Using Predicted Risk to Assess Disparities in the Utilization and Delivery of Community-Based Services

Background & Purpose. The Family First Prevention Services Act (FFPSA) aims to support families to prevent foster care entry. FFPSA provides transition funding to states, including California, that operated under Title IV-E waivers. In place since 2007, California’s waiver allowed the Los Angeles County Department of Children and Family Services (DCFS) considerable flexibility in program design and service delivery. Given changes expected in California’s approach to FFPSA, the current paper examines current community-based service programs, including the potential role a predictive risk model could play in identifying families to prioritize for services. Specifically, this analysis focuses on historical patterns of referral, service completion, and outcomes for families engaged in two key programs: (1) Prevention and Aftercare (P&A), which is designed to provide primary prevention supports for families in the community, as well as secondary and tertiary prevention services for families involved with the child welfare system; (2) Family Preservation (FP), which provides intensive services to improve family functioning and protect children while they remain in or safely transition back home.

Methods. De-identified study data were derived from: (1) Child Welfare Services/Child Management System (CWS/CMS) records available through an active data sharing agreement with the California Department of Social Services; and (2) records from the Family Centered Services (FCS) database abstracted by DCFS. Data extracts included families who were referred for alleged abuse or neglect between 2016 and 2019. Each family was classified based on a risk score generated through a predictive risk model. Families were then longitudinally followed through FCS data to assess patterns of service referrals and engagement.

Results. Overall, approximately one-third of eligible families (families with an open in-home services case) were referred to FP services between 2016-2019. Families who successfully completed services were significantly less likely to experience out-of-home placement. For P&A, about 14% of all eligible families (all DCFS families excluding those with out-of-home placement) were referred for services. Families referred to P&A who successfully completed services were significantly less likely to have subsequent DCFS involvement compared with families who were not referred or did not complete services. Additional analyses indicated a graded relationship between risk level (estimated by the county's risk stratification model) and the share of families referred for FP and P&A services. Yet even at the highest levels of risk, fewer than 1 in 5 families was connected to services.

Conclusion & Implications. Transition to FFPSA presents an opportunity for local jurisdictions to rethink the array of services and partnerships needed to support and serve families. Yet, with limited service slots available, efforts to identify families who will most benefit from prevention services will also be an important dimension. Findings from the current study suggest that using risk stratification approaches may be useful to integrating the work of contracted services with other resources.