CCOULD Lessons Learned

The goal of the CCOULD project was to provide technical assistance to two states in linking the Medicaid administrative claims of children and their caregivers with case-level data from the child welfare system, and to combine the state-specific datasets into a multi-state, deidentified dataset for secondary data analysis by researchers. Linked data on services from child welfare and Medicaid can help us understand the effectiveness of interventions and improve coordination to address critical challenges faced by families at risk of child welfare system involvement, including parental substance use. This report describes the experiences of the two states in linking Medicaid and child welfare data and summarizes the lessons learned across both states.

• Florida created a linked dataset covering 2016 through 2021 that included 80% of children (N = 669,425) who had been in contact with Florida’s child-welfare system. Florida linked 57,512 caregivers of those children to Medicaid records. Of the 669,425 Florida children who were linked to Medicaid, 8.5% also had at least one caregiver linked to Medicaid.
• Kentucky created a linked data set that covered the period 2016 through 2021 that included 81% of children (N = 216,738) who had been in contact with Kentucky’s child-welfare system. Kentucky linked 46,809 caregivers of those children to Medicaid records. Of the 216,738 Kentucky children linked to Medicaid, 21.3% had a least one caregiver linked to Medicaid.
• The experience in linking data from two states through the CCOULD project provides a solid foundation for future linking efforts. As the value of integrating data across Medicaid and child welfare program becomes more apparent, more and more states may begin to build from this foundation.

Data linkage was facilitated by the creation of a common data model (CDM) that employed the same variables and file structure between the two states. Child and caregiver child welfare records were linked to Medicaid records using deterministic linkage.  The data was harmonized and de-identified before being submitted to the National Data Archive on Child Abuse and Neglect (NDACAN). Researchers interested in analyzing these data may apply for the data here .  

• Linkage of children to parents could be enhanced if Medicaid claims data had a family identifier. Because of a lack of family identifier in Medicaid claims, children and caregivers could only be linked if the caregiver’s identifying information was included in the child welfare data.
• Future linkage efforts could focus more attention on highlighting the specific operational value to states and their constituents of linking their data. A thorough understanding of states’ data and operational readiness to undertake data linkage is also an important consideration.

Mark, T.L., Dolan, M., Allaire, B., Bradley, C. (2022). Linking Child Welfare and Medicaid Data: Lessons Learned from Two States. OPRE Report 2023-038, Washington, DC: Office of Planning, Research, and Evaluation, Administration for Children and Families, Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services.