OBJECTIVE: To compare mortality risk in patients with Parkinson’s disease psychosis (PDP) after initiation of pimavanserin or comparator atypical antipsychotics.
BACKGROUND: Pimavanserin is approved in the US to treat hallucinations and delusions associated with Parkinson’s disease psychosis (PDP). All US antipsychotic labels include boxed warnings of mortality risk in older patients with dementia-related psychosis.
METHODS: This cohort study identified patients aged ≥65 years with PDP initiating pimavanserin or a comparator atypical antipsychotic in US Medicare claims (2016-2021). A subcohort of LTC/SNF residents was identified. Cox proportional hazards models were used to estimate hazard ratios (HRs) comparing all-cause mortality in propensity score–matched treatment groups. Cumulative incidence curves, time period–specific relative risk, and risk difference estimates evaluated risk over time.
RESULTS: This study identified 4,384 pimavanserin initiators and 28,042 comparator initiators in the overall PDP cohort and 921 pimavanserin initiators and 7,963 comparator initiators in the LTC/SNF subcohort. After propensity score matching, the overall PDP cohort included 4,381 patients per treatment group, the LTC/SNF subcohort included 905 patients per treatment group, and all baseline characteristics were well-balanced. The matched HR for mortality (pimavanserin versus comparator) was 0.76 (95% CI, 0.68-0.85) in the overall PDP cohort and 0.90 (95% CI, 0.74-1.10) in the LTC/SNF subcohort. In the overall PDP cohort, time period–specific relative risks and risk differences showed that pimavanserin initiators had a lower risk of mortality during the first 365 days of follow-up.
CONCLUSIONS: This active-comparator study suggests an overall lower mortality risk in older patients with PDP treated with pimavanserin than in patients treated with other atypical antipsychotics through at least the first year of treatment. No meaningful differences in mortality risk were observed between groups in the LTC/SNF subcohort. Regardless of treatment, mortality was higher in the LTC/SNF subcohort than in the overall PDP cohort.
