Objective
This study used group variations in child injury fatality rates to assess racial bias in the population of children identified as victims of maltreatment.
Methods
Injury fatality and maltreatment data from California were compiled for the years 1998–2007. Death and maltreatment risk ratios (RRs) and 95% confidence intervals (CIs) were computed by race and age. Rates of excess child injury mortality by race were derived from three different baseline rates of death. Substantiations per excess injury death were calculated.
Results
Compared with white children, black children faced a risk of substantiated maltreatment that was more than twice as great (black RR: 2.39, 95% CI 2.37, 2.42) and were fatally injured at nearly twice the rate (black RR: 1.89, 95% CI 1.68, 2.12). Per excess death, however, black children had rates of substantiated maltreatment allegations that were equivalent to or lower than the rates for white children.
Conclusions
These data support claims that, at least in California, black-white racial disparities observed in maltreatment rates reflect real group differences in risk. These data provide no evidence of systematic racial bias in the child protective services'' substantiation process.In the United States, data collected by child protective services (CPS) consistently indicate that black children are abused and neglected at roughly twice the rate of white children.
1–3 Yet, these data reflect only maltreatment victims known to CPS, with other sources suggesting the count of abused and neglected children may be far higher.
4,5 Unknown is whether the group of maltreated children identified by CPS is racially representative of the broader population of maltreated children.In an effort to overcome the limitations of CPS data, the National Incidence Study of Child Abuse and Neglect (NIS) has been conducted on four separate occasions between 1980 and 2006.
4,6–8 The NIS attempts to estimate the actual number and characteristics of maltreated children based on cases reported to CPS, as well as knowledge of abused and neglected children ascertained from other professionals. Despite large black-white disparities observed in CPS-reported rates of maltreatment, the first three waves of the NIS failed to identify any statistically significant racial differences in the community incidence of abuse and neglect.
6–8 As such, it was widely concluded that there must be a racially biased identification of black children by CPS.
9–11 For the first time, however, the most recent wave of the NIS found that black children were maltreated at significantly higher rates than white children, reporting “strong and pervasive race differences in the incidence of maltreatment.”
4The purpose of this study was to examine racial disparities in child maltreatment using a new, population-based and public health-oriented approach. Given the incomplete nature of CPS data,
5,12,13 the contradictory claims and sampling controversies that surround the NIS,
14–16 and research arising from other sources suggesting possible reporting and CPS substantiation bias of black children,
17–19 it seemed an opportune time to consider other data that might inform an understanding of racial disparities. In this ecological study, data from two public surveillance systems were used to assess possible bias in one source (data collected by CPS) by using the more complete and objective information available in the other (vital statistics death data). Rates of excess injury mortality were based on different assumptions as to the fraction of injury deaths that were preventable. The ratios of substantiations per excess death were compared across races. It was hypothesized that racial variability in rates of excess injury death would be observed—arising from disproportionate burdens of poverty, substance use, and other risk factors—but that absent widespread bias on the part of CPS, there would be a fairly consistent count of substantiations for each excess death.
相似文献