Tying Social Determinants of ACO Patients With High-Need, High-Cost Care

Accountable care organizations (ACOs) often care for patients with complex, chronic conditions that can lead to high expenditures and utilization of care. During a session at the National Association of ACOs Fall 2018 conference, being held October 3-5 in Washington, DC, panelists discussed how ACOs can design and implement strategies that deliver high-quality, low-cost care for these patients.

Accountable care organizations (ACOs) often care for many patients with complex, chronic conditions that can lead to high expenditures and utilization of care. In the United States, 1% of the population accounts for 21% of healthcare expenditures and the costliest 5% account for more than half of all health spending, according to a report published by the Agency for Healthcare Research and Quality.

During a session at the National Association of ACOs (NAACOS) Fall 2018 conference, being held October 3-5 in Washington, DC, panelists discussed how ACOs can design and implement strategies that deliver high-quality, low-cost care for these patients.

Led by Rob Mechanic, MBA, executive director, Institute for Accountable Care, participants in this conversation included Melinda Abrams, vice president and director, Health Care Delivery System Reform program, The Commonwealth Fund; Sanjay Doddamani, MD, FACC, FACP, CMO, Geisinger at Home, Geisinger Health System, and CMO, Keystone ACO; Amy Russell, MD, chief of community medicine, Mission Health Partners (MHP); and Wilson Gabbard, HCS director, population health operations, University of North Carolina Health Care.

Authors of a recently published commentary in The American Journal of Managed Care® argue that data-sharing networks that merge a patient’s clinical and genomic data with social determinants of health information can improve both clinical and financial outcomes.1 While data sharing is vital to this process, especially for accountable care partnerships, the authors advise caution with regard to patient privacy and data security.

Additional coverage from the NAACOS meeting.

Mechanic highlighted characteristics of some of these high-need patients who may not be physically independent and may be socially isolated. The Institute, he said, conducted a survey that showed that ACOs care a lot for these patients, not just because they are high utilizers. Of the 135 ACOs that participated, 85% found they were successfully targeting high-cost individuals.

“Per the survey, while ACOs are being innovative with managing high-need patients, they are not sure they are reaching them at the right time,” Mechanic added.

The bottom line is that ACOs think they are having a positive impact on high-need, high-cost (HNHC) patients, but they are uncertain about what’s working, he said.

“The system is failing HNHC patients and their needs are not being appropriately met,” Abrams said. She added that The Commonwealth Fund pivoted focus to HNHC patients because of the growth in value-based payments; this steered greater attention to high-cost patients and provided an opportunity for greater flexibility within payment models including traditional Medicare or Medicare Advantage.

Doddamani, a cardiovascular physician by training, said that he realized that some of their sickest patients aren’t at the hospitals but are in the community. “We identified 3 well-defined buckets: medically complex, elderly, and high-need patients. For each of them we realized that the workforce is tied to what the patients want. However, for some of the strategies, such as home-based medical care, clinical and non-clinical care providers can work together.”

He then described how their ACO has developed alliances with home-based care providers and home-based palliative care providers to reach these HNHC patients.

“We have tried a variety of tailored payment models,” Russell said. “We also have a care management team that focus on things like transportation for our patients,” which she said providers may not understand or always think about. Additionally, their inpatient and outpatient nurse practitioners are involved in transition programs that follow patients all the way home.

“How are you identifying the right people for the right funding,” Mechanic asked.

Abrams pointed out that while it would seem obvious that models that work are ones that are targeted for the patients they serve, “the effect is diluted when the model is applied to the general population.” She said that they tried to identify subgroups within the high-need group based on the heterogeneity that exists: nonelderly adults with disability, frail elderly, homeless with diabetes, seriously ill, and others.

“These subgroups were isolated based on both cost and clinical needs and we opted to go for patients with high clinical needs. These were derived from claims data, so the caveat is that social determinants of health [SDOH] are not a part of these data.”

Identifying the right patients is vital, according to Doddamani. “It’s intensive work, but we need to step back and focus on a smaller population.”

The model that they created evaluates 6 or more serious conditions that are associated with catastrophic downstream events, with considerations for a home hospice where patients mainly need symptom support. Doddamani asked the audience to imagine an S-shaped curve, with the highest-risk patients at the top, defined as the most likely to die in the next 6 months; then the mid-level risk patients, who can be provided pharmacotherapy and care management; followed by the low-risk patients at the bottom of the curve who can be provided social support and care management and can be stably maintained.

Mechanic asked the panelists to describe each of their approaches to SDOH. “Have you found a good way to collect data around it?”

“We track gaps in the SDOH space, such as transportation issues, to intervene there,” Russell said, and then they evaluate the impact it has on patient outcomes. “That to me is one of the ways to support the high-risk populations, but it can help focus on specific issues and support these patients better.”

“Leveraging community partners is vital,” Doddamani said. “We often find that our community partners do not have access to patient information, so providing them that support is appreciated by these community resources as it is vital to their viability.”

When asked about the most important thing that their organization is doing right now, Russell said that it’s still in early stages for their ACO and that they don’t have data to support what’s been most impactful. She did point out, however, that community healthcare providers have been most impactful in connecting providers with patients.

Wilson also said that they lack data at this point but believes that some of the additional services that they provide patients—nurse call line, navigating urgent care to reduce out-of-pocket expenses, engaging with them to ensure they have medications for the weekend, and those with transportation issues have transportation—are making an impact on outcomes.

Abrams shared their work with the Bipartisan Policy Center, which she said is helping them understand what policy or supplemental changes are needed for HNHC patients, although the current focus is on Medicare Advantage patients. “The results could be extrapolated for traditional Medicare,” she said.

She asked the attendees to identify and share the barriers with The Commonwealth Fund “so we can document them and share with the Bipartisan Policy Center to change policy and move the dial on some of the existing roadblocks.”


Khullar D, Kaushal R. “Precision health” for high-need, high-cost patients. Am J Manag Care. 2018;24(9):396-398.

Related Videos
dr mohamad mohty
dr jennifer vidrine
dr ryan jacobs
dr brian koffman
John Hood, PhD
drs almeida, landgren, and lee
Melissa Jones, MD on Artificial Intelligence and Sleep Studies
SLEEP 2024 Recap Video
dr surya bhatt
Related Content
CH LogoCenter for Biosimilars Logo