The Post-Acute Window: Suicidal Ideation and Behaviour After Psilocybin. A Rapid Computational Evidence Synthesis
Odin Loch
Zenodo (CERN European Organization for Nuclear Research) October 24, 2026 preprint DOI: 10.5281/zenodo.23612453 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Rapid computational evidence synthesis (single analyst, not a dual-reviewer systematic review) |
|---|---|
| Population | 285 interventional, non-microdose psilocybin trial registrations; posted results for NCT03775200; openFDA adverse-event reports |
| Intervention | Psilocybin |
| Dose | 25 mg, 10 mg, 1 mg |
| Duration | Post-dose windows of days 2–21; NCT03775200 time frame written as up to 12 weeks |
| Measures | suicidality assessment (parser-recognised, unnamed); suicidality scale (unnamed) |
| Topics | Psilocybin |
| Registration | NCT03775200 |
| Key findings | Public sources do not yet support a window-specific incidence of suicidal ideation or behaviour after psilocybin. Only 10 of 285 trial registrations (3.5%) scheduled a parser-recognised suicidality assessment within days 2–21 after dosing, and the NCT03775200 appendix schedules week 1 and week 2 assessments but leaves those weeks inside wider spans without separate counts. The three serious suicidal-behaviour events in the 25 mg arm were dated around days 29, 58 and 63. |
Abstract
Background. Averaging the mean change in a suicidality score can hide a subgroup whose suicidal thoughts worsen in a particular stretch of days after psilocybin. Trials often do not enrol people with psychosis, bipolar disorder, or current high suicide risk. Objective. To assemble, from public sources only, what can actually be counted about treatment-emergent suicidal ideation or behaviour in pre-specified post-dose windows, and to record how often registered trials say they measure that window and whom they exclude. Methods. This is a rapid computational evidence synthesis by one analyst, not a dual-reviewer systematic review and not a PROSPERO-registered review. It does not claim full PRISMA compliance. PubMed, a labelled Europe PMC translation, OpenAlex, ClinicalTrials.gov and openFDA were queried. The supplementary appendix to the 2022 phase 2b trial was read from a local PDF. Arm-level odds ratios were estimated only where numerators and denominators were present. No continuity correction was used. Poison-centre reports of mushroom ingestion as a means of self-harm were out of scope. Results. PubMed returned 1384 records. Of 285 interventional, non-microdose registrations, 10/285 (3.5 per cent; Wilson 95% CI 1.9 to 6.3) contained a parser-recognised suicidality assessment on a day or week inside days 2–21 after dosing. No Mantel-Haenszel pool met the pre-specified three-table rule. Posted results for NCT03775200, over a time frame written as up to 12 weeks, showed serious suicidal behaviour in 3/79 at 25 mg, 0/75 at 10 mg and 0/79 at 1 mg. Non-serious suicidal ideation was 5/79, 3/75 and 4/79. The journal article splits that programme into the dosing day, day 2 to week 3, and after week 3. The supplementary appendix schedules the suicidality scale at a week 1 visit and a week 2 visit, and it does not publish separate counts for those two weeks. Any worsening from day 1 to week 3 was 11/79, 13/75 and 7/79. Any worsening after week 3 up to week 12 was 12/79, 12/75 and 12/79. The three serious suicidal-behaviour events in the 25 mg arm are dated around day 29, day 58 and day 63, so they are not in week 1 or week 2. Week 1 and week 2 are still not separable as an incidence. In openFDA, suicidal ideation had two reports (reporting odds ratio 3.40, Woolf 95% CI 0.84 to 13.73); the pre-specified signal rule was not met. Conclusions. Public sources do not yet support a window-specific incidence of suicidal ideation or behaviour after psilocybin. The appendix shows the two early weeks were on the assessment schedule and leaves them inside wider spans. Explicit in-window schedules are uncommon in registry text, and exclusions of psychosis, bipolar disorder and current suicidality are common. The record contains the preprint (PDF and Markdown) and the analysis code that reproduces the counts (MIT licence for the code).
Comparable studies
Other systematic reviews and meta-analyses on psilocybin, most cited first.
| Study | Year | Design | Participants |
|---|---|---|---|
| Therapeutic effects of classic serotonergic psychedelics: A systematic review of modern‐era clinical studies Patients with cancer- or illness-related anxiety and depression, major depressive... | 2020 | Systematic review | n = 188 |
| Blinding and expectancy confounds in psychedelic randomized controlled trials | 2021 | Systematic review | |
| Antidepressive, anxiolytic, and antiaddictive effects of ayahuasca, psilocybin and lysergic acid diethylamide (LSD): a systematic review of clinical trials published in the last 25 years Patients from included clinical trials | 2016 | Systematic review | |
| The experimental effects of psilocybin on symptoms of anxiety and depression: A meta-analysis People with elevated symptoms of anxiety and depression | 2020 | Meta-analysis | n = 117 |
| Default Mode Network Modulation by Psychedelics: A Systematic Review Participants from included studies | 2022 | Systematic review |