Frequently Asked Questions

Not independently in the fully adjusted model for psychosis diagnoses. In unadjusted analyses, hallucinogen-related admissions were associated with higher rates of psychosis between 30 days and 6 months after the index admission, but the association weakened after accounting for baseline psychiatric history and other clinical factors. The unadjusted hazard ratio for postindex psychosis diagnosis was 1.22 (95% CI, 1.19–1.25), fell to 1.12 (95% CI, 1.09–1.15) after adjustment for baseline psychotic disorder diagnoses, and was no longer significant after additional adjustment for baseline clinical characteristics and demographics (HR=0.97; 95% CI, 0.95–1.00).

Prior psychotic disorders were substantially more common before hallucinogen-related admissions than before other substance-related admissions. Baseline psychotic disorders were present in 19.2% of the hallucinogen-related admissions cohort versus 6.3% of the nonhallucinogen cohort, and baseline psychosis admissions were 13.9% versus 3.7%, respectively. These differences support the study's conclusion that preexisting psychiatric vulnerability materially confounds the crude hallucinogen-psychosis association.

Between 30 days and 6 months after the index substance-related admission, 16.4% of patients with a hallucinogen-related admission had at least 1 psychosis diagnosis, compared with 6.6% of patients with nonhallucinogen substance-related admissions (P<.001). During the same period, 9.3% versus 3.3% had at least 1 psychosis-related admission (P<.001). These are unadjusted comparisons and do not account for the higher baseline psychiatric burden in the hallucinogen-related cohort.

In adjusted models, baseline psychiatric and substance-related comorbidity were consistent predictors of later psychosis. The strongest reported predictors were baseline psychotic admissions (HRs 1.28–1.35), stimulant use disorders (HRs 1.27–1.32), and opioid use disorder (HRs 1.28–1.32). The study therefore suggests that underlying psychiatric vulnerability and co-occurring substance use mattered more than the hallucinogen-related admission itself when estimating short-term psychosis risk.

Polysubstance involvement was very common. Of the 6,184 individuals with at least 1 hallucinogen-related admission, only 695 patients, or about 10%, had hallucinogen-related admissions without any nonhallucinogen substance-related admissions during enrollment. The authors interpreted this as evidence that most patients with hallucinogen-related admissions likely had broader substance-use complexity rather than isolated hallucinogen exposure.

Baseline substance-related claims were frequent in the year before the first hallucinogen-related admission. In that cohort, 22.5% had opioid use disorder, 36.6% had cannabis-related disorder, 16.2% had stimulant-related disorder, and 28.1% had alcohol use disorder before the index hallucinogen-related admission. These rates were higher than those seen before nonhallucinogen substance-related admissions, where the corresponding baseline SUD rates ranged from 6.0% to 19.6% (P<.001 for all comparisons).

Timing is difficult to interpret because the hallucinogen-related admissions cohort was young, with a median age of 24 years, which overlaps with the usual age of onset for schizophrenia spectrum disorders. The authors note that prodromal psychotic symptoms can emerge years before formal diagnosis, so observational claims data often cannot determine whether hallucinogen exposure preceded psychosis, coincided with it, or followed early manifestations of illness. This limits causal inference even when an association is observed.

This was a retrospective cohort study using the Merative MarketScan Commercial and Multi-State Medicaid Databases from 2015 through 2019. The analytic cohort included 6,184 people with at least 1 hallucinogen-related admission and 267,282 people with at least 1 nonhallucinogen substance-related admission, all with at least 180 days of continuous enrollment before the index admission so baseline psychiatric history could be assessed. The primary outcome was any psychotic disorder diagnosed between 30 days and 6 months after the index admission, and the models adjusted for demographics, medical comorbidity, and baseline mood, anxiety, psychotic, and co-occurring substance use disorders.

No. The study found that psychosis-related admissions captured only a subset of patients with baseline psychotic disorders. Specifically, 10,719 of 18,131 people with baseline psychotic disorders had psychosis-related admissions, meaning many others had psychosis identified outside inpatient settings. The authors therefore argue that outpatient psychiatric diagnoses add important confounder control beyond prior admissions alone.

The study has several important limitations. It could not distinguish among hallucinogen subtypes such as psilocybin, LSD, MDMA, PCP, ayahuasca, and ketamine; it included only people with substance-related admissions, so it does not generalize to most recreational hallucinogen use; and the primary follow-up period was 6 months, which may miss longer-term outcomes. The authors also note possible misclassification because hallucinogen detection in hospital settings may be less reliable than testing for opioids, alcohol, stimulants, or cannabis, and observational administrative data cannot fully resolve residual confounding or prodromal illness.