reporting in PAP trials, to guide treatment, policy and research. Aims To assess side-effects reporting quality in PAP trials for psychiatric conditions, comparing published articles and ClinicalTrials.gov records. Method A PROSPERO-registered review (no. CRD42023458960) included English-language PAP trials (2005–2024) identified via Embase, CENTRAL, PubMed and reference searches. Reporting quality was assessed using the CONSORT Harms extension, categorised as either high (17–21), moderate (12–16), low (7–11) or very low (0–6). Randomised controlled trials underwent risk of bias analysis, and descriptive statistics compared side-effects across sources. Results Twenty-four trials were included. Reporting quality was high in six studies, moderate in four, low in nine and very low in five. All randomised controlled trials ( n = 9) showed high risk of bias for side-effects outcomes. Variability in reporting hindered comparisons between articles and ClinicalTrials.gov, underscoring the need for standardisation. Overall, there was no evidence of systematic underreporting of side-effects in published articles compared with trial registers. Conclusions Side-effects reporting in PAP trials is inconsistent but is improving over time. Existing evidence has a high risk of bias. Future trials should align with best-practice guidelines for side-effects reporting. Discussions with patients should prioritise findings from high-quality studies and emphasise the current uncertainty regarding PAP side-effects.