Can Routinely Collected Ambulance Data about Assaults Contribute to Reduction in Community Violence?

• In Event: Innovative Approaches for Assessing Crime at Place

Abstract

Can police-emergency-department interagency data-sharing be used to reduce community-violence using a hotspots methodology? We present the results of a 12-month (2012) analysis of spatiotemporal clusters of police and emergency calls for service using hotspots methodology and assessing the degree of incident overlap. 3,775 violent crime incidents and 775 assault incidents analysed using spatiotemporal clustering with k-means++ algorithm and Spearman’s rho. We show that spatiotemporal location of calls for services to the police and the ambulance service are equally highly concentrated in a small number of geographical areas, primarily within intra-agency hotspots (33% and 53%, respectively) but across agencies’ hotspots as well (25% and 15%, respectively). Datasets are statistically correlated with one another at the 0.57 and 0.34 levels, with 50% overlap when adjusted for the number of hotspots. At least one in every two police hotspots does not have an ambulance hotspot overlapping with it, suggesting half of assault spatiotemporal concentrations are unknown to the police. Out findings suggest that data-sharing could lead to both reduced community violence by way of prevention, particularly of more severe assaults, and improved efficiency of resource deployment.

Authors

• Cristobal Weinborn, University of Cambridge

• Barak Ariel, Cambridge University / Hebrew University