Integrating Medicare and Medicaid Coverage for Dual Eligibles: Recommendations for Reform

In the United States, 12.2 million individuals known as dual eligibles receive health insurance through Medicare and Medicaid.¹ Dual eligibles qualify for Medicare based on age, disability, or a diagnosis of end-stage renal disease, and qualify for Medicaid because they have low incomes and assets. Due to advanced age, health risks, and low socioeconomic status, dual eligibles often have complex care needs, including needs for chronic disease management, behavioral health care, and long-term care. Dual eligibles are also a costly population to insure: in 2020, they accounted for 34% of Medicare and 30% of Medicaid spending but represented 19% of Medicare and 14% of Medicaid beneficiaries.¹

Policymakers and clinicians have expressed concern that insurance coverage for dual eligibles is poorly configured for coordinating care because Medicare and Medicaid are administered separately and pay for different services.² Specifically, Medicare is a federal program that covers inpatient, outpatient, and pharmacy spending, while Medicaid is administered by states (with federal and state funding) and covers long-term care, community-based services and supports, and some behavioral health care.^1,2

Concerns that this bifurcated system contributes to fragmented and inefficient care have led policymakers to test integrated coverage models, in which a single managed care organization covers Medicare and Medicaid spending for dual eligibles.² The expectation is that integrated models have incentives to deliver care efficiently and are easier for patients and caregivers to navigate.³ Currently, three types of integrated models exist. However, the structure and availability of these models vary, and together, they cover only 10% of dual eligibles.²

Efforts to reform coverage for dual eligibles are accelerating. Several reform proposals have been introduced in Congress, and 2022 a bipartisan group of Senators requested input from policy experts on coverage issues for dual eligibles, existing integrated models, and the feasibility of expanding these models.

This Viewpoint provides an overview of existing integrated models and their performance as measured by quality, utilization, and spending. It then outlines several ways policymakers could improve coverage for dual eligibles by expanding enrollment in existing integrated models and enacting pragmatic reforms to promote higher quality, efficient, and more equitable care in them.

Existing Integrated Models and their Performance

Three integrated models currently serve dual eligibles. The first and most established of these models, the Program of All-Inclusive Care for the Elderly (PACE), coordinates medical, long-term care, and social support services via community-based adult day health centers. PACE serves frail older adults—almost all of whom are dual eligibles—who qualify for a nursing home level of care. Enrollment in PACE grew from approximately 33,000 to 60,000 individuals between 2013 and 2020, and PACE programs now operate in 30 states.⁴ However, it remains the smallest integrated model by number of enrollees.

Second, Fully Integrated Dual Eligible Special Needs Plans (FIDE-SNPs), which are a subset of Medicare Advantage Dual Eligible Special Needs Plans (D-SNPs), have capitation contracts with states to cover Medicaid-financed long-term care and behavioral health services.² Thus, FIDE-SNPs manage Medicare and Medicaid spending. Enrollment in FIDE-SNPs grew from 135,000 enrollees in 2016 to 392,000 enrollees across 12 states in 2023.⁵ These plans are a focus of evolving integration policy. For example, new regulations by the Centers for Medicare and Medicaid Services (CMS) require FIDE-SNPs to exclusively serve aligned enrollees (i.e., individuals receiving Medicare and Medicaid from the same plan) by 2025.⁶

Third, Medicare-Medicaid Plans (MMPs) are managed care plans that receive capitated payments for Medicare- and Medicaid-covered services. These plans were temporary demonstrations that states tested in partnership with CMS starting in 2013.⁴ Eleven states tested MMPs, although by 2023 MMPs operated in 9 states and served approximately 298,000 enrollees. All MMPs are scheduled to end by 2025 and are expected to be converted into D-SNPs.⁵

While evidence on integrated models is mixed, several studies found an association between integrated coverage and reduced nursing facility use, which is typically the most expensive Medicaid-financed care for dual eligibles.^3,7,8 Among studies that found a reduction in nursing home use, some also found that enrollees in integrated models received higher levels of community-based supportive services and outpatient care.^3,7 These findings suggest that integrated models may be effective in reducing nursing home use by meeting dual eligibles’ needs for long-term care and supportive services in the community.

