Effects of Medicaid Coverage and State-Level Delivery Approaches on Healthcare Quality, Outcomes, and Costs for Adults with Diabetes

Despite tremendous advances in diabetes care over the past 25 years, many Americans continue to have poor control over modifiable risk factors such as blood glucose, blood pressure, and cholesterol. As a result, diabetes continues to be a leading cause of death and many preventable complications, such as heart disease, stroke, cognitive impairment, chronic kidney disease, visual impairment, and amputation. Unfortunately, many adults with diabetes still lack access to a coordinated healthcare team and do not complete evidence-based preventive services, receive newer medications or technologies that may improve disease care, participate in daily self-management behaviors, or follow-up longitudinally for early detection and treatment of comorbidities and complications. This gap between evidence and practice disproportionately impacts racial and ethnic minorities, those with disabilities, the aged, and the poor. Policies that expand eligibility requirements for health insurance, encourage continuous coverage, or lower the acquisition costs of high value medications have the potential to improve healthcare engagement, the delivery of evidence-based services, and receipt of newer medication therapies with the promise of improving health outcomes. As the nation’s single largest health insurer, Medicaid programs provide coverage for 72 million people, including vulnerable groups such as the disabled, aged, and poor. Under federal guidance, states have a great deal of flexibility in designing and administering their programs, and the Affordable Care Act and other health reforms have further catalyzed considerable state-level variation in Medicaid eligibility, coverage, population management services, and financing. This variation presents a tremendous opportunity for natural experiments designed to evaluate the comparative effectiveness and costs of policy shifts on the care and outcomes of diabetes among vulnerable patient groups for whom the burden of diabetes remains highest. We propose to leverage our existing access to longitudinal Medicaid claims data for more than 6 million people in up to 27 states, along with linked electronic health records from large health systems in 3 of those states to conduct rigorous natural experiments that will examine effects on diabetes care processes, health outcomes, and costs following 1) Medicaid eligibility expansions implemented differentially by 3 states; 2) state Medicaid policies resulting in 12-month continuous coverage eligibility for adults; and 3) state policies to expand coverage for SGLT2 inhibitors and/or GLP 1RA under Medicaid “best price” rules. Natural experiments focused on adoption of the income-eligibility waiver and adoption of coverage for SGLT2i and GLP1ra class medications will compare implementing to non-implementing states using difference-in-differences (DiD) designs. Our evaluation of insurance expansion will exploit both age- and income-based Medicaid eligibility thresholds to conduct a novel regression discontinuity/DiD hybrid design with matching and multiple imputation methods to identify comparison groups both within and between the implementing states. Through direct engagement of collaborators and data contributors in all participating states, our studies have been designed around high priority questions of stakeholders who are implementing the programs, and our results will be communicated through these same channels to address any obstacles during conduct of the study and to guide policy decisions following the study.