Multilevel Model of Factors Influencing Foster Care Decision-Making

The current study aims to use the decision-making ecology (DME) to understand what decision-making factors influence state child welfare and foster care entry rates. We conducted a multilevel logistic regression model with 4,473 children from a rural northeastern state to understand the DME. Significant child-level predictors of foster care entry were having an unsafe determination on the Structured Decision-Making (SDM)(R) safety assessment, high SDM risk score, neglect, physical abuse, and child age below 5 years. In addition to safety considerations, community poverty and caseworker orientation toward child removal were all significant predictors of foster care. The study uncovered contextual system-level factors contributing to foster care placement that are policy malleable and, if addressed, could improve family preservation and prevent out-of-home placement.