Health system reforms, such as value-based payment, can worsen or improve existing health care disparities, even if policy changes do not target the disparities themselves. A new study by Rachel Werner, Genevieve Kanter, and Daniel Polsky adds to our understanding of how Medicare’s accountable care organizations (ACOs) interact with existing disparities, and its findings are reassuring.

Medicare ACOs are groups of providers (hospitals and physician practices) that agree to be financially accountable for the cost and quality of care for assigned beneficiaries. Hitting financial and quality benchmarks allows ACOs to share in the savings, and less frequently, missing those benchmarks can result in penalties. In 2018, 656 ACOs delivered care to nearly one-third of all Medicare fee-for-service patients. Existing evidence indicates that Medicare ACOs have produced small savings while maintaining or improving quality.

But are the benefits of Medicare ACOs reaching socially and clinically vulnerable groups?  A previous study found that physicians were less likely to join commercial and Medicare ACOs if they practiced in areas where a higher percentage of the population was black, living in poverty, uninsured, disabled, or had less than a high school education. Health disparities could worsen if the benefits of ACOs do not extend to vulnerable groups.

A more detailed picture emerges from the new study in JAMA Network Open, which compares the patient panels of physician groups that joined and did not join ACOs as part of the Medicare Shared Savings Program (MSSP) from 2012 through 2014. Werner and colleagues estimate how different levels of socioeconomic disadvantage and poor health within a practice’s patient panel predicted ACO participation. Their results suggest that physicians who join ACOs are no less likely to care for minority and low-income patients, and more likely to care for sicker patients.

At the practice level, physician groups that joined an ACO had 5.1% more dual eligible patients (i.e., poor patients eligible for both Medicare and Medicaid) and 4% more medically high-risk patients. At the patient level, medically complex patients were more likely to be attributed to an ACO. There was no relationship between race or poverty and the likelihood that a patient was attributed to an ACO.

Their results suggest that physicians who join ACOs are no less likely to care for minority and low-income patients, and more likely to care for sicker patients.

Why would practices caring for more chronically ill patients be more likely to participate in an ACO? Because spending and quality benchmarks are based on the practice’s prior costs, ACOs allow providers with high-cost patients to capture savings by improving the care of sub-optimally managed patients. In the short term, this will likely benefit these patients, but may only result in a 1-time gain in efficiency for the practice.

However, further research will have to investigate if, once physicians join ACOs, they start cream-skimming healthier patients. It is also unknown if vulnerable populations within ACOs benefit as much as more advantaged groups, and whether physicians and hospitals that serve a disproportionate share of low-income, minority, and clinically severe populations see gains from shared savings on par with other practices. As ACOs, bundled payments, and other payment reforms gain a foothold in the market, continued monitoring is needed to safeguard vulnerable patients and to guard against unintended consequences.