OBJECTIVES : Relapse after an acute episode in schizophrenia remains a major public health problem. This health outcomes study assessed future relapse risk and adherence to antipsychotic (AP) medication among Medicaid patients prescribed atypical long-acting injectables (LAI) vs. oral APs after a schizophrenia-related hospitalization.

METHODS : We conducted a retrospective cohort study using 2010-2016 Marketscan® Medicaid claims data. We identified patients (18-64 years) with a schizophrenia diagnosis, no recent LAI exposure and a prescription fill for an atypical LAI or oral AP within 45 days of a schizophrenia-related hospitalization. Relapse (defined as a schizophrenia-related hospitalization or ED visit) risk and time to relapse from index discharge date were assessed using Kaplan-Meier and Cox proportional-hazard models for up to 12 months after the index date. Adherence patterns, such as proportion of days covered (PDC), discontinuation and persistence, were also assessed during the 12 month follow-up period.

RESULTS : A total of 7,740 patients were included. LAIs were not often started following the acute event (LAI=278[3.6%]), compared with oral APs. Patients who started on LAIs were younger, predominantly male, and had fewer mental-health comorbidities and prior mental-health hospitalizations. After adjusting for covariates, the LAI patients had a 21% lower risk of relapse (p=0.032) and 27% lower risk of re-hospitalization (p=0.013). No differences were observed for ED visits. Mean times to first relapse, re-hospitalization, and treatment discontinuation were 16.2 (p=0.051), 18.7 (p=0.016) and 17.2 (p=0.046) days longer for LAI patients, compared with oral AP patients. In terms of continuity of medication, LAI patients had a 23% lower risk of discontinuation (p=0.005) and slightly better PDC (4%, p=0.014), compared with oral AP patients.

CONCLUSIONS : Initiating on LAIs, compared to oral APs, following an acute psychotic event is associated with a reduced risk of relapse (specifically re-hospitalization) and treatment discontinuation, as well as longer times to these outcomes.