LAUSR

Introduction The medical diagnoses and frequency of emergency department visits made by children who are later given a diagnosis of maltreatment do not differ much from those of nonabused children. However, the type of medical complaints and frequency of emergency medical services (EMS) use by child homicide victims before their death are not known. We compared EMS use between child homicide victims and children who died from natural causes before their death. Methods This was a retrospective case-control study of children 0 to 5 years old who died in Houston, Texas, from 2005 to 2010. Cases were child homicide victims. Controls were children who died from natural causes. We reviewed death data and EMS and child protective services (CPS) encounter information before the victim's death. The association between death type (natural vs homicide) and EMS use was assessed using Poisson regression with EMS count adjusted for exposure time. Results There were 89 child homicides and 183 natural deaths. Age at death was significantly higher for homicides than natural deaths (1.1 vs 0.2 y, P < 0.001). Homicide victims used EMS services (39% vs 14%, P < 0.001) and had previous CPS investigations (55% vs 7%, P < 0.001) significantly more often than children who died from natural causes. Poisson regression, after adjustment for age, revealed that the homicide group had more EMS calls than the natural death group (β = 0.55; 95% confidence interval, 0.04–1.07; P = 0.03). However, the EMS use frequency and working assessments were not helpful in identifying maltreatment victims. Conclusions Child homicide victims use EMS more often and have a higher number of CPS investigations before their death than children who die from natural causes. However, the frequency and nature of EMS medical complaints are not helpful in identifying maltreatment.