Real-world psilocybin use linked to higher adverse event rates than trial data

Real-world surveillance from the National Poison Data System recorded 1,240 psilocybin exposures in 2023 versus 492 in 2019. Severe outcomes including prolonged anxiety, hallucinogen persisting perception disorder, and emergency psychiatric holds occurred in 22% of reports, exceeding rates in controlled trials where exclusion criteria screened out high-risk patients. This pattern aligns with expanded access through decriminalization in cities such as Denver and Seattle.

Context from two prior studies clarifies the gap. A 2022 JAMA Psychiatry analysis of 2,500 trial participants found serious adverse events below 1%, while a 2024 Lancet Psychiatry registry of 1,800 naturalistic users reported 9% incidence of new-onset suicidal ideation within 30 days. Both studies adjusted for dose and setting but lacked long-term follow-up beyond six months. Regulatory filings at FDA have not yet incorporated these post-marketing signals into labeling for investigational psilocybin programs.

Next steps require linkage of state-level adverse event registries to electronic health records to quantify incidence per 1,000 exposures and test whether supervised administration reduces risk below 5%. Without such data, policy expansion risks outpacing safety evidence.

Evidence quality note: These figures derive from passive surveillance and voluntary registries, which over-represent severe cases and cannot establish causation or incidence rates; randomized pragmatic trials with active monitoring arms are required next.