End-of-life (EOL) care is ripe for transformation by accountable care organizations (ACOs), which have the right incentives to tackle the widespread variation in use, quality, and costs that now characterize health care at the end of life. Despite significant attention to ACOs from researchers and policy makers, little is known about how ACOs are approaching EOL care. While ACOs have been shown to affect the use of certain health care services, particularly in the postacute environment, discussions of the impact of ACOs on EOL care are conspicuously absent. Undoubtedly, the lack of attention to EOL care is attributable, in part, to ACOs being relatively nascent entities. However, as ACOs evolve and become more risk-bearing, they are increasingly likely to adopt strategic approaches aimed at improving the EOL experiences of beneficiaries while ensuring efficient use of health care services. In this blog post, we show how the design of the Medicare hospice benefit contributes to the extreme variation in hospice use and describe how the incentives built into the ACO model may yield three strategic and innovative approaches to EOL care.

Current EOL Experiences Are Marked By Wide Variation In Use And A Fixed Menu Of Services

Despite consensus on the value of hospice care for terminally ill individuals, there is substantial variation in the EOL experiences of Medicare beneficiaries and in their use of the Medicare hospice benefit. Although hospice care is a benefit for individuals determined to be in the last six months of life, the median length-of-stay in hospice care is approximately 17 days, while the average length of stay is around 72 days. The variation in hospice use combined with significant increases in Medicare spending on hospice care raise concerns about both underutilization and overutilization of hospice services.

While some variation should be expected as the needs of individuals at end of life are different, part of this variation is attributable to the design of the Medicare hospice benefit. First, the Medicare hospice benefit is a prognosis-based benefit. Using prognosis to determine eligibility for services contributes to variation because terminal illnesses do not always follow a predictable course. This ambiguity around prognosis has contributed to excessively long and expensive lengths-of-stay for individuals with certain diagnoses, particularly those living with dementia. Second, because electing hospice services requires waiving coverage for curative treatment, many individuals only receive hospice care for the last days of life. Thus, structural aspects of the Medicare hospice benefit contribute to variation at both ends of the distribution.

Similarly, hospice payment policy may also contribute to this variation. A beneficiary enrolled in a hospice program has access to a uniform set of services and benefits including medications, medical equipment, supplies related to the terminal prognosis, and care from a specialized interdisciplinary team. Irrespective of the intensity of services required to care for a terminal individual, Medicare pays for hospice using a per diem structure based on four levels of care. In this payment structure, incentives may not reflect a person’s acuity or intensity of services since reimbursement is unrelated to actual service use. While hospice payment reform commenced in 2016 in hopes of better aligning incentives to match intensity of services, the reforms are modest at best.

Put succinctly, the uniform set of hospice services combined with its per diem reimbursement structure have done little to invite innovation in EOL care. These supply-side challenges, combined with the substantial demand-side variability in use suggest an inefficient allocation of hospice care in the current delivery environment.

The Inclusion Of Hospice Care In ACOs

Opportunities for innovation that could reduce variation in EOL experiences and hospice use have been somewhat limited as hospice care has been “carved out” of Medicare Advantage. This exclusion, however, does not exist in Medicare’s Pioneer Shared Savings Program (SSP) and Next Generation ACO models. The inclusion of hospice in those models means that hospice expenditures are included when calculating the total Medicare expenditure for each beneficiary assigned to the ACO. This inclusion should incentivize Medicare ACOs to think more critically about not only the initiation of hospice but what that care looks like.

Strategic Approaches To Serious Illnesses And EOL Care

Because of the role of hospice and palliative care in the value equation, strategic approaches to administering this care are likely to become increasingly important. Given the trends in EOL care and hospice use and the ACO framework, we see three likely strategic approaches for ACOs:

Selective Focus On Quality Hospice Providers

The inclusion of hospice care into the ACO accountability structure incentivizes ACO providers to refer beneficiaries to hospice organizations that practice in a way that is consistent with the ACO model’s mission to promote coordinated, high-quality, and economically efficient care. While ACOs currently cannot constrain patient choice, providers have an opportunity to influence this decision by becoming knowledgeable about the local market for hospice services in their areas. The siloed nature of the Medicare hospice benefit does not incentivize providers to acquire this knowledge, but ACOs’ inclusion of hospice services and the impending release of hospice quality reporting data by the Centers for Medicare and Medicaid Services incentivize and facilitate this awareness. If Medicare ACOs truly collaborate with hospice providers that perform demonstratively better in key areas of quality, the result will be two-fold. First, the EOL experiences for beneficiaries will improve as they will be receiving care from high-performing hospice providers. Second, poor-performing hospice providers will lose market share and consequently will be forced with the choice of improving or becoming irrelevant in their communities. The selective affiliations of ACOs with high-quality hospices, if pervasive enough, may also improve the hospice market for non-ACO beneficiaries since it would incentivize competition among hospices based on quality.

