Comparison of intracranial computed tomographic (CT) findings in pediatric abusive and accidental head trauma

Background. Child abuse specialists rely heavily on diagnostic neuroimaging. Objectives. Study objectives were: (1) to compare the frequencies of six specific intracranial CT abnormalities in accidental and nonaccidental pediatric head trauma, and (2) to assess interobserver agreement regarding these CT findings. Materials and methods. Three pediatric radiologists blindly and independently reviewed cranial CT scans of pediatric patients who sustained closed head trauma between 1991 and 1994. All patients were less than 4 years of age. Study cases included thirty-nine (50 %) with nonaccidental head trauma and thirty-nine (50 %) with accidental head trauma. Each scan was evaluated for the presence or absence of the following six intracranial findings: (1) interhemispheric falx hemorrhage, (2) subdural hemorrhage, (3) large (non-acute) extra-axial fluid, (4) basal ganglia edema, (5) posterior fossa hemorrhage, and (6) frontal-parietal shearing tear(s). Interobserver agreement was calculated as the percentage of total cases in which all reviewers agreed a specific CT finding was present or absent. Diagnosis required independent agreement by all three pediatric radiologists. The frequencies of these six intracranial CT abnormalities were compared between the two study groups by Chi-square analysis and Fisher's exact test. Results. Interobserver agreement between radiologists was greater than 80 % for all. lesions evaluated, with the exception of frontal-parietal shearing tear(s). Interhemispheric fair hemorrhage, subdural hemorrhage, large (non-acute) extra-axial fluid, and basal ganglia edema were discovered significantly more frequently in non-accidental trauma (P less than or equal to.05). Conclusion. Although not specific for child abuse, discovery of these intracranial CT abnormalities in young patients should prompt careful evaluation of family and injury circumstances for indicators of nonaccidental trauma.