Identifying Infants at Risk of Adverse Outcomes Using Administrative Data: Findings from Washington State

This report summarizes a set of predictive models which used administrative data available at birth to predict adverse outcomes for children on Medicaid. The three outcomes of interest were: 1) infant mortality, 2) Child Protective Services involvement before age four, and 3) developmental disorder diagnosis before age four. The findings demonstrate the utility of combining birth certificate information and other state administrative data about maternal and child factors such as low birth weight, maternal child welfare involvement, and number of previous births to identify infants that are most at risk of adverse outcomes.