Geospatial relationship of trauma and violent crime: An analysis of violent crime and trauma center utilization

Abstract

Introduction Descriptive epidemiologic and geographic analysis utilizing geographic information science (GIS) has been used to determine the utilization of trauma systems and to spatially describe patterns of trauma and crime. We examined the relationship between spatial components of criminality and injuries in order to evaluate the optimal trauma center location and determine a correlation between reported violent crime and trauma center utilization. Methods All adult trauma and violent crime (VC) encounters in a defined area over a single year were included. Geospatial statistics pattern analysis tools of Median Center (MC) and the Average Nearest Neighbor analysis (ANNa) were used to determine if mapped points occurred in complete spatial randomness or were clustered in a significant pattern. Results ANNa of VC resulted in a z-score of –20.54 and a p-value of <0.001, indicating a <1% likelihood that violent crimes were distributed randomly. Further ANNa yielded a zscore of –5.67 and p-value of <0.001 for injuries. Our trauma center is 1.45 miles from the MC of VC and 2.28 miles from the MC for injuries. Spatial autocorrelation failed to demonstrate a direct relationship between criminality and trauma center utilization with a z-score of 0.030 and p-value of 0.98. Conclusion While not statistically significant, the spatial trends of violent crime and trauma center utilization demonstrated a clear pattern. GIS is a powerful tool for the trauma director, and examination of the local regional patterns of trauma should be undertaken by health systems to assist with optimizing outreach, expansion, and response times.