Accountable care organizations (ACOs) are an increasingly dominant feature of the health care delivery system. As of the third quarter of 2019, nearly 1,000 ACOs covered approximately 44 million lives in the United States. Evidence regarding Medicare ACOs and their commercial counterparts indicates these models can achieve savings while meeting quality standards, although savings amounts vary by ACO characteristics (see previous work here and here).
Given their breadth and their incentives for care coordination, ACOs could be a powerful vehicle for improving palliative and serious illness care. This is because many of the most useful services for people with serious illness, such as 24/7 access to clinical support, home-based services, or caregiver education, are not typically reimbursed (or have small payments) under fee-for-service arrangements, whereas ACOs could receive shared savings for implementing these strategies. Indeed, some ACOs have shown success through exactly this approach. As a whole, though, early research suggested most ACOs were not focused on care management strategies targeting high-risk individuals (for example, those with multiple chronic conditions or high rates of utilization) but rather were focusing on utilization reduction across all patient groups, particularly for inpatient and postacute care.
Our more recent research found ACOs are widely implementing strategies to identify individuals with serious illness, but far fewer are implementing specific care strategies for the now-identified population. Many barriers prevented ACOs from launching serious illness programs, such as challenges accessing upfront capital for infrastructure and new workforce, difficulty making the business case to encourage buy-in from ACO leadership, and data limitations in accurately identifying the serious illness population.
In this post, we dig deeper into those results and examine how ACOs are prioritizing and implementing palliative and serious illness care strategies, breaking down the results by ACO type. In our prior research, we have seen that overall quality and savings results vary across ACO characteristics, so ACOs would likely vary in their serious illness approaches as well. This was born out in our case studies, where we saw dramatically different approaches to care for the seriously ill across different types of ACOs.
To better understand how different types of ACOs prioritize and implement palliative care beyond our previous case studies, we used data from the 2018 Annual ACO Survey conducted by Leavitt Partners and the National Association of ACOs. This survey has been used since 2017 to understand ACO trends. For this post, we examined how ACO priorities and serious illness strategies vary by ACO type—integrated (that is, ACOs built around an integrated delivery network), physician group-led, or hospital-led.
ACOs Prioritized ED Visits And Inpatient Admissions—An Opportunity To Make The Connection To Palliative Care?
The survey asked ACOs to identify their top priorities for improving efficiency and reducing total cost of care. As shown in exhibit 1, across ACO types the top priorities were: reducing avoidable emergency department (ED) visits and avoidable inpatient admissions; and preventing readmissions through better care transitions. Physician-led ACOs were particularly committed to these areas with more than three-quarters selecting avoidable ED visits and inpatient admissions and approximately 60 percent selecting prevention of readmissions as top priorities. Other frequently selected priorities included “active management of high-need high-cost patients” and, for hospital-led ACOs especially, “management/reduction of post-acute care spending and quality.”
Notably, only a small percentage of ACOs (10 percent) selected palliative care/hospice as a top priority. By ACO type, 13 percent of integrated, 3 percent of hospital-led, and 9 percent of physician-led ACOs identified palliative/hospice care as a priority for improving efficiency and reducing costs.
In our recent series of case studies on ACOs with programs targeting the seriously ill, ACOs with palliative care and serious illness programs often judged the success of their programs through impacts on ED and hospital use, along with broader patient outcomes. This finding highlights the opportunity to expand palliative care use by communicating to ACO leadership that such care affects other types of short-term use, which is important for success under value-based payment, and for patient experience and quality of life.
Exhibit 1: Selected top priorities by ACO type
Source: Authors’ analysis of data from 2018 Annual ACO Survey conducted by Leavitt Partners and the National Association of ACOs.
Different Types Of ACOs Have Different Palliative Care Capabilities
In the survey, ACOs identified whether they offered hospital- or community-based palliative care (see exhibit 2). Hospital-led and integrated ACOs reported similar rates of having widely or partially implemented hospital-based palliative care, but physician group-led ACOs reported much lower rates. The rates for ACOs that included hospitals (that is, hospital-led and integrated ACOs) are relatively consistent with other estimates of the prevalence of palliative care teams in hospitals. The lower rate from physician group-led ACOs isn’t surprising given that these types of ACOs generally do not contain a hospital inside their network, so implementing hospital-based palliative care would require partnerships or additional contracting. Integrated ACOs, which often have access to hospital facilities with palliative care teams but also have larger outpatient physician networks, had the highest rate of prioritizing palliative care (13 percent—see exhibit 1), and more than half (67 percent) reported offering hospital-based palliative care services.
The rates of offering community-based palliative care among all ACO types were lower than those for hospital-based palliative care, with 10 percent or fewer of ACOs reporting that they widely implemented such services. Again, physician-led ACOs were least likely to offer this type of care, although the gap was smaller than the gap observed for hospital-based palliative care.
Notably, most hospital- and community-based palliative care was reported as being only partially implemented, suggesting that ACOs are still in the early stages of establishing such programs. Our previous work indicates ACOs need an initial start-up period of three to four years to build infrastructure and become accustomed to taking on financial risk for their general beneficiary population. ACOs may not be able to add programs for the seriously ill until after this initial start-up period. As ACOs continue to mature, it will be important to track and examine their palliative care strategies to see whether these services become more widely available.
Exhibit 2: Hospital-based and community-based palliative care offerings by ACO type
Source: Authors’ analysis of data from 2018 Annual ACO Survey conducted by Leavitt Partners and the National Association of ACOs.
