Evaluating multiple alternative responses strategies to 911 calls within Durham's Holistic Empathetic Assistance Response Team (HEART) program: A quasi-experimental study

Abstract

Alternative response programs are designed to bridge the gap between the crisis needs of 911 callers and public safety goals of first responders. Despite widespread implementation, rigorous research on program outcomes remains limited. This study used a quasi-experimental design to examine the incident-level outcomes associated with Durham, North Carolina's Holistic Empathetic Assistance Response Team (HEART). HEART includes three coordinated response approaches: crisis call diversion by clinicians embedded in the 911 call center, unarmed community response teams staffed by clinicians, peers, and medical personnel, and co-response teams pairing clinicians with specially trained officers. Using administrative data, we compared HEART-eligible calls that received a response from a HEART team with similar eligible calls handled by patrol when HEART units were unavailable. Analyses estimate outcomes separately for the community response team (CRT) and co-response (COR) approaches, with a more limited analysis of crisis call diversion. Findings differed by response type. Across CRT and COR approaches, we found modest differences in 911 response times, longer on-scene times on average, and higher likelihood of transports compared to patrol. CRT response was associated with lower probabilities of offense reports, arrests, and citations compared to patrol. COR response was associated with lower probabilities of offense reports, and no difference in arrests compared to patrol response. These findings suggest that alternative response programs can reduce some forms of criminal-legal involvement while preserving timely response, but effects varied by response model and call context.