The Differential Impact of Medicaid Expansion on Disparities in Outcomes Following Peripheral Vascular Intervention

Abstract

Peripheral artery disease (PAD) disproportionately affects nonwhite, Hispanic/Latino, and low socioeconomic status patients, who are less likely to have insurance and routine healthcare visits. Medicaid expansion (ME) has improved insurance rates and access to care, potentially benefitting these patients. We sought to assess the impact of ME on disparities in outcomes after peripheral vascular intervention (PVI) for PAD. A retrospective analysis of prospectively-collected Vascular Quality Initiative PVI procedures between 2011 and 2019 was conducted. The sample was restricted to first-record procedures in adults under the age 65 in states that expanded Medicaid on January 1, 2014 (ME group) or had not expanded before January 1, 2019 (non-expansion [NE] group). ME and NE groups were compared between pre-expansion (2011-2013) and post-expansion (2014- 2019) time periods to assess baseline demographic and operative differences. We used difference-in-differences multivariable logistic regression adjusted for patient factors and clinical center and year fixed effects. Our primary outcome was 1-year major amputation. Secondary outcomes included trends in presentation, 30-day mortality, 1-year mortality, and 1-year primary and secondary patency. Outcomes were stratified by race and ethnicity. We examined 34,313 PVI procedures, including 20,378 with follow-up data. Rates of Medicaid insurance increased post-expansion in ME and NE states (ME 16.7% to 23.0%, P<0.001; NE 10.0% to 11.9%, P=0.013) while rates of self-pay decreased in ME states only (ME 4.6% to 1.8%, P<0.001; NE 8.1% to 8.4%, P=0.620). Adjusted difference-in-differences analysis revealed lower odds of urgent/emergent PVI among all patients and all nonwhite patients in ME states post-expansion compared to NE states (all: odds ratio [OR] 0.53 [95% confidence interval 0.33-0.87], P=0.011; nonwhite: OR 0.41 [0.19-0.88], P=0.023). No differences were observed for 1-year major amputation (OR 0.70 [0.43-1.14], P=0.152), primary patency (OR 0.93 [0.63-1.38], P=0.726), or secondary patency (OR 1.29 [0.69-2.41], P=0.431). Odds of 1-year mortality were higher in ME states post-expansion compared to NE states (OR 2.50 [1.07-5.87], P=0.035), although 30-day mortality was not different (OR 2.04 [0.60-6.90], P=0.253). Notably, odds of 1-year major amputation among Hispanic/Latino patients decreased in ME states post-expansion compared to NE states (OR 0.11 [0.01-0.86], P=0.036). ME was associated with lower odds of 1-year major amputation among Hispanic/Latino patients who underwent PVI for PAD. ME was also associated with lower odds of urgent/emergent procedures among patients overall and nonwhite patients specifically. However, 1-year mortality increased in the overall cohort. Further study is needed to corroborate our findings that ME may have benefits for certain underserved populations with PAD.