Building capacity for injury prevention: a process evaluation of a replication of the Cardiff Violence Prevention Programme in the Southeastern USA.

OBJECTIVES: Violence is a major public health problem in the USA. In 2016, more than 1.6 million assault-related injuries were treated in US emergency departments (EDs). Unfortunately, information about the magnitude and patterns of violent incidents is often incomplete and underreported to law enforcement (LE). In an effort to identify more complete information on violence for the development of prevention programme, a cross-sectoral Cardiff Violence Prevention Programme (Cardiff Model) partnership was established at a large, urban ED with a level I trauma designation and local metropolitan LE agency in the Atlanta, Georgia metropolitan area. The Cardiff Model is a promising violence prevention approach that promotes combining injury data from hospitals and LE. The objective was to describe the Cardiff Model implementation and collaboration between hospital and LE partners. METHODS: The Cardiff Model was replicated in the USA. A process evaluation was conducted by reviewing project materials, nurse surveys and interviews and ED-LE records. RESULTS: Cardiff Model replication centred around four activities: (1) collaboration between the hospital and LE to form a community safety partnership locally called the US Injury Prevention Partnership; (2) building hospital capacity for data collection; (3) data aggregation and analysis and (4) developing and implementing violence prevention interventions based on the data. CONCLUSIONS: The Cardiff Model can be implemented in the USA for sustainable violent injury data surveillance and sharing. Key components include building a strong ED-LE partnership, communicating with each other and hospital staff, engaging in capacity building and sustainability planning.

Ability of crime, demographic and business data to forecast areas of increased violence.

Identifying geographic areas and time periods of increased violence is of considerable importance in prevention planning. This study compared the performance of multiple data sources to prospectively forecast areas of increased interpersonal violence. We used 2011-2014 data from a large metropolitan county on interpersonal violence (homicide, assault, rape and robbery) and forecasted violence at the level of census block-groups and over a one-month moving time window. Inputs to a Random Forest model included historical crime records from the police department, demographic data from the US Census Bureau, and administrative data on licensed businesses. Among 279 block groups, a model utilizing all data sources was found to prospectively improve the identification of the top 5% most violent block-group months (positive predictive value = 52.1%; negative predictive value = 97.5%; sensitivity = 43.4%; specificity = 98.2%). Predictive modelling with simple inputs can help communities more efficiently focus violence prevention resources geographically.