Abstract for: Shaping policy and capital investment in Oregon’s acute mental health system using a hybrid discrete event / system dynamics simulation model

Oregon's behavioral health system faces a severe crisis—ranking first nationally in adult mental illness prevalence (30.2%) and near the top for substance use disorders (22.6%). Capacity constraints create cascading bottlenecks, leaving patients boarding in emergency departments (EDs) or discharged to inappropriate care levels. System complexity makes identifying effective interventions difficult, and traditional benchmarking fails to capture local dynamics. This project uses a hybrid discrete event simulation and system dynamics (DES-SD) model in Simio describing patient flow through Oregon's acute mental health continuum across seven regional trauma regions. Individual patients carry attributes (diagnosis, age, legal status, acuity) that drive routing and length-of-stay decisions. Parameters are drawn from statewide administrative data, electronic health record extracts, and 66 key informant interviews, with iterative validation by a statewide expert panel. The model is substantially complete and calibrated against ED boarding times, state hospital waitlist lengths, and residential occupancy rates. Approximately 100 policy scenarios have been generated, covering bed supply expansion, residential treatment capacity additions, mobile crisis expansion, and state hospital access changes. An interactive dashboard is in development, with findings already shared with Oregon policymakers and health system leaders. This project advances hybrid DES-SD methodology by demonstrating how object-oriented sub-models enable statewide simulation without sacrificing individual-level fidelity. It extends prior psychiatric bed modeling by incorporating patient heterogeneity, legal system interfaces, and regional variation. The interactive scenario dashboard is designed for ongoing use by Oregon policymakers, supporting data-driven decision-making under conditions of system complexity and uncertainty.