Background: The Eckerd Rapid Safety Feedback Process (ERSF) is an approach to child welfare practice that combines predictive risk modeling (PRM) with real-time quality assurance and staff coaching during a child protective services investigation. During the case investigation, quality assurance (QA) reviewers provide guidance to staff by assessing key dimensions of their investigative practices and providing coaching sessions as needed.
Objective: To determine to what degree ERSF reduced the incidence of children who experienced high severity physical, sexual, or neglect maltreatment (HS-CAN) among children known to a state child welfare agency, within 12 months of the agency's involvement in a previous investigation.
Methods: This multi-year quasi-experimental study evaluated the impact of the ERSF Process on risk of repeat high severity maltreatment among children and youth identified by a PRM algorithm as high risk for experiencing repeat maltreatment and receiving the additional QA intervention, compared to those receiving the usual standard of care.
Results: The ERSF Process did not lower the rate of HS-CAN among children identified as high risk. The joint odds ratio for receiving ERSF given being high risk was 1.05, i.e. a point estimate indicating higher probability of future HS-CAN.
Conclusions: In this statewide implementation of ERSF, we did not find a positive effect of this particular PRM-based intervention. Future maltreatment, even within 12 months of a prior investigation, may be too distal an outcome for a PRM and QA process designed to produce a high-quality CPS investigation and safety plan.
Keywords: Child abuse; Child welfare; Predictive risk modeling; Program evaluation; Real-time quality assurance.
Copyright © 2022. Published by Elsevier Ltd.