Advancing Equity For The Dually Eligible Population In Alternative Payment Models

Andara Puchino

Editor’s Note

This article is the latest among a series in response to the latest developments in policy and research affecting the dual-eligible population. Other authors will contribute to the series as well.

The series is produced with the support of Arnold Ventures. Included articles are reviewed and edited by Health Affairs Forefront staff; the opinions expressed are those of the authors.

The series will run through August 30, 2022; submissions are accepted on a rolling basis.


The nearly one in five Medicare beneficiaries dually enrolled in Medicaid (duals) are a medically and socially vulnerable patient population who experience inequitable access to care and marked health outcome disparities. Duals are 5.0 times more likely to have a disability and 2.6 times more likely to be of minority race or ethnicity than their non-dual counterparts. They have markedly worse health outcomes, being two times more likely to be hospitalized or die after adjusting for comorbidities, than nondual beneficiaries.

There is strong evidence that, compared to non-duals, duals have less access to high-quality care including some of the most innovative accountable care organization (ACO) models. Duals are also less likely to access needed specialty care than other Medicare beneficiaries. Furthermore, the standard risk adjustment and payment model used by Medicare to pay ACOs and Medicare Advantage plans, as well as to adjust clinician performance measures in the Merit-Based Incentive Payment System (MIPS), systematically underpredicts dual enrollees’ annual Medicare costs by as much as 35percent, effectively institutionalizing underpayment of providers and plans caring for this population.

A key way of reducing disparities in outcomes for dual enrollees is by improving their access to high-quality medical care and innovative payment and care delivery models within Medicare. As value-based and alternative payment models, along with integrated care delivery approaches such as ACOs, become the new normal among both public and private payers, there is a growing need to ensure these strategies actively promote equity in care access and quality for socially and medically vulnerable patient populations.

In this article, we argue that the Centers for Medicare and Medicaid Services (CMS) should take action within its quality measurement and incentive programs to remedy such inequities by updating current incentive structures within Medicare that disincentivize care of duals and to instead reward clinicians and health systems for providing high-quality care to this population.

Quality measurement and payment—when conducted with an equity lens—can provide clinicians and health care leaders with the roadmap for where to focus their energy. Sharpening that focus is essential to achieving improvements in patient health outcomes, especially those that are due to unequal access to high-quality care or to lack of health care investment in disadvantaged communities. If we truly aim to reduce longstanding and pervasive racial, income-based, ability-based, rural-urban, and other inequities in health, our goal cannot just be to do no harm. Rather, we must actively shift these programs to incent a meaningful focus on dual enrollees and their specific needs.

Why Quality Measurement?

It is perhaps a truism that we can’t improve what we can’t measure, but measurement also provides a point of focus in the complex, chaotic ecosystem of health care. As first demonstrated by Avedis Donabedian, one way to prioritize and coordinate improvement efforts is to anchor the work of health care professionals and policy makers in concrete, measurable goals related to the modifiable structures and processes of medical care for our most vulnerable.

Of course, it is also a truism that once a “test” is created, people will teach to it, sometimes to the exclusion of all else. Since that behavior is unlikely to change, it should instead be leveraged. We should work hard to come up with the right test—one that is simple, important, and improves health equity for vulnerable and disadvantaged patients.

What does the right test look like? To answer this question, let’s start with a few key principles:

1. Measure What You Want To Change

If we want to leverage value-based and alternative payment models to improve equity, then we should define and measure the specific forms of equity we want to improve, such as equal care access and quality for disadvantaged patients. Currently, there are thousands of quality measures available to clinicians and practices from CMS alone. Under the MIPS, practices can choose from nearly 400 measures on which to report. Nonetheless, multiple studies indicated that the current MIPS measures appear to exacerbate rather than reduce inequities in payments to clinicians with high caseloads of dual, poor, and racial or ethnic minority enrollees, potentially threatening their access to care.

What are equity-focused measures? Such measures may specifically focus on duals, in areas where there are particularly striking inequities in access, quality, or modifiable clinical outcomes. For example, dual enrollees with disabilities face many more barriers in accessing primary and specialty care in the ambulatory setting than their peers—perhaps because clinicians and practices are unaware of patients’ basic needs or the accessibility requirements they are entitled to under the American with Disabilities Act (ADA). In addition, many clinical practices in the US are not functionally compliant with the ADA. Quality measures may assess practice improvements for accessibility and accommodations for patients with disabilities, training for staff on disability-relevant needs, and patient experience scores among dual enrollees with disabilities.

Diabetes control is another salient example: Duals with diabetes have some of the highest rates of diabetes complications compared to patients of any other insurance payer groups. Prior research indicates that regular ongoing ambulatory care evaluation and management of patients with diabetes by primary care and specialist clinicians is crucial for their care quality and clinical outcomes. A measure that specifically incents clinicians and ACOs to focus on improving diabetes control for duals in addition to broader population measures will lead practices to prioritize engaging this vulnerable population.

Although setting up in-office appointments with a diabetes educator may be sufficient for many patients, establishment of a community health worker program to work with patients in their homes on medication regimens and food choices may be needed to fully engage dual beneficiaries. Neither are bad programs, but a community health worker program is much more likely to improve care and outcomes among those who need it most. However, the community health worker program may require a larger investment of resources; financial incentives tied to quality measures may help justify and finance such investments.

An equity-focused measure could also explicitly measure gaps in care access and quality for disadvantaged patients such as duals. In addition to measuring absolute improvement within the broader population of interest, such a measure would compare the gap between dual and non-dual beneficiaries and ask practices to close the gap. This approach would also create an additional incentive to help ensure the rising tide of quality improvement at the practice level does indeed lift all boats.