Expanding Access To Palliative Care

Any strategic approach to EOL care will raise the issue of how hospice fits into the care continuum and opportunities for the delivery of palliative care outside of the Medicare hospice benefit. Palliative care is specialized medical care for people living with a serious illness. It is provided by a team of highly trained professionals and is appropriate for individuals of any age and any stage of a serious illness and can be provided along with curative treatment. Because of the growing body of evidence that palliative care teams improve clinical outcomes while reducing unnecessary, unwanted, and potentially harmful medical care for seriously ill individuals, it should play an indispensable role in ACOs and in population health strategies more generally.

The business case for palliative care is becoming increasingly clear, with consistent evidence of palliative care associated with reductions in hospitalizations, emergency department visits, and costs. Creating a strategy for serious illness care and access to specialist-level palliative care for seriously ill individuals is contingent on ACOs and other risk-bearing entities seeing the value of palliative care services. Scaling palliative care, particularly in community settings, has been hampered by the inadequacy of reimbursement in a fee-for-service environment. It is not difficult to imagine ACOs being the vehicle by which palliative care becomes widely accessible. One can envision a more central role for palliative care as a bridge to hospice, with a gradual increase in palliative care intensity leading to a possible but not inevitable transition to hospice care.

The integration of palliative care and palliative care principles into the curative care plans for seriously ill individuals should be extremely attractive to ACOs because of the established effectiveness of palliative care across all “triple aim” domains. This includes direct evidence of reductions in readmissions, a Medicare ACO quality indicator. Deliberate approaches to expand access to palliative care outside of the Medicare hospice benefit would create an ideal continuum of care for seriously ill beneficiaries that would improve quality of life, diminish unwanted and harmful health care services, and likely result in more efficient use of hospice care.

Condition-Specific EOL Care Pathways

The variation in hospice use and lengths-of-stay is in part attributable to different trajectories of terminal illnesses. Unsurprisingly, longer lengths-of-stay are more common for conditions such as dementia and chronic obstructive pulmonary disease. Yet, despite the clear diversity in the EOL needs of beneficiaries with different conditions, the hospice benefit is stubbornly one size fits all. This is no doubt inefficient, but again, the exclusion of hospice care from most payment models does not create a market for innovation. However, the inclusion of hospice in ACOs may inspire an unpacking of the off-the-shelf hospice benefit, with hospice reimagined as a suite of services tailored to beneficiary needs.

The ACO Framework As Antidote To The Structural Barriers To EOL Care

The ACO model has the right incentives to reduce variation, spur innovation, and improve quality in EOL care. It also gives providers and delivery systems the latitude to expand access to palliative care outside of the Medicare hospice benefit by encouraging ACO providers to coordinate hospice care for their beneficiaries and shifting the care delivery emphasis away from what services will be reimbursed and toward what services will bring downstream value. Ultimately, this shift in perspective toward EOL care may reduce the variation we currently observe in who receives services and the intensity and duration of that care.

While the ACO model in its current form might motivate providers to address EOL care via the channels mentioned earlier, it is notable that the Medicare SSP, Pioneer, and Next Generation quality indicators do not specifically address EOL care. While filling this gap is a difficult task, it is critically important that quality measures germane to seriously ill and terminally ill individuals are identified and integrated into Medicare programs. An overarching concern about identifying the “right” mix of indicators is what they don’t incentivize. For example, the absence of EOL care-specific quality indicators could incentivize providers to focus on other areas that are included explicitly in the quality indicator set, thereby overlooking EOL care even though there are clearly identified areas for improvement. Unquestionably, integrating quality measures that are relevant to serious and terminal illnesses will help spur ACOs to think critically about how to tackle EOL care. In addition, ACOs must perceive the financial benefits from such an endeavor to be worth the effort involved in such a dramatic overhaul; existing concerns that the incentives are not lucrative enough to spur participation and delivery reform would also stifle new EOL approaches.


In short, in a market that is moving away from paying for care based solely on the quantity provided, the EOL care realm remains a largely unventured but promising area for ACO innovation. The variation in hospice use means that there are delivery inefficiencies that, if corrected, could generate both clinical and financial improvements, which are precisely the types of advances rewarded in the ACO reimbursement structure. Furthermore, hospice inclusion in Medicare ACOs finally provides a platform and the incentives to learn more about local hospice markets, rethink how palliative care should be used to improve use by seriously ill individuals and streamline the timely initiation of hospice care, and change the static hospice benefit to reflect the very different EOL needs of individuals with different serious or terminal conditions.

Individuals with a serious or terminal condition are a population with a demonstrated need for care coordination that is currently woefully unmet in the fee-for-service environment. While the ACO model has generated little momentum toward improving EOL care, the conditions are right for a more thoughtful integration of palliative and hospice services into the ACO care continuum.