People who qualify for both Medicare and Medicaid — known as dual-eligible beneficiaries — frequently have complex and costly health care needs. Since 2011, the federal Centers for Medicare and Medicaid Services (CMS) has collaborated with 13 states to integrate benefits for this low-income population through a demonstration called the Financial Alignment Initiative. The most common demonstration model has a simple goal: combine Medicare and Medicaid benefits in one managed care plan with a financial incentive to manage members’ overall health care use and outcomes across medical, behavioral health, and long-term care. A federal–state contract holds each participating plan accountable for coordinating care, improving quality, and curbing spending growth by linking plan reimbursement to measures.
CMS announced this week that the demonstration can be extended for current participating states and will open to new states interested in launching integrated plans or other alignment approaches. More than 1 million of the nation’s 7.6 million dual-eligible beneficiaries with full Medicaid benefits have been eligible to participate in the initial demonstration plans. Administering these integrated benefits involved significant operational changes for both Medicare and Medicaid.
As states consider strategies for aligning benefits, it is important to consider the interim evaluation results for four states that launched new Medicare–Medicaid managed care plans (California, Illinois, Massachusetts, and Ohio). Here’s what evaluators have learned from the early years of these plans:
While fewer beneficiaries participated in demonstration plans than expected, most plan members reported high levels of satisfaction.
When plans launched, states could transfer dual-eligible beneficiaries into demonstration plans unless beneficiaries opted out. More beneficiaries than expected opted out of demonstration plans immediately or disenrolled later. Concerns about provider access and uncertainty about the benefit changes factored into beneficiaries’ decisions to opt out. For Medicare, beneficiaries could return to fee-for-service Medicare or other Medicare Advantage plans; for Medicaid, some states still required Medicaid managed care participation for beneficiaries who opted out of demonstration plans.
Across all demonstration states, about a quarter of the eligible 1.2 million beneficiaries enrolled, with participation rates varying among states. Ohio, which had the highest participation rate among eligible beneficiaries (62% in October 2016), passively enrolled beneficiaries into demonstration plans a year after the state had mandated Medicaid managed care for dual-eligible beneficiaries. Participation among eligible individuals was lower than expected in California (23%), which also implemented mandatory Medicaid managed care in demonstration counties, and in Illinois (29%) and Massachusetts (13%), where beneficiaries could completely opt out of managed care for both Medicaid and Medicare.
Among beneficiaries who participated, surveys indicate that the percentage of members who gave their plan the highest ratings increased from 51 percent in 2015 to 59 percent in 2016. This 2016 figure is comparable to performance ratings for other Medicare Advantage plans.
Increased accountability for care coordination processes is difficult to link to improved outcomes.
The demonstration focuses on improving care coordination. A small percentage of payments to plans was conditional on meeting care coordination metrics. However, these metrics do not hold plans accountable for linking care coordination to better health outcomes and may actually limit plans’ flexibility to optimize care coordination. Plans were required to complete health-risk assessments for their members, which became expensive and time-consuming when members were difficult to reach. Plans reported that they provided documented care coordination plans for 50 percent to 60 percent of members. Members were often unaware of these care coordination services. In half of all plans, fewer than 30 percent of members reported receiving assistance with care coordination.
Much of the responsibility for ongoing benefit coordination falls to individual care coordinators, with variation in whether plans provide significant resources, like robust electronic heath records, to support these efforts. Data indicate considerable turnover among care coordinators (17% in Ohio and 22% in Illinois by the second demonstration year).
Similar to trends across the entire Medicare population, hospital use declined among dual-eligible beneficiaries who qualified to participate in the demonstration plans.
In Illinois, Massachusetts, and Ohio, inpatient use decreased among dual-eligible beneficiaries participating in the demonstration. This pattern follows national trends of declining inpatient use across the entire Medicare population. However, it is difficult to determine whether the demonstration is accelerating this decline, especially in states where only a small share of eligible beneficiaries participated. Evaluation reports suggest there may have been modest Medicare savings in Illinois (about 2%), but none in California or Ohio.
So far, there aren’t enough data to understand how the demonstration is affecting beneficiaries who use long-term services and supports.
MedPAC reported that beneficiaries with long-term nursing home use are as likely to participate in demonstration plans as other eligible beneficiaries, suggesting that plans are reaching these populations. The evaluation reports mixed findings in Illinois and Ohio on whether the demonstrations lowered or increased use of long-term nursing home services. In plan member surveys, beneficiaries with long-term services and supports report some difficulties accessing durable medical equipment, which can be a critical service for beneficiaries with mobility impairments.
After the significant effort involved in launching the demonstrations, it is frustrating to have few clear indications of whether they are achieving better quality care with greater efficiency. Reports to date find it plausible, but not definitive, that plans are making progress toward important goals like moderating inpatient use. Although it is concerning that so many eligible beneficiaries have opted out, the smaller launch provided an opportunity to work out operational details. One early lesson learned is how disruptive changing benefits can be for dual-eligible beneficiaries. Further upheaval will likely be caused by ending state demonstrations and reassigning those members to new coverage. The best path forward may be to use extensions to allow the demonstration plans more time to develop while awaiting more definitive results about outcomes.
The author would like to acknowledge Marc Cohen, Ph.D., David Grabowski, Ph.D., Ann Hwang, M.D., Leena Sharma, M.P.P., and David Stevenson, Ph.D. for their input.
