Previous emergency department use among homicide victims and offenders: a case-control study.

STUDY OBJECTIVE We differentiate risk factors for future homicide victimization and offending, and we measure emergency department (ED) use among homicide victims, offenders, and controls. METHODS The design was a matched case-control study conducted in Bernalillo County, NM, and its university-affiliated health sciences center and hospital. All Bernalillo County homicide victims (N=124) and offenders (N=138) identified between January 1996 and December 2001 who were linked to university physician billing records and who had health care use during the 3 years before the homicide incident were included as cases. Randomly selected age-matched (+/-1 year) and sex-matched subjects with health care use within 3 years of their matched pair's homicide were included as controls. Main outcome measures were the number and type of ED visits by cases and controls. RESULTS Among the 124 victims and 138 offenders who used health care, most were men (80%) and averaged 27.7 years of age. Victims and offenders had similar health care use and were grouped for final analyses. Cases (victims and offenders) were more likely to have had an ED visit within 3 years of the homicide (85%) compared with controls (59%) (odds ratio [OR] 4.3; 95% confidence interval [CI] 3.0 to 6.2). Within previous ED visits, assault (OR 4.5; 95% CI 2.9 to 7.0), firearm injury (OR 13.6; 95% CI 4.9 to 37.7), and substance abuse (OR 3.7; 95% CI 2.2 to 6.0) were associated with future homicide. ED visits clustered closer to the homicide incident for cases (median days before the homicide -402 days; 95% CI -434 to -364) compared with controls (median -487 days; 95% CI -498 to -474). CONCLUSION Patients with ED visits for assault, firearm injuries, and substance abuse are at increased risk for homicide and often have an escalating number of visits leading up to the homicide event. ED-based identification and referral programs similar to those used for intimate partner violence or other preventive strategies should be considered for this high-risk population.