Predictors of Service Engagement and Disengagement in Community-Based Coordinated Specialty Care for Early Psychosis in Texas.

Abstract

Coordinated specialty care (CSC) is an evidence-based, early intervention approach for individuals with a recent onset of psychosis. This study identifies program and individual characteristics that predict (1) initial engagement in care; (2) the use of peer and family partner services; and (3) time to disengagement. Administrative data representing CSC encounters at 22 community mental health programs were analyzed. Logistic regressions were used to model initial CSC engagement, as well as the use of peer or family partner services, given initial engagement. Cox proportional hazards regression was used to quantify program disengagement, with possible time-varying effects of peer service use. Identified predictors of initial engagement included: race/ethnicity, age, diagnosis, program urbanicity, and program maturity. Identified predictors of disengagement in the first year included: diagnosis, program urbanicity, and program maturity. Peer and family partner services did not significantly predict disengagement. These results suggest equity issues for some individuals, while also highlighting the importance of program characteristics on pathways to care. Future research should include program factors as key predictors for engagement, explore both person-centered and program-centered strategies to maximize engagement, and optimize and examine the role of individual and family peer services in engagement. Such a shift would align with the call to understand pathways to care from a community-level rather than individual lens.