Using analytics to reduce child abuse risk
Connecting state and local government leaders
Florida's Department of Children and Families is studying the use of predictive analytics to reduce child fatalities.
In an effort to reduce the number of child fatalities caused by abuse and neglect, Florida’s Department of Children and Families (DCF) is turning to predictive analytics.
DCF recently released a report analyzing child fatalities reported to the Florida Abuse Hotline from Jan. 1, 2007 to June 30, 2013. The data was used to confirm general trends in child fatalities and to determine key risk factors, with the goal of improving child welfare practices by using predictive analytics tools and techniques.
Interim DCF Secretary Mike Carroll said, "This recent study has helped us determine which children are at a greater risk and offers an additional tool for DCF and partners to better assess and provide for the safety of children."
The five-week project was completed using analytical software from SAS of Cary, N.C., and the assistance of Atlanta-based consulting firm North Highland. Additional support was provided by the private, non-profit Child Welfare Policy and Practice Group (CWG). Input from from several child protection experts in the medical, legal, community-based care, law enforcement, prevention and quality assurance professions, was also provided
The project primarily analyzed data from the Florida Safe Families Network and the Child Death Review Database. Three secondary summary-level sources also were referenced from the Department of Health, Abuse Hotline and Child Protective Investigators. Children and families who had prior involvement in protective investigations – as well as those who did not – were included in the study. The number of child fatalities was compared to the entire population of children involved in child protective investigations during the same time period.
The SAS analysis considered factors such as prior removals due to sexual abuse, prior removal due to drug abuse and physical or mental disabilities. Additional variables included whether elements such as gender, ethnicity, number of siblings, etc. increased or decreased the odds of death.
“Our research showed the tremendous positive effect of a visit from a caseworker,” said SAS analytics expert Albert Blackmon, who presented the findings to The Federal Commission to Eliminate Child Abuse and Neglect Fatalities. “But child protective services agencies across the country are overburdened. Analytics can help caseworkers identify the most at-risk kids as well as pinpoint the services that can lead to the most positive outcomes.”
The report concluded that DCF needs to develop a comprehensive child welfare heightened risk analytics model to improve child welfare and family safety practices.
“By incorporating a wider set of data, a richer and more meaningful analysis can be done. Some of the sources used in the Child Fatality Trend Analysis would be expanded as well as inclusion of new sources that could indicate risk factors like financial hardship, substance abuse or mental health stresses on the family, or household members with a history of crime or violence,” noted the report. The report also suggested improvements in data entry, historical tracking and user training.
In response to the high child abuse death rates, late last month Florida Governor Rick Scott signed a bill (Senate Bill 1666) overhauling DCF.
The bill calls for adding 270 child protective investigators, establishing a new Assistant Secretary for Child Welfare at DCF, creating a “Critical Incident Rapid Response Team” to quickly investigate child abuse deaths where the child was known to the child welfare system and strengthening provisions relating to safety plans for children who have been abused or neglected.