Source: Bechtel K, Stoessel K, Leventhal JM, et al. Characteristics that distinguish accidental from abusive injury in hospitalized young children with head trauma. Pediatrics. 2004;114:165–168.A majority of children under the age of 2 who have serious or fatal head trauma have been victims of abuse.1,2 Nevertheless, major head injuries may occur in this same age group from accidental causes. These authors present a prospective study of children less than 2 years of age who were hospitalized at Yale-New Haven Children’s Hospital, Conn, with a diagnosis of head injury and who had computed tomography of the head performed as part of their evaluation. The children were classified as victims of inflicted or accidental head trauma based upon predetermined criteria that did not include the presence or absence of retinal hemorrhages. All patients in the study underwent a dilated ophthalmologic examination by a pediatric ophthalmologist.The 67 children in the “accidental injury” group included 65 who fell (of these, 47 fell ≤4 feet), 1 child whose head was run over by a car, and 1 passenger in a motor vehicle crash. The 15 children in the “inflicted injury” group included 12 for whom no mechanism of injury was provided by the caretakers, 1 who was reported to have fallen 4 inches, 1 who was reported to have hit himself in the head with a piece of wood, and 1 child whose father admitted to shaking the child 48 hours prior to hospitalization.There were no differences between the 2 groups with regard to mean age, gender, or ethnicity. Significantly more patients in the “abusive injury” group had subdural hemorrhage (SDH) (80% versus 27%), retinal hemorrhage (RH) (60% versus 10%), seizures (53% versus 6%), and abnormal mental status at presentation (53% versus 10%). Among children with RH, those due to abuse were significantly more likely to be bilateral (6/9 versus 1/7); extend to the periphery of the retina (4/9 versus 0/7); and include pre-retinal hemorrhage (5/9 versus 0/7), premacular hemorrhage (3/9 versus 0/7), or vitreous hemorrhage (2/9 versus 0/7). Accidentally injured children were significantly more likely than abused children to have a skull fracture (60% versus 27%) or scalp hematoma (51% versus 7%).In recent years, a small but vocal segment of the medical community has charged that the conventional understanding of findings attributed to abusive head trauma are the result of circular reasoning.3–5 Certain findings, eg RH, have been linked with abuse,6 and so, if a child comes in and has RH, we conclude that the child has been abused. Critics assert that adequate controlled trials to test this hypothesis have not been done, and that no attempt has been made to establish whether certain patterns of RH are indicative of abuse.3–5 This study addresses those concerns. Most importantly, the allocation of children to the “inflicted injury” or “accidental injury” groups was based upon findings and considerations other than those being analyzed.We can conclude from the data presented in this article that RH may occur in some accidental injuries, although with far less frequency than they are seen in abused children. When RH does occur in accidental head trauma, it is not likely to present with the findings most typical of the abused child with head trauma, ie, multilayer, bilateral hemorrhages that extend to the periphery of the retina.This study contains important information for the pediatrician who faces the quandary of when to suspect child abuse in the head-injured child. While the authors did not present likelihood ratios, they can be readily calculated from the data they present. While positive likelihood ratios (+LR) for certain history and examination findings cannot conclusively confirm the diagnosis of child abuse (RH, +LR = 6.0 [95% CI, 3.9–9.1]; seizure, +LR = 8.3 [95% CI, 5.2–13.4]; and mental status decrease, +LR = 5.3 [95% CI, 3.3–8.6]), any combination of 2 of these findings virtually rules in a diagnosis of child abuse based on the prevalence of child abuse in head injury that has been previously reported.1