Mortality from Child Abuse is Underestimated
Abstract & Commentary
Synopsis:Review of North Carolina medical examiner records revealed that the actual rate of death from child abuse was more than three times higher than officially listed in state statutes.
Source: Herman-Giddens ME, et al. Underascertainment of child abuse mortality in the United States. JAMA 1999;282:463-467.
Child abuse is often difficult to diagnose. parents seldom provide truthful explanations about how the injuries have occurred, and physicians often find it difficult to consider such a diagnosis. When a child dies of injuries, it can be even more difficult to confront a family about the possibility of abuse.
It is this issue of the underascertainment of fatal child abuse that is addressed by Herman-Giddens and colleagues. They reviewed the North Carolina Medical Examiners (ME) system to locate cases of homicides of children younger than 11 years of age due to suspected abuse. Over a 10-year period (1985-1994), Herman-Giddens et al found that 220 of the 259 homicides listed in the ME’s registry were due to abuse. This figure was 3.2 times higher than the 68 children who were officially listed in the state’s statistics as dying from abuse. Thus, there was a substantial underreporting (by 59%) of child abuse deaths of all the homicides. The rate of deaths due to abuse increased 12.5% per year from 1.5 per 100,000 children in 1985 to 2.8 in 1994. African-American children had a rate three times higher than did white children. Parents made up 63% of the assailants, mothers’ boyfriends 18%, relatives and friends 12%, and babysitters 6%. Overall, two-thirds of the perpetrators were male.
Herman-Giddens et al extrapolated their North Carolina data to estimate the number of child abuse deaths in the entire United States. For the 10-year period, the number of deaths was estimated to be 9467 compared to 2973 listed in the mortality statistics for the country.
Comment by John M. Leventhal, MD, FAAP
Clearly, one of the first steps in understanding a problem is an accurate count of its occurrence. For many years, pediatricians have been aware that child abuse is underreported and undercounted. Not surprisingly, this study shows that deaths due to abuse are undercounted as well, but I was surprised at the extent of the underascertainment. I also found the results a bit puzzling. If the deaths were classified correctly, as due to abuse, by the ME’s office, why was this correct information not passed on to North Carolina’s vital records?
It would have been helpful to learn more about the details of the cases, particularly information about whether some of these deaths might have been preventable. For example, what proportion of cases had been seen by a physician for a suspicious injury prior to the child’s death?
What are the implications of these results for pediatricians who seldom care for patients who die? I would suggest two take-home messages. First, in all unexplained deaths, particularly in young children, child abuse should be considered in the differential diagnosis. Second, communities should have a systematic approach to reviewing the clinical and social characteristics of all unexplained deaths. Although child fatality review boards have been established in many states to help ensure that unexplained deaths are reviewed in a comprehensive manner, these review panels do not exist in all communities and, in many communities, only children who have been reported to protective services are eligible for review. All child deaths need a comprehensive review; pediatricians can work with their legislators, protective services agency, the ME’s office, and others to help accomplish such a systematic review. Such initiatives could lead to more comprehensive data concerning this kind of carnage and perhaps enable a proactive approach in prevention.
True statements concerning homicides resulting from child abuse include all of the following except:
a. The perpetrator is usually a relative.
b. They are currently accurately reported in state and national vital statistics.
c. The perpetrator is usually a male.
d. They have increased in incidence between 1984 and 1995.