Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Research Evidence topic
No spam. Unsubscribe anytime.
Researchers tell committee the evidence is mixed but non‑police responses most consistently reduce ED transports
Summary
Two university researchers told the Community Safety Committee that randomized and controlled evidence remains limited, but studies suggest non‑police mobile crisis responses and some co‑responder models reduce emergency‑department transports and on‑scene arrests. They urged routine evaluation and locally aligned performance measures.
University researchers summarized the emerging evidence on crisis response models for the Community Safety Committee on Oct. 29, telling lawmakers that the literature supports certain operational goals but has limited ability to draw causal conclusions across diverse local contexts.
Dr. Evan Louder (George Mason University) said the literature falls into three broad models: police‑only (often CIT‑trained officers), co‑responder teams (police plus clinicians/EMS) and non‑police mobile crisis teams (clinicians and peers responding without officers). "The state of the literature is very limited in its rigor and ability to make causal inferences," Louder said, but he noted that non‑police responses have the most consistent evidence for reducing emergency‑department transport and achieving on‑scene resolution without arrest. He recommended routine, local evaluation and the establishment of clear dispatch pathways, on‑scene diversion options and continuity of care after contact.
Dr. James Pine (Stanford) emphasized program counts and recent empirical findings: "As of 2024, there were about 1,100 mental health crisis response programs operating in the U.S., and about 650 of those were dispatched directly through 911 or 988," he said. Pine described studies from Denver, Eugene (Cahoots), San Mateo County and others that show reductions in involuntary detention and emergency transports for some programs; effects on long‑term criminal justice outcomes are less consistent.
What researchers recommended: (1) set clear program goals and match metrics to those goals (e.g., on‑scene resolution, ED diversion, client linkage to care), (2) invest in routine data collection and evaluation capacity, and (3) study how design elements (dispatch method, rural vs. urban delivery, clinician composition, follow‑up services) influence outcomes in context.
Policy relevance: researchers told the committee that strong local implementation fidelity and interagency data sharing are critical to measure success and guide scaling decisions; they cautioned against assuming a single model fits all communities.
