Improving Family Stability and Substance Use Recovery for Families in the Child Welfare System: Impact of Ohio's Statewide System Improvement Program

Abstract

Research ObjectiveFamily drug courts (FDCs) and other multi‐system strategies are increasingly used to address the needs of families involved in the child welfare system who are affected by substance use disorders. Ohio’s Statewide System Improvement Program (SSIP) was established to expand the scale and service scope of existing FDCs and coordinating care across court, child welfare, and behavioral health systems. This study was conducted to evaluate the impact of SSIP on family stability and recover outcomes over time.Study DesignPropensity Score Matching was applied to compare outcomes for eligible families with and without exposure to SSIP and identify characteristics of families most likely to enroll in FDC. Logistic and linear regression analyses were conducted to identify differences in temporal trends among the SSIP intervention and matched comparison groups.Population StudiedAdministrative data from state agencies and local FDCs were linked using a deterministic linkage process. Statewide Automated Child Welfare Information System (SACWIS) data were gathered to identify cases of child maltreatment[i] where substance use was a contributing factor[ii] and to assess family stability (eg, out‐of‐home placement, subsequent maltreatment, and reunification). Medicaid administrative data provided information about clinical diagnoses and treatment. Behavioral health treatment episode data provided information related to social determinants of health (eg, employment).Principal FindingsThe SSIP FDC participants demonstrated significant improvement (reduction) over time in out‐of‐home placement (2015‐2017), while there was no improvement over time in the comparison group (OR = 1.70, P < .05). Counties that implemented SSIP demonstrated improvement over time (2014‐2017) relative to the comparison group in the following areas: behavioral health treatment retention (β = 0.26, P < .01); medication‐assisted treatment retention (β = 0.45, P < .05); length‐of‐stay in out‐of‐home placement (OR = 1.15, P < .05); and placement with kin (OR = 1.22, P < .01).Among 1,695 family case episodes eligible for FDC, only 15.6% were enrolled in a FDC. Standardized differences scores revealed family characteristics associated with FDC enrollment including: white race (P < .001), history of child maltreatment (P < .001), and parent with mental illness (P