An EMS-Based Crisis Response Model for Mental Health-Related EMS Calls: A Quasi-Experimental Study.
Psychiatric services (Washington, D.C.) · 2026-04-17 · Quasi-experimental
Abstract
OBJECTIVE: The Behavioral Health Emergency Assistance Response Division (B-HEARD) program is an emergency medical services (EMS)-based alternative to police-led responses for nonviolent mental health crises in New York City. This study aimed to examine B-HEARD's effects on nonviolent mental health-related EMS call rates. METHODS: Data from the EMS Incident Dispatch Data file (January 2019-December 2024) were analyzed. The sample included 76 New York Police Department precincts, of which 31 adopted B-HEARD during the study period; the remaining 45 precincts did not adopt the program. A staggered-adoption, difference-in-differences design was used to compare monthly precinct-level rates of nonviolent mental health-related EMS calls between precincts that adopted B-HEARD and those that did not. Estimates were disaggregated by implementation wave and time since adoption. RESULTS: Estimated effects were heterogeneous over time and across implementation waves. When aggregated by implementation wave, early-adopting precincts did not show statistically significant changes. By contrast, statistically significant reductions in call rate were observed among B-HEARD-adopting precincts in March 2023. When aggregated by calendar month, statistically significant call-rate reductions were concentrated in certain months during 2024. Dynamic event-time estimates suggested that, when present, reductions emerged approximately 1 year after program implementation. CONCLUSIONS: The findings suggest that EMS-based, nonpolice crisis response programs may reduce nonviolent mental health-related EMS calls under specific operational and contextual conditions. The heterogeneous and delayed effects observed provide a benchmark for future evaluations aimed at identifying when and where these models translate into measurable, system-level change.