47.06 Geographic and Socioeconomic Distribution of Pediatric Firearm Injuries in Arizona

A. Safavi1, T. O’Keeffe1, R. S. Friese1, B. Joseph1  1University Of Arizona,Division Of Trauma, Critical Care, Emergency Surgery, And Burns, Department Of Surgery,Tucson, AZ, USA

Introduction:
Strict firearm legislation has been shown to decrease pediatric firearm injuries. However other factors may also play part, such as the effect of socioeconomic status.  The aim of this study is to identify the geographic distribution of pediatric firearm injuries in Arizona and examine the association of firearm incidence with household income of Arizonian families. 

Methods:
Patients younger than 18 years old with firearm injuries in Arizona were identified from the 2008-2012 Arizona trauma registry.  Primary outcome of interest was incidence of firearm injury per county and zip code of residence.  Secondary outcome was the socioeconomic status of the household of residence. Using US census bureau data, mean and median income for individual counties and zip codes were obtained.  Descriptive and linear regression analysis was then used to determine the association between socioeconomic status and incidence of pediatric firearm injury.

Results:
577 children with firearm injuries (male: female 493:84, mean age 15.6+3.5) were included, of which 31 (5%) were self inflicted. 279 (46%) of children were Caucasian followed by African American 68 (12%) and Native American 47 (8%).  Among counties, Pima (12.7) Yuma (10.2) and Maricopa (9.7) had the highest number of pediatric firearm injury per 100,000 populations. When analyzing by zip code of residence, 349 (60%) and 479 (83%) of injured children were residing in zip codes below 50 (53,893$) and 75 (67,843$) percentile of Arizona mean household income respectively. This was coherent with subgroup analysis of individual counties.  Residing in zip code with mean household annual income below 50 percentile was found to be predictive of higher pediatric firearm injury incidence in linear regression analysis  (β coefficient, 2.09; 95% confidence interval, 0.5-3.6; p = 0.007).

Conclusion:
Low household income is an independent predictor of pediatric firearm injury. Interventions to specifically target this high risk population may lead to more impactful intervention programs.