Recognition of Child Abuse: Title and subTitle BreakNotes From the Field

ADVANCES IN medicine are generally the result of a process of repeated observation, investigation, and correlation. These initial descriptive steps are followed by more sophisticated hypothesis generation and testing, which continues the spiral heading toward definitive relationships between symptoms, physical and laboratory findings, and treatment of the condition. The rapidity with which this process progresses is dependent on factors such as disease incidence, morbidity, and professional and public interest.

John Caffey, MD, a radiologist, can probably be credited with initiating medical concern regarding the problem of child abuse. His description of the association of skeletal fractures and subdural hematomas in 6 children made it clear that trauma was the underlying cause, but did not so clearly make the connection to abuse by the parents.¹ Over the next quarter century, through repeated observation, investigation, and correlation, the source of the trauma became more apparent, and the problem was stated clearly in the landmark article by Kempe and coworkers,² "The Battered Child Syndrome," in 1962. Since that time, our knowledge of the condition of child abuse and neglect has expanded greatly as recognition of the condition, and various factors associated with it, has increased.

While the awareness of child abuse has increased the likelihood of including it in one's differential diagnosis of etiology of injury, there are still many situations in which professionals are uncertain of the cause. Physical abuse continues to be a situation in which perpetrators deny guilt, and the victim may be unable to communicate how his or her injuries were received. It is also a condition that many physicians are still loath to identify. Among the reasons for not reporting are denial that abuse could occur, distaste for involvement with protective services agencies, and concern about reporting cases of which they are uncertain. Physicians and nurses seek descriptive information that can help them ascribe some degree of likelihood that a child's injuries might or might not be inflicted. Such supporting data can help health professionals more accurately report or not report cases of suspected abuse.

In this issue of the ARCHIVES are 2 articles that add to our knowledge of childhood injuries and the likelihood of child abuse. Reece and Sege³ report their experience with 6 years of children hospitalized for head injuries. Their data reflect a large cohort seen over several years and reduce confusion by categorizing those cases often considered neglectful (lack of supervision, failure to use a seat belt) as accidental. The categorization as definite abuse or accident and data analysis thereafter attempt to draw a clear line between the two. The findings support the premise that there are characteristics that are seen more frequently in injuries caused by abuse than in accidental cases. Severity of injury and complications such as intracranial bleeding and death were all increased in the abuse group. Of significance to physicians initially encountering these patients are the findings that almost 20% of head injuries in hospitalized children are caused by abuse, and that in more than half of the abuse cases, family members reported no history to account for the injuries. Of importance to all health professionals is the finding that mortality for children hospitalized for head injuries secondary to abuse is high (13%), not only higher than for accidental head injuries but higher than for many hospitalized conditions such as meningitis.

DiScala and colleagues⁴ describe a 10-year retrospective of children hospitalized for trauma across the country. Their data reflect the experience of the National Pediatric Trauma Registry, a sample that seems to reflect the national condition more accurately than a single hospital. They noted that the child abuse group represents more than 10% of those children hospitalized for trauma and found a mortality rate for those hospitalized similar to that seen in the Reece and Sege study. Previously described relationships with child abuse of findings such as prematurity and retinal hemorrhages are given further credence. Other findings were supportive of past reports in the literature concerning the comparative increased severity of injuries, the tendency for younger children to be victims of abuse, and the relative safety of children in day care centers.

An important limitation of both studies relates to the populations examined. These are children with injuries significant enough to warrant hospitalization. Those children seen in health care settings—office settings, urgent care centers, and emergency departments—and not requiring hospitalization are not included in this analysis. Do injuries secondary to child abuse account for an even larger percentage of nonhospitalized cases, or is it smaller, in part because of perpetrators not bringing the child to a treatment facility? While this question cannot be answered by these studies, other conclusions can help us with these cases. The findings suggestive of abuse—especially the lack of explanation for a child's injuries—are important for all physicians to know and consider in evaluating pediatric trauma. The associated morbidity and mortality should make clear to every physician the significance of child abuse as a public health problem. Whether through increased identification of findings suggestive of child abuse or through increased appreciation of the morbidity of the condition, there should be more identification and reporting of suspected abuse.

Observation, investigation, correlation. These studies help renew our attention to the significance of the problem of child abuse and provide additional guidance in our recognition of cases in children younger than 5 years.