Replication Data for: The Association Between Medicaid Expansion and Cardiovascular Interventions: The Michigan Experience

We evaluated Michigan's Medicaid expansion as it relates to access and outcomes for 7,558 coronary artery bypass grafting (CABG) operations at 33 hospitals, and 45,183 percutaneous coronary interventions (PCI) at 47 hospitals. We excluded patients >=65 years and those with Medicare, non-Medicaid state or Federal government coverage (e.g., military), or non-US insurance. Clinical data, from each institution’s participation in CABG and PCI collaboratives, represents procedures performed between April 2012 – March 2014 and April 2014 – March 2016 (24 months before/after expansion).

We evaluated changes in access (i.e., number of procedures) by insurance type (private, Medicaid, uninsured), and used risk adjustment models for CABG(3) and PCI(4) to calculate adjusted mortality (CABG: operative mortality; PCI: during the index admission) before and after expansion. G-computation was utilized to estimate the relationship between Medicaid expansion and the number of patients presenting quarterly for revascularization within insurance coverage type (Medicaid, private, uninsured). A generalized estimating equation (GEE) model using autoregressive covariance structure was fitted, including insurance type, calendar quarter, Medicaid expansion, and interaction terms (two-way and three-way interactions among quarter, Medicaid expansion, and insurance coverage type) as predictors.