Measuring Income Disparities After ACO Formation
An independent published methods example using Medicare claims, practice attribution, and triple differences to evaluate changes in income disparities.
Home / Case Studies / Measuring Income Disparities After ACO Formation
Source published August 20, 2026 · Resource reviewed September 7, 2026
The research question
Did ACO formation change the gap in care outcomes between lower- and higher-income Medicare patients?
Data files and cohort construction
The study used a 20% sample of 2010–2019 fee-for-service claims, MD-PPAS practice files, MSSP provider files, and enrollment information. Attribution used the preceding year’s plurality of primary care services at a practice TIN or clinic CCN. ACO assignment retained organizations from their first two participation years. Ever receiving Part D Low-Income Subsidy eligibility defined low income. Patient-years with any MA enrollment or incomplete Parts A and B coverage were excluded.
Measures and analytic design
Outcomes were potentially avoidable emergency visits, ambulatory care–sensitive admissions, and 30-day unplanned readmissions. Triple differences compared income gaps before and after ACO formation against changes at non-ACO organizations in the same hospital-referral regions.
Robustness checks and limitations
Checks included cohort-specific estimates, event studies, Sun–Abraham estimates, and patient-composition analyses. ACO-specific estimates used empirical Bayes shrinkage. Adjusted results did not show an overall reduction in disparities. Subsidy eligibility is an income proxy, and causal interpretation requires comparable counterfactual trends.
Practical application: show all four groups
The following are FastHSR implementation considerations, not additional procedures claimed for the published study. A useful reporting table has four groups: lower-income ACO patients, higher-income ACO patients, and the corresponding two non-ACO groups. Show each group’s denominator and outcome rate in each period. Then calculate the change in the income gap for ACOs and the change in the income gap for comparison organizations. The difference between those changes is the quantity of interest.
Keep attribution and measure definitions reviewable
Create a separate attribution audit showing the service window, clinician-to-practice linkage, tie rules, and organizations retained in each year. Maintain a versioned definition for each outcome, including diagnosis exclusions and readmission eligibility. Describe whether the denominator is people, person-years, or qualifying discharges. Do not combine rates with different denominators into a single percentage without an explicit transformation.
How this can guide decisions
This design can help an evaluator ask whether an improvement reached patients with fewer financial resources. Report group-specific changes alongside the disparity estimate: a narrowing gap can occur while both groups worsen. Review changing patient mix and baseline trends before attributing the change to a program. Organization-level estimates also need uncertainty intervals; small samples should not drive a performance ranking.
Source article and supplement
Medicare Shared Savings Program and Disparities for Patients With Low Income. JAMA Network Open. 2026;9(8):e2629369. doi:10.1001/jamanetworkopen.2026.29369.
The full article was reviewed. The separate supplement download could not be retrieved during this scan; descriptions of its contents rely on the article. Use the journal’s supplement access point to verify exact specifications before replication.
Frequently asked question
What does a triple difference add?
It compares how a gap between two patient groups changes in participating organizations with how that same gap changes in comparison organizations.