Use of a gun prevalence proxy measure to examine the spatial distribution of homicide in urban los angeles county, 2000-2009

Introduction: Homicide is a leading cause of death in Los Angeles County with gun homicide accounting for approximately 80% of homicide deaths. This study analyzes the relationship between a gun prevalence proxy and homicide deaths at the census tract level in the urban areas of Los Angeles County, California. Methods: Homicide in the contiguous, residential, urban census tracts in Los Angeles County was examined using death certificate data for residents of Los Angeles County (2000-2009), the Population Estimates and Projection Systems (PEPS), and 2000 Census – Sample Data. Race, education, poverty, and a previously validated proxy measure for gun prevalence were examined at the census tract level. Poisson log-linear regression with a working spatial covariance structure incorporating Euclidean-based distance weights was used to account for spatial effects. Results: After adjusting for race, education, poverty, and spatial effects, the gun prevalence proxy remained significantly associated with homicide at the census tract level (p <0.007). Multivariable analysis showed that race, poverty, education, and the proxy measure of gun prevalence were all significantly associated with homicide when adjusting for model covariates. After adjusting for spatial effects, race, education, and the gun prevalence proxy all remained significantly associated with homicide. Conclusions: Although the proxy measure of gun prevalence was positively associated with homicide at the census tract level, the measure underestimates homicide in predominately black, urban neighborhoods in Los Angeles. Further study is needed to develop a robust measure of gun prevalence suitable for small area analysis in residentially segregated urban areas.