Abstract for: Ten years of Modeling to Learn: Improving flow through psychotherapy and pharmacotherapy in behavioral health systems

Specialty mental health at VHA help Veterans access evidence-based treatment for PTSD and SUD to reduce prolonged suffering and decrease suicide risk. Providers face overwhelming caseloads, workforce shortages, and increased burnout causing bottlenecks and treatment gaps. These challenges stem from stock and flow structures, dynamic feedback loops, time delays, and various emergent behaviors. We use system dynamics to identify leverage points and address these dynamic problems apart from hiring/adding resources. Mental healthcare can be conceived of as a leaky aging chain, with evidence-based practice (EBP) dictating a specific number of appointments per month for x months for specific treatments, with some percentage of patients failing to continue after each appointment. Despite substantial incentives and years of effort to encourage providers to follow EBP, most instead supply a “maintenance dose” of therapy (where a provider sees more patients, but at a sub-optimal frequently). For psychologists treating PTSD, modeling time-bound therapy phases providers to better understand how treatment patient-by-patient decisions unknowingly create accumulations that drive down care access and quality over time. For psychiatrists, adjusting the number of weeks between return visits within clinical practice guidelines, adjustments in medication management cohort (e.g., pts with depression), can have outsized impacts on providers’ ability to meet the needs another (e.g., pts with OUD). Modeling to Learn (MTL) was developed using participatory modeling. Data and modeling infrastructure was scaled nationally in 2017. Over time, MTL releases evolved from supporting clinical decisions of frontline psychiatrists, psychologists, social workers and nurses for licensure credit, to supporting operations improvement decisions of healthcare leaders. MTL open source guides, code, model documentation, and video explainers are available in both frontline facilitation and leadership consult releases on GitHub.