Medicare Shared Savings Program and Disparities for Patients With Low Income

Importance: Little is known about whether accountable care organizations (ACOs) in the Medicare Shared Savings Program (MSSP) have affected socioeconomic health care disparities. Objective: To investigate whether attribution to an ACO was associated with greater improvement in outcomes for patients with low vs higher income. Design, Setting, and Participants: This cohort study employed triple difference-in-differences analyses using 2010 to 2019 Medicare data to investigate whether quality differences between patients with low income and patients with higher income attributed to ACOs changed after ACO formation relative to patients at non-ACO practices or clinics. In exploratory analysis, we calculated 3-year outcomes (reflecting 1 ACO contract period) for each ACO and examined whether changes in disparities were associated with ACO characteristics such as size, rurality, inclusion of a hospital, share of patients with low income, presence of federally qualified health centers, and level of baseline disparities. Data analyses were performed between January 1, 2025, and February 27, 2026. Exposure: Attribution to a medical practice or clinic in an ACO during the first 2 years of ACO tenure. Main Outcomes and Measures: Differential changes in utilization-based quality metrics for patients with low vs higher income. Outcomes included potentially avoidable emergency department visits, ambulatory care-sensitive admissions, and 30-day unplanned hospital readmissions. Results: There were 585 ACOs in the study. Three years after ACO formation, there was no differential change in outcomes for patients with low vs higher income at ACOs compared with non-ACOs. However, there was considerable variation; 76 ACOs (13.0%) were associated with reduced outcomes disparities in at least 1 quality measure while maintaining or improving quality for patients with higher income; 68 ACOs (11.6%) were associated with worsened outcomes disparities on at least 1 quality measure while maintaining or worsening quality for patients with higher income. ACOs that were associated with reduced disparities had a 0.36 (95% CI, 0.17-0.56; P < .001) higher SD-level z score of baseline disparities (eg, a 16.2% higher disparity in preventable emergency department visits) compared with ACOs that were associated with increased disparities but were similar on other characteristics. Conclusions and Relevance: In this cohort study of MSSP ACOs, ACO formation was not associated with reductions in income-based disparities in quality of care during the program's first decade. However, the differences in quality of care narrowed for patients with low vs higher income at a subset of ACOs.