Abstract
BACKGROUND: Discharge from early psychosis intervention is a critical stage of treatment that may occur for a variety of reasons. This study characterizes reasons for discharge among participants in early psychosis intervention programs participating in the Early Psychosis Intervention Network (EPINET) which comprises >100 programs in the United States organized under 8 academic hubs. STUDY DESIGN: We analyzed 1787 discharges, focusing on program completion, unilateral termination by the client/family, and lost contact with the client/family. We performed exploratory analyses of demographic, clinical, and functional predictors of discharge reason. Variables predictive of discharge type were included in multilevel logistic regressions, allowing for the estimation of predictors of discharge reason and variability in rates by program and hub. STUDY RESULTS: An estimated 20%-30% of enrolled patients completed the program. Program completion rates were higher among participants who were older on admission, had lower negative symptoms severity, spent more time in education, employment, or training, and who were covered by private insurance (a close proxy for socioeconomic status). Programs were more likely to lose contact with male participants, Black participants, and participants who were never covered by private insurance. After accounting for patient-level factors, there was substantial program-level variation in all 3 discharge outcomes, and hub-level variability in the proportion of participants who completed the program. The impact of race on program completion varied substantially by program. CONCLUSIONS: Participants were discharged from early psychosis intervention services for diverse reasons, some of which were associated with sociocultural factors. Disengagement is a widespread problem affecting all hubs.