(PS6-99) Exploring Predictors of Treatment Dropout in Primary Care Behavioral Health Using Andersen’s Behavioral Model of Health Services

Introduction: Approximately one in five adults in the United States report suffering from mental health problems, yet fewer than half of those in need receive appropriate services. Results from a recent meta-analysis indicate that on average, 19.7% of individuals drop out from outpatient mental health services, suggesting that even when individuals seek treatment, they may not be engaging long enough to receive full benefits. Andersen’s Behavioral Model of Health Services describes three types of factors that may contribute to individuals ending services prematurely: predisposing (e.g., demographics), enabling (e.g., distance from clinic), and need (e.g., symptom severity) factors. One service delivery model that is thought to make mental health care more accessible is the Primary Care Behavioral Health (PCBH) model, where behavioral health consultants (BHCs) are embedded into primary care teams. However, research on predictors of dropout in this model is limited. The present study aims to examine predisposing, enabling, and need factors as predictors of dropout in PCBH.

Method: Data are taken from a larger study comparing help-seeking pathways and treatment experiences between community mental health clinics and PCBH clinics in South Central Appalachia. PCBH patients who initiate mental health services via a same-day warm handoff from their primary care provider and schedule at least one follow-up visit with the BHC are eligible for the present study, which involves a baseline interview and monthly follow-up contacts for up to four months. Predisposing factors assessed include age, sex, minority status, and stigma surrounding mental health care. Enabling factors assessed include travel time to clinic, cost of services, insurance status, perceived access to care, satisfaction with care, and therapeutic alliance. Need factors assessed include mental health symptoms and readiness for change. Dropout is assessed using a self-report item administered at the participant’s final study contact, in which they are asked if they are still enrolled in treatment, have discontinued by mutual agreement with their clinician, or have dropped out. Participant responses will be confirmed using data from electronic health records. The bivariate relationships between each predictor variable and treatment dropout will be examined (using two-samples t-tests for continuous predictor variables and chi-square analyses for dichotomous predictor variables).

Results: Data collection is currently underway; final results will be available in November. Currently enrolled participants (n = 13 to date) range in age from 18 to 75 years (M = 43.77, SD = 19.39). Most identify as white (92.3%) and non-Latino/a/x (100.0%).

Discussion: Prior research suggests predisposing, enabling, and need factors contribute to dropout in outpatient mental health settings. These factors likely play a similar role in PCBH settings, however, the research base in this area is lacking. Premature discontinuation of services has negative outcomes for both patients and the healthcare system. Given the negative effects of ending treatment early and the paucity of research on these domains in PCBH, more research addressing these concerns is needed.