Maximizing child abuse prevention resources with place-based predictive analytics
The ethics of predictive analytics in child welfare are hotly debated by those who envision families in crisis facing bias, disproportionality, or even unnecessary removal of children as the result. While there is historical truth in these concerns, there is also incredible potential for ethically responsible predictive tools.
A recent example of ethically responsible predictive analytics, designed to improve child welfare call screening decisions, is the Allegheny Family Screening Tool, detailed in the recent New York Times article, “Can an Algorithm Tell When Kids Are in Danger?” This predictive algorithm is unique and exceptional because it is a transparent tool that is community owned and open, thanks to its visionary creators, Emily Putnam-Hornstein and Rhema Vaithianathan.
All over the United States, there are “new and emerging techniques to harness data and technology to make dramatic improvements to child welfare practice and ultimately produce better outcomes for children and families,” says Christopher Teixeira and Matthew Boyas of the MITRE corporation, in a recent review article, Predictive Analytics in Child Welfare, An Assessment of Current Efforts, Challenges and Opportunities.
Many of these tools and techniques are geared toward an improved response when an allegation of child maltreatment is reported to child welfare, and could offer a tremendous benefit to the most vulnerable children. In fact, the development of predictive analytics for child welfare was one of the main recommendations in the final report from the Federal Commission to Eliminate Child Abuse and Neglect Fatalities.
Increasingly, the goal of children’s welfare administrations is to proactively provide support to vulnerable families before an allegation suspicion of child maltreatment is reported. Place-based predictive analytics may contribute to the prevention of child maltreatment by identifying places where child maltreatment is more likely to occur in the future based on contextual, not personal, risk factors.
Knowing precisely where to target prevention resources, and what to focus on, can maximize impact of prevention initiatives as demonstrated by the work of Joel Kaplan and Less Kennedy with risk terrain modeling.
Instead of looking for individuals with certain risk factors, place-based predictive analytics identifies small geographic areas where risk factors in the physical and social environment support behaviors related to child maltreatment. With risk factors based on the Adverse Childhood Experiences study (ACEs) this method has succeeded in predicting where child maltreatment is likely to occur in the future.
Knowing where to focus resources to make the biggest impact for prevention is a huge advantage. Place-based predictive analytics offer a level of focus not previously available for strategic allocation of successful community-based prevention programs such as the Casey Family Program’s 2020: Building Communities of Hope initiative, the Aces Interface initiative to Build Self Healing Communities, and Prevent Child Abuse America’s Healthy Families America.
State agencies, city governments, and children’s advocacy centers have been attracted to geospatial risk analysis because it draws on community factors rather than person-based concerns.
Location-based predictive analytics provide an opportunity to maximize resources in communities where they can have the most positive impact on the children and families served by children’s administrations. By focusing on reducing exposure to adverse experiences in addition to child maltreatment, this method looks at multiple needs of children and families in the communities they live in. Addressing a situation before it escalates to a call to child protective services is the goal.
Dyann Daley, MD is the Founder and CEO of Predict-Align-Prevent, a Texas-based 501(c)(3) nonprofit organization focused on the development of child abuse and neglect prevention solutions. Fundamental to her work are a commitment to open science, objective metrics, and child-centered outcomes. Dr. Daley is a practicing pediatric anesthesiologist, author, and speaker. Learn more and contact here.