Predicting outcomes of children in residential treatment: A comparison of a decision support algorithm and a multidisciplinary team decision model

Residential treatment is the most intensive and costly component of all child welfare systems per episode of care. At the same time, decisions to place in residential treatment centers are prioritized by the practice of least restrictive setting and best interest for children. There are, however, no standard evidence-based criteria for placing children in residential treatment. Clinical judgment, staffing dynamics, and other system factors are part of the decision-making process. Thus, some residential placements may be unnecessary and may be even harmful. The present study compares two models of decision-making, a multidisciplinary team approach and an objective decision support algorithm, and assesses outcomes when the two models either concur or not. Concordant decisions predicted greater clinical improvement than discordant decisions, but no differences were found in length of stay in placement. Policy implications for the decision-making process in child welfare are discussed.

► Two decision-making models for residential placement in child welfare were compared.
► Multidisciplinary team vs. decision support algorithm based on clinical ratings.
► Concordant decisions predicted significant clinical improvement over time.
► Discordant decisions predicted fewer clinical gains.
► Evidence-based placement decision-making can benefit child welfare outcome.