197 - Leveraging Medicaid data to evaluate dyadic care models in early childhood: creating mother-child linkages

Poster Number: 197
Publication Number: 197.319

Katherine Piwnica-Worms, NYU Grossman School of Medicine, New York, NY, United States; Mary M. McCord, NYC H+H/Gotham Health, New York, NY, United States; Renata Howland, New York University, New York, NY, United States; Arthur H. Fierman, New York University Grossman School of Medicine, New York, NY, United States; Marcy Stein-Albert, Icahn School of Medicine at Mount Sinai, Jamaica, NY, United States; John Billings, New York University Grossman School of Medicine, New York, NY, United States

Background: There is growing interest in how states and health systems can collaborate and leverage Medicaid data to inform sustainable value-based payment (VBP) models focusing on early child development, preventive and integrated mental health, and dyadic (caregiver-child) health. A key challenge is assessing claims-based outcomes and value across the dyad with a focus on prevention. 3-2-1 IMPACT, New York City Health + Hospitals’ (NYC H+H) version of the New York State (NYS) Department of Health’s dyadic care model, is leveraging Medicaid data on the path to advocating for a statewide early childhood VBP model.

Objective: We assess the feasibility of linking the mother-child dyad in Medicaid data, a first step in our plan to evaluate utilization, cost, and quality outcomes following participation in IMPACT.

Design/Methods: We used NYS Medicaid fee-for-service and encounter claims to identify children (1-35 months) with a CPT-coded well visit in 2018 at 3 IMPACT and 3 control sites. Using enrollment information and historical claims, children were linked with potential mothers if the child and mother had matching NYS Medicaid Case numbers and the child’s date of birth fell within 2 days of the mother’s hospital admission delivery date. Dyadic linkages were cross-checked with NYC H+H’s electronic medical record, where linkages are autogenerated through the child’s 5th birthday when a mother delivers at a NYC H+H facility.

Results: Among 9,016 children ages 1-35 months who were enrolled in Medicaid and received care at an IMPACT or control site in 2018, 7,186 (79.7%) of them were linked successfully to their mothers. To evaluate the effect of IMPACT, we plan to analyze this cohort using a difference-in-difference approach. We will evaluate the short-term effects on ED utilization, site loyalty, cost of care, and Medicaid quality measures focused on pregnancy, mental health and early childhood preventive care by Quarter 1 2022, with plans to evaluate longer-term outcomes through Quarter 3 2023. Conclusion(s): Preliminary Medicaid analysis identified almost 80% of mother-child dyads in the baseline IMPACT program cohort. Analysis of linked mother-child dyads is essential to evaluate early childhood interventions in primary care but has not been widely adopted. Our findings suggest that evaluating dyadic care models utilizing Medicaid data is possible in NYS and IMPACT’s planned analysis could inform the growing interest in building data and evaluation structures for early childhood VBP models within Medicaid.