Injuries/Violence
Comparison of Interpersonal Violence and Homicide Rates Using FBI’s Uniform Crime Report (UCR) Data, Hospital Injury Data and Vital Statistics Across California Cities, 2007-2017. Saron Goitom* Saron Goitom Shelley Jung Jennifer Ahern
Interpersonal violence, including fatal and nonfatal injuries, is a critical public health problem in the US. Yet, the longstanding surveillance system for violent crime in the US, FBI’s Uniform Crime Reports (UCR), has limitations, including pervasive missingness, and unreliable estimates at smaller geographic levels. Leveraging hospital records and vital statistics may be an alternative source to identify interpersonal violence. In this study, we examined the agreement between violent crime as measured by FBI UCR versus hospital records and vital statistics in California.
We gathered violent crime data (murder, rape, aggravated assault, and robbery) from the FBI’s UCR, nonfatal interpersonal injury data from hospital discharge records, and homicide data from vital statistics across California cities (incorporated places) for 2007-2017. For both crime (FBI) and health-based (hospital records and vital statistics) violence data, annual interpersonal violence and homicide rates were calculated as a count per 100,000 people for each city with complete data across the study period (418 cities with complete violent crime data; 79 cities with complete homicide data) and averaged across all cities for each year. We examined trends over time, and calculated spearman correlation coefficients to assess the level of agreement between the two measures. This analysis was replicated for the top 4 largest cities in California.
Across the study period, there was high correlation between crime and health-based interpersonal violence measures (R = 0.90), with less agreement between homicide measures (R = 0.60). However, for the top 4 largest cities, there was high correlation between the homicide measures [Los Angeles (R = 0.94), San Diego (R = 0.88), San Jose (0.78), San Francisco (R = 0.64)], while the correlation between the interpersonal violence measures for these same cities varied [Los Angeles (R = 0.76), San Diego (R = 0.80), San Jose (-0.28), San Francisco (R = 0.03)]. Further investigation into these differences is important to accurately measure interpersonal violence, and critical in determining where public health intervention and resources are needed most.