Integrating Care at the End of Life: Should Medicare Advantage Include Hospice?

Since its creation in 1983, the Medicare hospice benefit has been “carved out” of Medicare’s managed care program, commonly known as “Medicare Advantage”. When a Medicare Advantage (MA) enrollee elects hospice, payments for both hospice and other services unrelated to the individual’s terminal condition revert to fee-for-service Medicare, and health plans remain liable only for the supplemental benefits they provide. Although the initial rationale for this approach is unclear, the policy has come to define end-of-life care for a substantial portion of Medicare beneficiaries. Approximately 417,000 MA enrollees died in 2011 (24% of Medicare deaths), almost half of whom used hospice.^a Consistent with broader efforts to integrate health care services across the continuum, the Medicare Payment Advisory Commission (MedPAC) recently discussed the possibility of ending the MA hospice carve-out.^b

Integrating hospice into the MA program has a number of potential strengths and limitations, both of which are discussed below. Moreover, in the context of such a change, important safeguards must be in place to ensure optimal end of life care for Medicare beneficiaries.

Plans currently have a strong incentive to encourage patients with terminal conditions to enroll in hospice, thereby ending the plans’ clinical and financial responsibilities for their care. Removing the hospice carve-out would require plans to coordinate care for all enrollees at the end of life, whether or not they elect hospice, and ideally would encourage plans to integrate hospice and other palliative services with the care they deliver to patients with advanced illness. Perhaps more important, by giving plans greater flexibility in their targeting and delivery of services, eliminating the MA hospice carve-out could reduce the difficult and arbitrary distinctions that Medicare hospice eligibility criteria force clinicians, patients, and families to make about having an expected prognosis of 6 months or less and about forgoing potentially life-prolonging therapies. (Reflecting these challenges to timely enrollment, 28% of Medicare hospice decedents enroll in hospice for three or fewer days.¹) An integrated hospice benefit also could diminish concerns about longer hospice stays in the context of per-diem hospice payments and shift attention to ensuring high quality end-of-life care.

Hospice and palliative services have been associated with higher quality of life, higher patient and family satisfaction, longer survival, and, for some populations, lower Medicare expenditures.^2,3 Instead of being a separate path that must be chosen by the beneficiary and certified by a physician, an enhanced MA benefit including a full array of hospice and other palliative services could be incorporated seamlessly into beneficiaries’ care and driven by their needs and preferences, not by a specific benefit’s eligibility criteria. For example, relative to the current benefit, plans may wish to extend hospice to patients with longer or more uncertain prognoses, offer hospice enrollees concurrent access to a broader range of palliative and therapeutic services, or incorporate palliative services more effectively at earlier points in advanced illness (i.e., not just at the end of life, when hospice is appropriate). Insurers have reported success with similar approaches for their under-65 commercial populations (primarily enrollees with cancer⁴) but have not been permitted to use them for their MA enrollees. A move to include hospice in MA benefits is consistent with Medicare’s current emphasis on eliminating payment silos, decreasing fragmentation across settings, and providing patient-centered care.

Although there are potential strengths with a hospice carve-in (ie, including hospice in Medicare Advantage plan benefits), there also are concerns. Provision of hospice would likely be shaped in part by the broader financial incentives created by capitated payments, underscoring both the importance and challenge of ensuring that risk adjustment methods account for the expected costs of patients with advanced illness. Plans might include only a subset of local hospice agencies in their contracted networks (perhaps negotiating lower rates in exchange for higher patient volume), which could limit choice for beneficiaries with advanced illness. If Medicare Advantage plans negotiate hospice rates that are lower than what Medicare currently pays, changes in the quantity and types of services provided could result and it is unclear how the quality of care could be affected. For example, a hallmark of successful hospice programs is the comprehensive and interdisciplinary nature of the teams that provide care, something that could be undercut if hospice services are offered individually and not as an integrated package. Plans also could conceivably incorporate cost sharing or utilization management into the provision of hospice and palliative services, something the traditional Medicare program has not done for the hospice benefit. More fundamentally, under a carve-in approach, the Medicare Advantage plan and not the hospice agency would drive the care planning process and oversee provision of end-of-life care for hospice enrollees, a change that could lessen the role of clinicians who have the most relevant, specialized experience.

To ensure that beneficiaries receive high-quality care at the end of life under a carve-in model, MA plans must be held accountable for end-of-life care quality. For this to happen, a combination of monitoring, public reporting, and performance incentives must be in place. At present, several key components of federal payment and oversight do not include quality measures with relevance to end-of-life care, including the Healthcare Effectiveness Data and Information Set (HEDIS) measures used to characterize MA plan quality, quality performance standards for Medicare’s Pioneer Accountable Care Organization (ACO) and Shared Savings Programs, and public reporting efforts such as Nursing Home Compare. Although the National Quality Forum endorsed an initial set of end-of-life care quality measures in 2012,⁵ attention should be given to the continued development and use of such measures, especially to ensure their relevance as individuals transition across settings of care. In the context of integrating hospice into MA, one key area for oversight pertains to assessing the adequacy of plans’ hospice and palliative care provider networks (e.g., the hospice agencies and palliative care physicians with which plans contract), something that is already done for MA plans along other dimensions. A complementary approach that would preserve beneficiaries’ freedom of choice more strongly is to offer individuals meeting hospice eligibility standards the ability to disenroll from their MA plans into the traditional Medicare hospice benefit if they determine their end-of-life care needs are not being met. Although such a provision might reduce the incentive for plans to provide high quality end-of-life care, stop-gap mechanisms that penalize plans with relatively high disenrollment rates or require plans to reimburse a portion of beneficiaries’ “out-of-network” care could be devised.

Our primary focus is on the potential integration of hospice into Medicare Advantage, but MA is not the only program aimed at improving the coordination of care for Medicare beneficiaries that excludes hospice. The same is generally true under the new State Integrated Care and Financial Alignment Demonstrations for dually-eligible (Medicare and Medicaid) beneficiaries, where most state proposals carve out hospice or explicitly exclude individuals electing hospice.⁶ Similarly, none of the models authorized for the Medicare Bundled Payments for Care Improvement Initiative include hospice and palliative services among the acute and post-acute services in the bundled payments. Although hospice is included in the spending targets under the Medicare Shared Savings and Pioneer ACO Programs, the extent to which hospice and palliative care will be integrated under these models is unclear and will depend, in part, on whether individuals with advanced illness remain assigned to ACOs by using enough services with ACO-contracted primary care physicians. In other words, ending the MA hospice carve-out could help achieve better integration of hospice and other palliative services into MA, but similar challenges would remain for other populations and in other parts of Medicare.

Since it was created 30 years ago, the Medicare hospice benefit has defined end-of-life care for an increasing number of Medicare beneficiaries at the end of their lives. Although not a prominent feature at the time, carving out hospice from MA has important implications today and potentially undercuts broader improvements in end-of-life care for MA enrollees. MA plans currently have little incentive and limited opportunity to integrate high-quality hospice and palliative services into the care they provide their members, around half of whom do not elect hospice at the end of their life. As the Medicare program moves toward increased integration and accountability, including hospice in the MA program and in integrated care efforts more broadly should be part of this conversation, provided sufficient safeguards can ensure such a change bolsters the quality of care beneficiaries receive at the end of life.