Natural history and clinical outcomes in patients with Alzheimer's disease-related psychosis by antipsychotic treatment status in the United States.

Background: While some literature on clinical outcomes in persons with dementia-related psychosis exists, little is known regarding Alzheimer's disease-related psychosis (ADP).

Objective: Describe demographic/clinical characteristics of adults with ADP and estimate incidence of clinical events by antipsychotic treatment status.

Methods: This cohort study identified adults ≥65 years with Alzheimer's disease and incident psychosis (US Medicare database [2013-2018]) and no prior exposure to antipsychotics. Two nonmutually exclusive ADP subcohorts included: patients who initiated treatment with antipsychotic medications (antipsychotic users) and those who remained untreated (antipsychotic nonusers). Baseline characteristics were evaluated before psychosis diagnosis in untreated patients and before antipsychotic initiation in treated patients. Incidence rates were estimated for falls and fractures (composite and separately), seizure/epilepsy (new onset and any), and mortality.

Results: 145,333 ADP antipsychotic nonusers and 49,452 antipsychotic users were identified. Both cohorts had similar baseline demographics; however, antipsychotic users versus nonusers had higher baseline skilled nursing facility use (40.3% and 27.8%), mood (72.7% and 62.1%) and anxiety (70.9% and 57.3%) disorders, falls/fractures (39.5% and 33.8%), urinary tract infections (55.1% and 47.0%), and frailty index scores (76.0% and 69.7%). Crude incidence rates (95% confidence interval)/100 person-years in antipsychotic users and nonusers were 70.0 (68.9-71.2) and 55.8 (55.4-56.1) (falls/fractures composite), 69.0 (67.9-70.1) and 54.9 (54.5-55.2) (falls), 38.6 (38.1-39.0) and 33.0 (32.7-33.2) (mortality), and 45.8 (44.9-46.7) and 54.2 (53.9-54.6) (any seizure/epilepsy).

Conclusions: Antipsychotic initiators with ADP had a higher burden of some baseline comorbidities; experienced higher incidence of falls, fractures, and mortality; and had lower incidence of seizure/epilepsy than antipsychotic nonusers.