OBJECTIVES: Over 50% of Parkinson’s disease (PD) patients develop psychosis over the course of the disease. This study evaluated patients who were prescribed antipsychotics to treat PD with psychosis.

METHODS: This retrospective claims data cohort study examined commercial and Medicare Advantage enrollees with PD and ≥1 claim for aripiprazole, clozapine, olanzapine, pimavanserin, quetiapine, or risperidone from June 2016 – June 2017; the first claim was the index date. Patients had continuous insurance enrollment for 12 months pre-index and ≥6 months post-index; patients in cohorts except pimavanserin had ≥2 pre-index PD diagnoses.

RESULTS: The sample comprised 919 patients: 47 aripiprazole, 4 clozapine (not analyzed), 79 olanzapine, 120 pimavanserin, 575 quetiapine, 94 risperidone. Mean age was 76.9 years; 54.1% were male. Patients initiating pimavanserin had lower mean Charlson comorbidity index score (2.2 versus 2.5-3.1; p<0.05), and lower percentage with claims for tremor/abnormal movements (8.3 versus 16.2-26.6%; p<0.01). Higher percentages of pimavanserin patients had PD-related office visits during pre-index (82.5% versus 62.8-73.9%; p<0.01) and post-index (1.87/person-year versus 0.83-1.7; p<0.001), and outpatient facility visits during post-index (1.10/person year versus 0.68-0.91; p<0.05). Pimavanserin patients had higher average pre-index PD medication use (prescriptions for 1.8 distinct PD medications versus 1.14-1.39; p<0.001) and higher post-index utilization rates for levodopa (3.48/person year versus 1.54-3.35; p<0.001), MAO-B inhibitor (0.25/person-year versus 0.05-0.11; p<0.001), and amantadine or droxidopa (0.23/person-year versus 0.07-0.09; p<0.01). The pimavanserin cohort had higher mean post-index all-cause per patient per month (PPPM) pharmacy costs ($2,523 PPPM versus $580-920; p<0.001), but lower medical costs ($1,451 PPPM versus $1,816-$2,983; p<0.01).

CONCLUSIONS: The pimavanserin cohort had better pre-index health status and higher engagement in PD-related ambulatory and pharmacy management during both pre-index and post-index periods. Pimavanserin patients had lower mean all-cause medical costs. All-cause and PD-related pharmacy costs were higher, likely due to pimavanserin’s brand status and high PD medication use.