Low-income, older adults, and people with disabilities are eligible for both Medicaid and Medicare. Although it might seem that having double coverage would be helpful, the reality for decades has been the opposite. Having multiple payers creates a fragmented system where no one entity is responsible, or has the financial incentive, to improve care. That leaves the sickest and poorest among us with a system that is costly, hard to access, and unable to improve the health of vulnerable people.  

“Integrated care” has long been touted as the solution, and Congress and the Centers for Medicare and Medicaid Services (CMS) have been promoting it since before the Affordable Care Act in 2010. Under an integrated care model, one entity (for example, a single managed care plan) is responsible for managing a beneficiary’s Medicaid and Medicare coverage, and, ideally, all their physical, behavioral, acute, and long-term care benefits. In 2022, two different bipartisan bills–the Comprehensive Care for Dual Eligible Individuals Act and the Advancing Integration in Medicare and Medicaid Act–were introduced (but not passed) to encourage more integrated care. Additional legislation is anticipated this year. 

While the policy community has embraced integrated care, evidence on the benefits of integrated care plans has been elusive. To examine whether there is a benefit, LDI Senior Fellow Eric Roberts recently examined the effects of enrolling in an integrated plan in a study published in JAMA Health Forum. Roberts and coauthors’ findings were not conclusive. They found that enrollment in an integrated plan was associated with increased use of home and community-based services when compared to a comparison group, but did not find any association with changes in care management and coordination, medication use for chronic conditions, hospital stays, or follow-up outpatient care after a hospital stay. Roberts discussed the study further in a series of questions below.

What led you to conduct this study?

Roberts: The dual-eligible population frames an important set of questions in health economics that have real-world implications for the care of vulnerable populations. 

Economists think a lot about where well-intended policies and programs can produce misaligned incentives. In the U.S. system, a prototypical example of misalignment is health insurance for the dual-eligible population. Although Medicare and Medicaid are both government insurance programs, they are financed separately and cover different services (for dual-eligible beneficiaries, Medicare pays for hospital and outpatient care while Medicaid pays for long-term nursing home care).

This system produces misaligned incentives and weakens incentives to coordinate care across programs. For example, Medicaid does not capture savings if the services it covers prevent hospitalizations, since hospital care is financed by Medicare. Likewise, Medicare cannot capture savings by preventing long-term nursing home stays after a hospitalization, because those costs fall on Medicaid. From a patient care perspective, this lack of coordination is concerning because dual-eligible beneficiaries have complex needs and may benefit substantially from coordinated care.

Integrated care plans are a potential solution. I wanted to see if these models worked as theory would predict.

Did the findings surprise you?

Roberts: Yes and no!

A fascinating aspect of research is seeing where theory does–or does not–bear out in practice. In theory, having one plan manage Medicare- and Medicaid-covered services should produce incentives to deliver an efficient mix of services. For example, an integrated plan should have an incentive to invest in care coordination and supportive services to prevent an individual going to the hospital or a nursing home, where care is very expensive. But actually implementing this requires time to build infrastructure, establish lines of communication between care managers, etc. While we saw some increases in who received community-based services and supports in the integrated plan, there was no evidence of improved care coordination nor of reductions in hospitalizations. Overall, the gains from integrating care were quite modest, though I should note that we looked only at outcomes over three years. Also, despite best efforts, it is really difficult to compare the groups who choose integrated care to those who do not and some of the lack of better outcomes in the integrated groups may come from underlying differences in the groups of beneficiaries.

What have other researchers found about dual-eligible beneficiaries?

Roberts: Most other studies of integrated care plans found similarly modest impacts on care. For example, most evidence suggests that integrated plans are effective in expanding who receives long-term services and supports in the community. There is also some evidence that, over the long term, dual-eligible beneficiaries in integrated plans might become less likely to become long-stay nursing home residents. But when it comes to meeting patient needs after a hospital stay, preventing hospitalizations, or improving chronic disease management–all important for dual-eligible beneficiaries–integrated plans overall don’t seem to perform better than the status quo.

What should future studies look at?

Roberts: We need to address two major gaps in the research.

First, we need to understand more about the experiences of patients enrolled in integrated care plans. We need to learn more about which aspects of integrated care plans are, or are not, meeting the day-to-day needs of patients. Research centering on the experiences of patients and caregivers will be indispensable for identifying what is needed to improve how integrated care plans function in practice.

Second, the research on integrated plans needs to keep pace with a rapidly evolving policy environment. The last five years have seen major changes in how federal and state policymakers have approached integration policy for dual-eligible beneficiaries, and more reforms are on the horizon. It will be important for the research to keep up so it can guide policymaking and implementation strategy.

Given the great needs of the dual-eligible population, should we expect integrated care to save money? And for whom? 

Roberts: This is a great question! At first blush, it would seem that plans delivering better care could realize savings from avoiding patient hospital and nursing home admissions. And certainly that is the hope over the long run.

But your question underscores the challenge: A large proportion of dual-eligible beneficiaries have complex medical and social needs that drive their high costs. Moreover, these needs may reflect the deeply entrenched effects of poverty on health. Addressing these needs often requires investment and time, making it unrealistic to expect that integrated care plans will save money, at least in the short- to medium-term. 

It is more realistic to consider that integrated care plans may help to rebalance the composition of Medicare and Medicaid spending so that plans manage patient needs more proactively through investments in community-based services. In turn, plans may be able to spend somewhat less over time on hospital and nursing home care. Moving this needle will be a bit like steering an aircraft carrier–deliberate, yet incremental.

Are there other lessons for policymakers in the short-term?

Roberts: The dual-eligible population is extremely diverse–both demographically and in terms of its health profile. A one-size-fits-all approach for integrated care programs will not likely work. I think there is a growing appreciation that integrated care models need to be tailored to different populations (e.g., people receiving long-term nursing home care versus those living in the community, people with intellectual and developmental disabilities, etc.). Embracing these nuances will help us design better models that meet the diverse needs of this population.

---

The study, “Changes in Care Associated with Integrating Medicare and Medicaid for Dual-Eligible Individuals,” was published on December 1, 2023 in JAMA Health Forum. Authors include Eric Roberts, Lingshu Xue, John Lovelace, Chris Kypriotis, Kathryn L. Connor, Qingfeng Liang, and David Grabowski.

---

Author

---