The survey also asked ACOs whether they were providing palliative care services directly, as opposed to through partnerships or contracting with non-ACO providers. Approximately half of ACOs said they provided some type of palliative care services in house, with responses ranging from 30 percent for physician group-led ACOs to 75 percent for hospital-led ACOs. Partnerships may be important for those ACOs that are not yet offering services. The Center to Advance Palliative Care recently identified nearly 900 community-based palliative care programs, indicating these programs are becoming more prevalent and may be potential partners for ACOs.
Successful Strategies Remain Unclear
The survey further asked ACOs about specific strategies they’ve implemented to manage their serious illness population. These strategies included the routine identification of seriously ill individuals, provision of 24/7 clinical response to these individuals, and provision of advance care planning services.
Across ACO types, physician group-led ACOs (72 percent) reported slightly higher rates of widespread implementation of routinely identifying seriously ill individuals than did hospital-led (59 percent) or integrated (55 percent) ACOs. It is unclear why physician group-led ACOs may be more inclined to identify the seriously ill given these ACOs reported the lowest rates of offering palliative care services (see exhibit 2). This finding may indicate physician-led ACOs are taking initial steps to implement palliative care by first identifying who within their patient population may benefit from such services. In contrast to identification, all ACOs were relatively similar in their limited use of 24/7 services and advance care planning.
These results show how palliative care fits into the current priorities of different types of ACOs. While palliative care is not currently a high priority, the findings highlight an opportunity to show ACO leaders how improved serious illness care can affect performance in other, higher-priority areas such as readmissions or utilization management. The results also illustrate that most ACOs are in the early stages of implementing various serious illness strategies.
The relative low priority placed on palliative care may be due to multiple factors. It may be that several years are needed for most organizations to build general accountable care infrastructure before they consider designing additional care programs—and many ACOs are still early in existence. Resource constraints may also underlie the variable levels of palliative care implementation across ACO types, but recent policy developments and programs may present new opportunities to further embed palliative and other types of serious illness care into ACO-delivered care.
For example, under the Pathways to Success Medicare Shared Savings Program (MSSP) rule, ACOs in that program have new options to expand their services outside of traditional health settings. Seriously ill individuals have functional limitations that can make traveling to clinic difficult, so offering services—including palliative care—through telehealth may be one avenue to reach these individuals. The new MSSP rule allows for the coverage of telehealth services for ACOs with prospectively assigned beneficiaries in non-rural areas. If these services are found to be effective, future regulatory actions could expand the coverage of telemedicine to additional beneficiary groups cared for by MSSP ACOs, such as those located in rural areas where access to palliative care can be limited.
The Primary Care First (PCF) model announced in April 2019 may also be leveraged in creative ways to address the needs of the seriously ill, and ACOs participating in the MSSP can also participate in this model. Under the PCF model, eligible practices will deliver advanced primary care services to medically complex beneficiaries. The Serious Illness Population (SIP) component of the model aims to proactively identify beneficiaries that would benefit from “transitional high touch, intensive intervention” to stabilize their health status. For this more intensive option within PCF, participating practices will receive enhanced first-visit fees and monthly professional population-based payments to account for the higher level of care delivered. Eligible ACOs may consider participating in, or partnering with practices participating in, PCF as a way to meet the needs of medically complex and seriously ill beneficiaries.
Under the Direct Contracting (DC) model, which had its Request for Applications released recently, the Centers for Medicare and Medicaid Services will contract directly with an accountable entity (or advanced ACO) to be responsible for the costs of care of Medicare beneficiaries, including those that are medically complex and seriously ill. In addition, DC entities will be eligible for a variety of payment rule waivers that may be helpful for better managing their serious ill population, such as the three-day skilled nursing facility rule waiver and the telehealth expansion waiver (explained in more detail in the linked request for applications). These features of the model, especially its high-need track, are meant to give the DC entity flexibility to address the complex needs of a seriously ill beneficiary. This may mean that the ACO partners with a third-party firm that offers serious illness care models, such as those provided by Landmark, Aspire, Turn-Key, and others. Under these models, a third-party firm contracts with private insurers to offer enhanced services to patients with high medical costs and needs. A multidisciplinary care team consisting of physicians, nurse practitioners, behavioral health specialists, social workers, and other providers offers a variety of in-home services including care evaluations, lab draws, IV medication administration, wound care, and minor procedures. This approach might be an option for ACOs participating in direct contracting to enhance team-based care services for the seriously ill.
These new policy tools are early in implementation and will be refined as results come in. They do provide new opportunities for ACOs to improve their palliative and serious illness care. However, payment models alone are not enough to drive adoption of palliative and serious illness care. Rather, the workforce, workflows, data infrastructure, and other factors must be in place. This is especially challenging for physician-led ACOs, which tend to be smaller and have less capital than other types of ACOs. Further spread will depend on tackling these organizational and infrastructure issues.
William K. Bleser, PhD, MSPH, received consulting fees from Merck on vaccine litigation research unrelated to this work. David Muhlestein is employed by Leavitt Partners, which consults about ACOs and works with providers and ACOs. Muhlestein is a visiting fellow at the Accountable Care Learning Collaborative. He has given speeches about ACO research to organizations, some of which provided travel expenses or small honoraria, and has attended multiple ACO conferences unrelated to this work. He has received grants from the Commonwealth Fund and the Moore Foundation for work on ACOs. Robert Saunders, PhD, has a consulting agreement with Yale-New Haven Health System for development of measures and development of quality measurement strategies for CMMI alternative payment models under Centers for Medicare and Medicaid Services contract No. 75FCMC18D0042/task order No. 75FCMC19F0003, “Quality Measure Development and Analytic Support,” Base Period. Support for this blog post was provided by the Gordon and Betty Moore Foundation.