An allegation of maltreatment, regardless of whether it is ever substantiated, is the strongest signal of a potential for a young child to experience a fatal injury, according to a recent study based on the records of more than 4.3 million children in California.
Children under 5 who are named in a maltreatment allegation are nearly six times more likely to die of infanticide and two times more likely to die from unintentional injuries as are other children, according to “Report of Maltreatment as a Risk Factor for Injury Death.”
Although some homicide prosecutors have tried to show “patterns of maltreatment” leading up to fatal injuries in their court cases, there had never previously been a mass body of data that showed that just the reporting of possible maltreatment is a stand-alone risk factor.
The study “suggests that children reported for maltreatment have a truly distinctive risk profile defined by much more than just birth into poverty,” study author Emily Put-nam-Hornstein said in the study’s “Implications” section. “At least in California, a report of maltreatment is more than just a marker of poverty – it is an independent signal of child risk.”
The results, coupled with other research on early predictors of maltreatment reports, suggest that child welfare systems could target preventive services at families before injury-related deaths happen.
“We would never be able to predict which kids who have contact with child protective services will actually die,” said Putnam-Hornstein. “But knowing we can predict contact and that contact is tied to death can have the positive benefit of reducing deaths.”
Millions of records
Putnam-Hornstein cross-referenced birth records of 4,317,321 California children born between 1999 and 2006 with state death records of nearly 26,000 children under the age of 5 who died during the same period. The cutoff age was 5 because 80 percent of child abuse victims who die in the United States do so by the age of 4.
She then matched those records against the California Child Welfare Performance Indicators Project’s database, which is housed at the University of California-Berkeley’s Center for Social Services Research and includes information on all contacts with the state’s child welfare agencies since 1994.
Of the 4.3 million children born in California between 1999 and 2006, most of those under 5 who died succumbed to disease or birth abnormalities, usually within 28 days of birth. Putnam-Hornstein found death certificates for nearly 2,000 who died of injuries.
Less than 1 percent of the children reported as maltreated to state child welfare authorities died during the years of the study. But re-ported children died of intentional injuries at a rate 5.9 times as high as that of other children, and were two times as likely to die from unintentional injuries.
Other health and demographic characteristics, most of them identifiable at birth, were also associated with injury deaths in this study. Male children were found to be more at risk than females to die of an injury, as were children who were in poor health, including those born with one or more birth abnormalities. Black children were overrep-resented among children who were fatally injured, while Hispanic and Asian/Pacific Islander children were underrepresented.
But when adjusting for those “risk factors at birth,” the study said, “a prior allegation to CPS [child protective services] proved to be the strongest independent risk factor for injury mortality before the age of five. Overall, 20 percent of children who died of injuries had been previously reported to CPS as maltreated. One-third of the 381 children who died of intentional injuries and 17 percent of the 1,438 children who died of unintentional injuries had been reported.
Putnam-Hornstein said she was not surprised that reporting a child for maltreatment was an indicator for a child whose death was counted as accidental or unintentional, suggesting that some form of neglect or abuse may have contributed to the child’s death – such as being left alone or not closely supervised and dying of a fall, or drowning or being hit by a car.
She draws two major conclusions from her findings. First, she suggests that the strong connection between reports and deaths is evidence that California’s process for reporting of maltreatment is effective in general, not a system that exposes poor families to the misinterpretations of others who have more.
“The question as to whether children are reported for maltreatment simply because they are poor has been passionately debated for decades,” she said in the study. “It is re-assuring that allegations … made to CPS are not entirely random events, with investigations disturbing the lives of families.”
She also said the “data suggest that public health prevention campaigns … might be fruitfully targeted to families reported to CPS.”
Findings from another study drawn from the California database suggest that many such families could be targeted early in a child’s life. Thirty-five percent of all children under 5 reported for maltreatment were reported before their first birthdays, and 25 percent were reported in their first three days of life, according to “Predictors of Child Protective Service Contact between Birth and Age Five,” which Putnam-Hornstein co-authored with Barbara Needell, who leads the Cal-ifornia Child Welfare Performance Indicators Project.
Stepping back further, that study identifies risk factors identifiable at birth that put children at a higher likelihood for CPS contact:
• Mothers with one or more preg-nancy terminations.
• Children born to younger mothers, especially teens.
• Mothers born outside the United States.
• Mothers who had not completed high school.
• A child falling third or fourth in birth order.
• Black and Native American parents, especially those covered by Medi-Cal, the state’s version of Medicaid.
• Late initiation of pre-natal care. Unexpectedly, lack of prenatal care was not a risk factor.
• Children whose paternity was not established.
Promising new tool
The death records for all children under 5 who died in California from 1999 to 2006 were examined to separate out all children who died of injuries. Putnam-Hornstein said, for example, that no death records listing Sudden Infant Death Syndrome as the cause of death were included in the study, though many experts contend that child homicides can sometimes be miscoded as SIDS.
“It is difficult, if not impossible, to distinguish at autopsy between SIDS and accidental or deliberate suffocation with a soft object,” wrote Kent Hymel in “Dis-tinguishing Sudden Infant Death Syndrome from Child Abuse Fatalities,” which appeared in a 2006 issue of Pediatrics.
The California Child Welfare Performance Indicators Project database is primarily used by California counties to gauge their own performance against state averages and other counties, but it has also been used as a source for other major research projects.
Richard Barth, dean of the School of Social Work at the University of Maryland, and Mark Courtney, now a professor at the University of Chicago School of Social Administration, helped establish the database and used it to show that black children were represented disproportionately, at least at the front end of the system.
In another report drawn from the database, Barth and Needell Courtney used the records to determine that older, poorer women who continued to have children were more likely to have those children end up in foster care.
Barth called Putnam-Hornstein’s work “important” and a prime example of how state vital statistics records can be used productively and without breaching important privacy concerns.
“Report of Maltreatment as a Risk Factor for Injury Death: A Prospective Birth Cohort Study,” Free, 12 pages, http://cmx.sagepub.com/con-tent/16/3/163.full.pdf+html.
“Predictors of child protective service contact between birth and age five: An examination of California’s birth cohort,” Child and Youth Services Review, 2011.