BACKGROUND: Fatal child abuse can be mistaken for sudden unexpected infant death (SUID) in the emergency department setting. It is unknown if there are characteristics that distinguish abusive from nonabusive causes of SUIDs in the emergency department. METHODS: Using a matched case-control design, we reviewed the medical examiner records of deaths of infants younger than 12 months who were found unresponsive at a residence, required cardiopulmonary resuscitation, and had a complete forensic autopsy between 2009 and 2015. Infants with a manner of death as homicide were cases; controls were those with the manner of death as accident, natural, or undetermined. Each case was matched with 5 controls based on age (months). Differences between cases and controls were evaluated with respect to demographic, parental, and household characteristics and clinical outcomes. RESULTS: We identified 12 cases (homicides) and 169 controls (nonhomicides), of which 60 were selected for the matched analysis. We found no significant differences between cases and controls with respect to age, race, sex, maternal substance use, Child Protective Services involvement prior to death, presence of male head of household, surviving siblings, or emergency medical services transport. Cases were more likely to have Child Protective Services involvement at the time of death (83% vs 38%; P = 0.01), sentinel injuries (odds ratio, 9.67; 95% confidence interval, 1.30-122.43), and return of spontaneous circulation (odds ratio, 29.99; 95% confidence interval, 3.70-241.30). CONCLUSIONS: Child Protective Services agency involvement at time of death, sentinel injury, and return of spontaneous circulation were more often associated with abusive causes of SUID. Further study is needed to confirm these findings.
BACKGROUND: Fatal child abuse can be mistaken for sudden unexpected infant death (SUID) in the emergency department setting. It is unknown if there are characteristics that distinguish abusive from nonabusive causes of SUIDs in the emergency department. METHODS: Using a matched case-control design, we reviewed the medical examiner records of deaths of infants younger than 12 months who were found unresponsive at a residence, required cardiopulmonary resuscitation, and had a complete forensic autopsy between 2009 and 2015. Infants with a manner of death as homicide were cases; controls were those with the manner of death as accident, natural, or undetermined. Each case was matched with 5 controls based on age (months). Differences between cases and controls were evaluated with respect to demographic, parental, and household characteristics and clinical outcomes. RESULTS: We identified 12 cases (homicides) and 169 controls (nonhomicides), of which 60 were selected for the matched analysis. We found no significant differences between cases and controls with respect to age, race, sex, maternal substance use, Child Protective Services involvement prior to death, presence of male head of household, surviving siblings, or emergency medical services transport. Cases were more likely to have Child Protective Services involvement at the time of death (83% vs 38%; P = 0.01), sentinel injuries (odds ratio, 9.67; 95% confidence interval, 1.30-122.43), and return of spontaneous circulation (odds ratio, 29.99; 95% confidence interval, 3.70-241.30). CONCLUSIONS: Child Protective Services agency involvement at time of death, sentinel injury, and return of spontaneous circulation were more often associated with abusive causes of SUID. Further study is needed to confirm these findings.