Geographic association of liquor licenses and gunshot wounds in Chicago.

BACKGROUND The association between alcohol and interpersonal violence is well established. Up to 80% of homicide perpetrators and victims are known to have used alcohol before the incident. However, the association between proximity to a liquor-selling establishment and gun violence is more controversial. METHODS Scene address data from the Illinois State Trauma Registry from 1999 to 2009 were used to geocode all gunshot wounds (GSWs) presenting to trauma centers in Chicago during the study period. These data were linked to publicly available US Census Demographic Data and City of Chicago Liquor Board data. A combination of ordinary least squares and geographically weighted regression was performed to identify "risk regions" throughout the study area. Logistic regression analysis was then performed to assess the independent effect of proximity to an establishment with a liquor license (LL) on trauma center admissions for GSWs. RESULTS A total of 11,744 GSWs were geocoded. No association between LLs and GSWs was identified for the city overall (odds ratio [OR] .97, 95% confidence interval [CI] .96 to .99). However, 5 distinct regions of influence between LLs and GSWs were found. In regions with the highest association, likelihood of a GSW near a packaged LL was extraordinarily high (OR 518.08, 95% CI 10.23 to 1,000), and tavern LLs were also very significant (OR 21.51, 95% CI 1.81 to 255.53). CONCLUSIONS We found that proximity to an establishment with an LL was a strong independent predictor of GSW incidence for many areas of the city, even after controlling for neighborhood characteristics. However, this association was not demonstrable for the entire city, and, in fact, marked regional variation was apparent. These data may contribute to our understanding of the interplay between alcohol and violent injury disparities.