“REACHing” for Equity — Moving from Regressive toward Progressive Value-Based Payment

Suhas Gondi, M.D., M.B.A.,  Karen Joynt Maddox, M.D., M.P.H.,  and Rishi K. Wadhera, M.D., M.P.P.

The movement toward value-based payment has been a defining feature of U.S. health care reform during the past decade. Despite substantial enthusiasm and investment, however, these efforts have been largely disappointing. Most Centers for Medicare and Medicaid Services (CMS) value-based payment models have failed to meaningfully reduce health care expenditures or improve quality of care.

Perhaps more concerning is that many value-based payment programs have been regressive, which has hampered the pursuit of health equity. For example, Medicare’s Merit-Based Incentive Payment System (MIPS) has disproportionately penalized outpatient clinicians who care for poor adults.¹ Similarly, all three of Medicare’s hospital value-based programs — the Hospital Readmissions Reduction Program, the Hospital Value-Based Purchasing Program, and the Hospital-Acquired Condition Reduction Program — have transferred resources away from safety-net hospitals and potentially widened inequities in care. By more frequently penalizing institutions caring for high proportions of Black adults, many of these programs have also unintentionally perpetuated structural racism.²

Value-based payment initiatives have failed to advance health equity in large part because equity wasn’t prioritized during their design and implementation. Since many such payment programs are budget-neutral, they create winners and losers — rewarding some providers at the expense of others. Preexisting structural differences often influence who wins and who loses. Practices and hospitals with the infrastructure and resources to rapidly adapt to logistic and reporting requirements are much more likely to succeed in new programs. In addition, because the spending targets used in some payment models are based on a provider’s prior spending levels, providers whose patients have historically used fewer services because of a lack of access to care may face unrealistic expectations. When such disadvantages are ignored, baseline disparities may become structurally embedded in new payment mechanisms, which unfairly advantages certain providers. These issues are magnified by Medicare’s current risk-adjustment approach, which doesn’t adequately account for all medical and social risk factors that influence spending and outcomes. In failing to explicitly consider equity, value-based payment programs implicitly prioritize well-resourced clinicians and health systems.

The responses elicited by value-based models also have important consequences for health equity. Although the financial penalties included in these models are intended to encourage providers to deliver high-quality care, the desire to avoid losses may also promote “gaming” that disproportionately harms low-income and historically marginalized populations. For example, recent evidence suggests that Medicare’s Comprehensive Care for Joint Replacement model may have impeded access to knee and hip replacements for Black adults (who tend to have a higher burden of medical and social risk factors than White adults), thereby widening long-standing racial disparities.³ In the MIPS program, well-resourced providers tend to strategically choose quality measures that maximize their scores on the basis of their existing performance — a process that often requires data analytics, external consultants, and other investments — which allows them to earn financial rewards without necessarily improving care.⁴ This dynamic disadvantages providers who care for low-income populations, since they tend to have fewer resources to dedicate to score optimization.¹ More broadly, value-based models have encouraged increases in coding intensity (which do not necessarily reflect true changes in patients’ medical complexity) that most likely benefit well-resourced health systems with more robust coding capabilities.

A bright spot in the value-based payment era has been accountable care organizations (ACOs) — groups of providers that are given incentives to reduce spending below a benchmark — some of which have produced savings for Medicare. Health-equity concerns persist, however. Recent evidence suggests that some ACOs may strategically drop “high risk” beneficiaries (e.g., those with multiple chronic conditions and high expected medical spending) or clinicians whose panels consist of large numbers of such patients in order to reduce spending and increase their chances of earning shared savings.⁵ In the absence of explicit incentives to invest in equity, value-based payment models can elicit responses that widen disparities.

In an important shift, the Center for Medicare and Medicaid Innovation recently announced a new model — the ACO Realizing Equity, Access, and Community Health (ACO REACH) model — partly in response to concerns about the inequitable effects of value-based payment programs. This model explicitly names promoting equity — not just value — as a central goal. Several provisions of ACO REACH could help advance health equity.

First, the model includes a new “health equity benchmark adjustment” that supports ACOs caring for socioeconomically disadvantaged patients. This approach represents a marked shift from earlier payment models; it acknowledges that providers may need to spend more — not less — to care for members of marginalized populations. Specifically, CMS will increase spending benchmarks by $30 per month for each ACO member in the top decile of disadvantage. A smaller downward adjustment ($6 per month) will be applied for each member in the bottom five deciles. This calculation will incorporate both individual and neighborhood-level markers of socioeconomic disadvantage. The net effect will be higher spending benchmarks for ACOs caring for the most disadvantaged patient populations, which means a higher likelihood of shared savings for these provider groups. By capturing a broad range of social risk factors affecting health care use, this approach should reduce disincentives to serve members of marginalized groups.

Second, ACO REACH will require participating ACOs to develop and implement a health equity plan that involves identifying disparities in their patient populations, establishing an equity strategy, and adopting initiatives to reduce disparities. CMS is therefore taking a novel step toward using payment reform as a lever to foster local efforts that promote equity.

Third, CMS is requiring ACOs to collect and submit data on patient-reported demographics and social determinants of health. The lack of granular and reliable data on race, ethnic group, and health-related social needs has hindered health-equity efforts. Mandating collection of these data could facilitate the evidence-based implementation of equity-focused interventions.

Although these provisions represent an important shift, their effects will be limited to providers and patients who participate in the new model. For these reforms to have broader influence, policymakers will need to decide which provisions to apply to other payment models and in what form. This process will require careful implementation and rigorous evaluation to answer key questions. One such question will be whether the proposed benchmark adjustment is sufficient to encourage providers to care for low-income patients and members of other underserved groups and large enough to allow providers to meaningfully invest in the health of these populations. The value-based care movement has traditionally prioritized reducing spending, but advancing health equity demands spending more on underserved groups with unmet needs. If the benchmark adjustment isn’t adequate to finance this increased spending, ACO REACH may not achieve its goals.

Another consideration will be the types of investments that providers make in response to the benchmark adjustment. Facing incentives to care for disadvantaged patient populations, ACOs may respond in productive ways (for instance, improving care delivery for such populations) or in ways focused on profit generation (for instance, marketing more aggressively to them). In addition, the types of organizations that participate in ACO REACH will matter. Voluntary participation has limited the effects of other payment models, since providers who stand to benefit tend to join and those who perform poorly tend to drop out. Finally, it remains to be seen whether the health equity plan requirement will motivate real action. Although this idea is promising in theory, similar requirements — such as community-benefit and needs-assessment requirements for nonprofit hospitals — have proven weak in practice. Without proper oversight, this provision may become another administrative checkbox.

Value-based payment models implemented over the past decade have often been regressive, moving dollars away from patients, providers, and communities with fewer resources and toward those with more. ACO REACH reflects policymakers’ efforts to mitigate this unintended consequence. It also lays a foundation for further steps to address the long history of underinvestment in the health of low-income and marginalized populations. Could this new approach to value-based payment be a tool for redistributing health care resources in a progressive way that meaningfully advances health equity? This may be the central question in the next decade of payment reform.

Disclosure forms provided by the authors are available at NEJM.org.

This article was published on July 9, 2022, at NEJM.org.

Author Affiliations

From the Richard A. and Susan F. Smith Center for Outcomes Research, Beth Israel Deaconess Medical Center (S.G., R.K.W.), and Brigham and Women’s Hospital (S.G.) — both in Boston; and the Cardiovascular Division, John T. Milliken Department of Medicine, School of Medicine, and the Center for Health Economics and Policy, Washington University in St. Louis, St. Louis (K.J.M.).

Supplementary Material