This focus highlights the importance of including both equity and excellence in quality measurement. Excellence incentivizes improving average quality for all; equity incentivizes closing the gaps for vulnerable groups. Together, measurement of excellence and equity incentivize quality improvement for everyone in a way that ensures vulnerable and disadvantaged patients are not left behind. In this way, quality improvement should not be viewed as a zero-sum game where one group’s gain is another group’s loss.

Finally, an equity-focused measure could be based on access to care rather than processes or outcomes of care, given that we know access is a key part of achieving optimal health outcomes. Practices could be directly rewarded for caring for a high proportion of high-risk patients, rather than disincentivized from doing so under a value paradigm.

2. Be Cautious Of Unintended Consequences Of Cost Measurement

Using costs as the primary measure of achievement in value-based payment programs has the potential to create significant unintended consequences. We do not want—and should not incent—cost reduction that is harmful to patients. Cost measurement should thus be done carefully and intentionally, and always in concert with key quality metrics.

In addition, many Black and Hispanic patients, as well as patients living in poverty, with disabilities, and in rural areas already have inappropriately low medical use in the outpatient setting that reflects unmet need. Thus, pushing across-the-board cost reductions could worsen inequities. Indeed, recent studies validate this concern, finding that, although providers in both the Shared Savings Program and Maryland Global Payment Model saved more money among dually enrolled beneficiaries than they did among nondual beneficiaries, providers also delivered lower quality of care among duals. This indicates that the priorities of policy makers are not translating to a focus on quality improvement for vulnerable patients.

Similarly, although cost reduction in the Medicare program may be a priority for policy makers, quality improvement is of far more importance to patients and to the clinicians who care for them. Few clinicians and even fewer patients are positively motivated by programs that aim primarily to reduce health care spending. Focusing on quality measurement and improvement rather than on cost reduction creates much better conditions for clinician and patient buy-in and engagement. Quality measures could be selected where improvements are likely to reduce costs—for example, with a measure of preventable hospitalizations—and achieve the same focus on reducing use but in a less blunt manner.

3. Reward Both Average And Equitable Quality Improvement Rather Than Just Achievement

Beyond selecting measures, there are also equity implications in how we define success on those measures. If we measure the proportion of patients who meet a hypertension control goal, we are incenting practices to find patients on the margin of control and make small tweaks to their regimen to meet a specific cutoff. If we define success as improving average blood pressure in a population and on reducing the proportion of people with extremely high blood pressure, we incent practices to seek out their most complex patients for interventions to improve quality for all, including the most vulnerable and difficult to reach. While based on the same underlying data collection, the latter approach is likely to have broader, deeper implications for equity in improving care for the most at-risk patients.

Rewarding quality improvement also has the benefit of avoiding the need for risk adjustment, which as discussed below, has complex intersections with equity. Focusing on improvement also creates more meaningful opportunities for collaboration because one group’s success does not diminish the chances of success for another.

4. If Comparisons Must Be Made, Risk Adjust With An Intentional Focus On Equity

Comparisons need not necessarily be made between clinics or hospitals with highly dissimilar patients, as noted above, if programs focus on improvement. Nonetheless, if comparisons must be made for either statutory or strategic reasons, risk adjustment should be constructed with an intentional focus on equity. To achieve equitable risk adjustment, conditions that disproportionately impact vulnerable populations need to be included in risk models. Elements such as functional status, cognitive status, and social risk are critical components of outcome models that will more fairly reflect the resources needed by patients, and clinics’ actual performance, thus preventing the assessment of unfair penalties and underpayment of clinics disproportionately serving vulnerable patients.

Researchers and policy makers alike have increasingly recognized the impact of social determinants on health outcomes. Given the profound impact of a lifetime of poverty, racism, and stress on people’s bodies, it is unrealistic to think that a single hospitalization or clinical encounter can equalize clinical outcomes between historically marginalized populations and their peers. So, we must continue working to improve both the underlying social conditions that create inequity and the ongoing ways in which the health care system perpetuates it. And in this work, we must at the same time, take social risk into account when using quality measures for comparative purposes. Failing to do so simply takes resources away from clinicians serving the highest-need populations, potential worsening inequity.

And, of course, just as measures can control for factors that are outside the control of the health system, models can explicitly incent action to remedy those factors that are under the control of the health system, especially the well-documented unequal access to high-quality care and systemic underpayment of clinicians and practices who serve duals.

Great Potential

Once the right measures are selected, and the best use and adjustment of those measures identified, policy makers and system leaders must design programs to use these measures in a way that improves equity. Currently, the vast majority of Medicare beneficiaries are covered by Medicare Advantage (MA) plans or cared for by ACOs in the traditional Medicare program—and we can expect nearly all beneficiaries to fall under such policies by the end of the decade. In these two programs, there is great potential to focus on quality improvement efforts that incorporate equity as a key priority. We are encouraged by CMS’s current ongoing efforts to do this in both MA and ACO programs.

Authors’ Note

Dr. Joynt Maddox receives research support from the National Heart, Lung, and Blood Institute (R01HL143421) and the National Institute on Aging (R01AG060935, R01AG063759, and R21AG065526), and from Humana. She also serves on the Health Policy Advisory Council for the Centene Corporation in St. Louis, Missouri. Dr. Johnston receives research support from the National Institute on Aging (R21AG065526) and the National Institute of Mental Health (R01MH125820).

Next Post

Keto diet vs low carb: what's the difference?

Keto diet vs low carb: what’s the difference and which is better? Although the two diets both involve cutting carbs and can help with weight loss, the keto diet (short for ketogenic) is a far more restrictive way of eating and involves limiting carb intake and eating a high amount […]