Assessing the spatial pattern and temporal stability of violence in Scotland. A data linkage study using ambulance data

ObjectiveViolence is not evenly distributed across space and time highlighting huge disparities across communities. This research aims to improve our understanding of the spatio-temporal patterns of violence in Scotland, and how area level characteristics, such as area deprivation and proximity to alcohol establishments, are associated with differences in violence across communities. MethodsTraditional research has largely focused on cross-sectional spatial measures or aggregated temporal data, which limits our understanding of how violent crime clusters shift over time. Using linked administrative data from ambulance call-outs for assault-related injuries, this project analyses the spatio-temporal patterns of violence in Scotland. Conditional autoregressive models (CAR) are used to identify clusters of violence and how they change over time. These models allow for the integration of spatial dependencies and temporal variations, providing a nuanced understanding of the interplay between violence and area-based characteristics such as deprivation and proximity to alcohol establishments. Results Although overall trends indicate a decline in violence over the study period, this redu