Using EMS data to identify persons at high risk for violent injury: Implications for surveillance and intervention

Objectives: Concern about mass violence has led to calls for increased mental health support for violence prevention. Health information systems may help identify persons at risk of violence who may benefit from mental support services. Emergency medical services (EMS) serve an increasingly vulnerable population that, compared to other ED users, has high mental health problems. We explored the potential utility of EMS data in identifying at-risk persons. Methods: We reviewed Baltimore's 9-1-1 dispatch data and fire department EMS records over 23 months (2008-10). Health data was categorized and violent trauma related incident data was analyzed. We examined associations of subtypes of violent incidents to patient demographics, repeat EMS use, and evidence of behavioral (mental or substance use) problems. Results: We identified 7,190 violent trauma incidents linked to patient records, comprising 57.7% assaults; 13.3% shootings; 11.8% stabbings; 2.5% abuse/sexual assaults; and 14.7% attempted or threatened self-harm including suicide. Young Black males were over-represented for all incident types except abuse/sexual assault and self-harm, for which there was a wider age range, and whites, and only moderately males, were over-represented. There was evidence of repeat EMS use and behavioral health problems for one-fifth of all persons; of persons involved in self-harm, 35.2% were repeat users and 82.3% had behavioral health problems. Conclusions: Our findings suggest EMS data may aid in identifying persons at risk of violence to themselves or others. High levels of behavioral health problems and emergency care recidivism make them important targets for enhanced assessment and preventive intervention.