However, the performance of integrated models on other domains of quality and utilization remains less established.⁷ Further, while some studies found little evidence linking integrated models to lower spending, evaluations have been limited by incomplete Medicaid spending data.⁷

Recommendations to expand and improve integrated models

Policymakers have an opportunity to improve coverage for dual eligibles by strengthening existing integrated coverage models and facilitating enrollment in them. Although some analysts called for developing an entirely new integrated program for dual eligibles, an incremental approach that capitalizes on existing models could allow policymakers to implement timely reforms and make refinements an ongoing basis. These reforms could include the following elements.

First, reforms should promote take-up of integrated models by expanding default enrollment, which allows integrated plans to automatically enroll individuals when they first become dual eligibles.⁴ To preserve choice, dual eligibles should remain able to opt out of plans to which they are automatically enrolled. Research on default enrollment in other health insurance settings has found this to be an effective lever for increasing insurance take-up.⁹

Second, states should encourage growth of integrated plans by leveraging the requirement that all D-SNPs have Medicaid contracts. As part of these contracts, states can establish capitation agreements with plans to cover Medicaid long-term care and behavioral health spending (required for FIDE-SNPs).^2,4 Other categories D-SNPs do not fully integrate Medicare and Medicaid spending, and therefore have less capacity to manage dual eligibles’ care.

Third, integrated plans should only serve aligned enrollees, who receive Medicare and Medicaid through the same plan, and limit enrollment to individuals with full Medicaid (72% of dual eligibles),¹ for whom Medicaid covers long-term care and behavioral health care. The Medicaid and CHIP Payment and Access Commission (MACPAC) also recommended this reform, noting that it would incentivize plans to manage the continuum of dual eligibles’ Medicare and Medicaid spending.⁴ Further, plans should develop mechanisms to enable individuals who experience a change in dual eligibility status (e.g., to partial Medicaid, which only covers Medicare out-of-pocket costs) to switch to a companion plan without coverage disruptions.

Fourth, CMS should fund training of navigational aids to help dual eligibles and families understand and select integrated plans, analogous to navigators on Affordable Care Act marketplaces. This would reduce reliance on brokers, who receive enrollment-based commissions, to facilitate plan selection.² Plans should also be required to provide resources to help new enrollees plan understand changes in coverage and provider networks.

Fifth, more reliable data is needed to monitor the effectiveness of integrated models and help dual eligibles who actively choose plans make informed enrollment decisions. To provide such data, CMS should offer technical assistance to Medicaid programs and plans to measure and report on utilization patterns and quality of care in integrated models, with a particular focus on evaluating care for subpopulations of dual eligibles with distinct care needs (e.g., behavioral health conditions or intellectual and developmental disabilities). This evidence could help identify which models advance health equity by improving outcomes for particularly vulnerable subgroups of dual eligibles.

Sixth, policymakers should pursue strategies to promote integration in fee-for-service Medicare, which covered 48% of dual eligibles in 2020.¹ For example, CMS could modify its Medicare Accountable Care Organization (ACO) model—the dominant fee-for-service Medicare alternative payment model—to allow ACO providers to assume financial risk for dual eligibles’ Medicare and Medicaid spending.

Conclusion

Integrating Medicare and Medicaid coverage offers an opportunity to manage spending and improve the quality and equity of care dual eligibles receive. Although research demonstrates benefits of integrated coverage, only 10% of dual eligibles are enrolled in integrated models, and opportunities remain to improve these models. Expanding and reforming existing integrated models, guided by input from stakeholders and ongoing research, could yield tangible and timely improvements in coverage for dual eligibles.