Black Children Are Disproportionately Identified as Victims of Child Abuse: A National Trauma Data Bank Study
Background: Child abuse is a significant cause of injury and death among children, but proper identification is often challenging. This study aims to assess whether racial disparities exist in the identification of possible child abuse.
Methods: National Trauma Data Bank (NTDB) 2010-2014 was queried for injured children <18 years old. ICD-9CM codes were used to identify patients with injuries due to suspected child abuse. The 2010 US Census Data was used to determine racial breakdown of the US population. The primary outcome was to determine if certain races were disproportionately identified as child abuse victims. Secondary analyses included comparisons of injury severity score (ISS), length of stay (LOS), and mortality by race.
Results: Between 2010 and 2014, 538,042 trauma patients were included in NTDB. Suspected child abuse victims accounted for 4,288 (0.8%) patients. Of these, 51.2% were white, 32.2% black, 1.2% Asian, 0.2% Native Hawaiian/Other Pacific Islander, 1.8% American Indian, and 13.4% other race. Black patients were disproportionately overrepresented, composing 12% of the US population, but 32.2% of suspected child abuse victims (p<0.001). Among children suspected of abuse, white children were more likely to have a serious ISS (15-25) compared with black children (11% vs. 22%, p<0.01). White children also had higher in-hospital mortality (11% vs. 8%, p=0.01). Despite this, black children were hospitalized longer (6.5 vs. 7.1 days, p< 0.01) even when controlling for ISS (1-15: 4.3 vs. 5.2 days, p< 0.01). There was no difference in ICU LOS.
Conclusion: Child abuse was disproportionately identified among black children. When compared to white children of abuse, black children had lower ISS and hospital mortality, however, had longer hospitalizations even with milder injuries. By identifying racial disparities in child abuse, further studies can focus on understanding the etiology of this disparity, and the possible impact of reporter biases.