Initiation patterns of long-acting injectable antipsychotics in a claims-based proxy cohort approximating first-episode schizophrenia: A longitudinal descriptive study using a Japanese claims database

Schizophr Res. 2026 Jul 7;296:67-74. doi: 10.1016/j.schres.2026.07.004. Online ahead of print.

ABSTRACT

Long-acting injectable antipsychotics (LAIs) may support treatment continuity in schizophrenia, yet real-world initiation patterns during the first-episode schizophrenia (FES) phase remain poorly characterized in Japan. Using the JMDC Claims Database (January 1, 2005-March 31, 2024), we conducted a longitudinal descriptive epidemiologic study of a claims-based proxy cohort approximating FES, defined by an operational diagnosis algorithm, temporal proximity between diagnosis and antipsychotic treatment initiation, and continuous enrollment with a pre-index lookback period. Suspected schizophrenia diagnoses were excluded. Primary descriptive measures were LAI initiation (≥1 LAI dispensing) during follow-up, time from schizophrenia diagnosis to first LAI dispensing, and annual antipsychotic dispensing trends. In an exploratory descriptive analysis restricted to patients switching from oral antipsychotics (OAP) to LAIs with 1-year observability before and after LAI initiation, we described treatment continuity (weekly proportion of days covered [PDC]) and psychiatric hospitalization. Among 24,226 patients, the median follow-up duration was 30 months (IQR, 13-58), and 437 (1.8%) had ≥1 LAI dispensing. The median time from diagnosis to first LAI dispensing was 12 months (IQR, 2-31), and the annual proportion with any LAI dispensing remained consistently <2% across the study period. Among OAP-to-LAI switchers, median PDC in the year before LAI initiation was 0.82 (IQR, 0.50-0.96), and 26.9% had hospitalization for medical care and protection during that period. In a Japanese insured claims population, LAI initiation was infrequent and typically occurred after OAP initiation. These findings describe LAI initiation patterns in newly observed and newly treated schizophrenia within an insured claims population.

PMID:42413308 | DOI:10.1016/j.schres.2026.07.004