Implications of Medicare Advantage for Patients With Alzheimer Disease and Related Dementias

The care of patients with Alzheimer disease and related dementias (ADRD) is complex and costly.^1,2 The number of US individuals with ADRD is growing rapidly and is estimated to increase from 4.5 million in 2010 to 13.8 million in 2050, with mean per-person health care costs often 3 times greater for patients with ADRD than for those without ADRD.³ Patients with ADRD and their caregivers often report that their care is uncoordinated and fragmented, and evidence suggests that burdensome transitions are frequent.² In the end, these transitions are often inefficient and only serve to cost more health care dollars without enhancing quality.

The work by Park and colleagues⁴ highlights the differences in health care utilization for patients with ADRD enrolled in traditional Medicare (TM) fee-for-service plans vs those who are enrolled in capitated Medicare Advantage (MA) plans. The proportion of Medicare beneficiaries enrolled in MA plans increased from 10.2 million in 2009 to 18.2 million in 2017.⁵ Some policy observers contend that capitated payments to MA plans provide strong incentives to provide higher-value care and allow flexibility for plans to invest in care coordination efforts and enhanced chronic disease management, such as social determinants of health spending. In contrast, critics of MA have expressed concerns that capitated payments may spur plans to restrict access to care and may, in fact, lead to the disenrollment of beneficiaries with complex health care needs, such as those with ADRD. Unfortunately, few studies have compared the quality and outcomes of care in Medicare Advantage and TM for patients with complex health conditions.

The study by Park et al⁴ was conducted using longitudinal data between 2010 and 2016, and the findings suggest that patients with ADRD enrolled in MA plans had 22.3 fewer visits to a medical provider and 2.3 fewer visits to the outpatient hospital setting than those enrolled in TM fee-for-service plans. Despite the reduced care, the analyses failed to detect statistically significant differences in health care status or care satisfaction between MA and TM beneficiaries, although the estimates could not exclude meaningful differences in these measures between the 2 populations. The authors, therefore, speculate that MA may have achieved cost savings through better coordination of care without jeopardizing health care quality.

Although this theory is possible, there are several limitations to this work⁴ that merit discussion. The authors rightly point out that MA plans have long had a history of skimming the cream off the top, offering incentives that attract healthier populations than those enrolled in TM plans. Although the investigators identified no differences in care satisfaction between those enrolled in MA vs TM, they could not account for those patients who potentially left, or were coerced to leave, MA plans to reenroll in TM. This is important, because prior work^6,7 has shown that MA beneficiaries who are dually eligible for Medicaid or who receive skilled nursing facility or home health care have high rates of disenrollment in the following year.

It Is also important to note that the population with ADRD is clinically heterogeneous, and patients with dementia can have a wide range of cognitive and functional impairments that range from those who are completely independent to profound disability. Park et al⁴ did attempt to control for this variability by including a measure of function in the form of an activities of daily living metric. Nevertheless, this functional scale is unlikely to account fully for the differences in clinical severity between TM and MA beneficiaries with ADRD. The analysis by Park et al⁴ identified substantially fewer functional limitations in the MA population, perhaps explaining why those patients had fewer health care visits.

The study by Park et al⁴ of patients with ADRD suggests that those enrolled in MA plans received fewer health services compared with patients enrolled in TM. These differences in care could reflect selection of a less-impaired slice of the population with ADRD, differential disenrollment of high-need patients with ADRD out of MA plans, restricted access to effective care, or more-efficient health care delivery. More research will need to be done to fully understand the implications of managed care for patients with ADRD and other complex medical conditions.