Reunification from foster care: Informing measures over time

Children's paths through the child welfare system to permanency (adoption, reunification or guardianship) are complicated. Currently, federal policy and the child welfare literature point to reunification with the birth parents as the preferred type of permanent exit. The time it takes to move children towards a permanent exit is an important metric for child welfare. Traditionally, survival analysis was the method utilized when examining exits over time. One of the primary assumptions necessary for survival analysis is that the censoring be uninformed (i.e. that there is no underlying mechanism driving the censoring events). This assumption is violated when the probability of one event occurring is not independent of another outcome. A multi-step statistical model was developed to overcome the informative censoring present in the analysis of child welfare exits to permanency. Results suggest that, for reunification, the standard survival analysis model overestimates the amount of time children spend in care prior to reunification by not accounting for informative censoring. Overestimating the time children spend in care could lead to incorrect caseload projection, inaccurate fiscal projections, and imprecise measurements of time to reunification.