September 1985

Should Child Abuse and Neglect Be Considered When a Child Dies Unexpectedly?

Author Affiliations

Janet Chiaramonte
From the Departments of Pediatrics (Drs Christoffel and Zieserl and Ms Chiaramonte) and Community Health and Preventive Medicine (Dr Christoffel), Northwestern University Medical School Chicago; and the Division of General and Emergency Pediatrics, Children's Memorial Hospital, Chicago (Drs Christoffel and Zieserl and Ms Chiaramonte).

Am J Dis Child. 1985;139(9):876-880. doi:10.1001/archpedi.1985.02140110030023

• Deaths during two years at a pediatric teaching hospital were studied to develop guidelines for clinicians who must decide when to explore the possibility of child abuse or neglect when a child dies unexpectedly. Unexpected deaths were defined as deaths occurring before arrival at the hospital or within ten days of hospitalization in children past the first month of life and unrelated to any previously known congenital anomaly or medical condition. There were 43 such deaths during two years. The majority (27) were due to natural causes. However, nine deaths were related to suspected child abuse or neglect (SCAN) and in three of those cases evidence of injury was found only at autopsy. The SCAN cases were more often mistaken for medical illness than for uninflicted injury, indicating that autopsy and postmortem skeletal survey are warranted whenever a child dies unexpectedly. The two factors "dead on arrival" and "1 year of age or less" had a positive predictive value for SCAN of 89% and a negative predictive value of 94%. The two factors identify a high-risk group requiring at least hospital-based Investigation into the possibility of abuse or neglect. Reporting for SCAN is warranted when (1) unsuspected trauma Is found post mortem, (2) there is direct physical or social evidence of child abuse or neglect, or (3) the child is in the high-risk group and hospital-based investigation fails to eliminate the possibility that maltreatment contributed to the child's death.

(AJDC 1985;139:876-880)