Reflections on 10 Years as a Long-Term Care ACO

When it launched in 2016, Long-Term Care ACO was the first ACO to focus on the population of Medicare beneficiaries who reside in long-term care nursing facilities, but now several other ACOs have entered that space. At the National Association of ACOs (NAACOS) fall meeting, Kristen Krzyzewski, chief strategy and program development officer for LTC ACO, highlighted some of the challenges and successes in navigating the Medicare Shared Savings Program with a long-term care focus. 

Krzyzewski noted that after LTC ACO entered the MSSP, it was the only ACO focused on this population for six years. “Since proving that there is significant opportunity, there are now six ACOs that have long-term care populations as greater than 60% of their total assigned lives, and it continues to expand. In 2024 these six ACOs generated over $300 million in savings,” she said. “There is a lot of opportunity, and we're happy to be a part of it.”

LTC ACO has been in the enhanced track of MSSP for a second agreement period, and has entered into into a third agreement period starting this year. “We also expanded in 2025 to include the assisted living facility population. That was something new for us,” she said. “A lot of our providers in this space were serving both populations, so it was important to expand. With that expansion we are one of the ACOs that have generated some of the highest savings on a per-beneficiary basis.”

She provided a quick snapshot of how her ACO’s population differs from a typical MSSP ACO. About 98% of its population is in a long-term care, institutional setting, and that compares to less than 1% in a typical ACO. Their percentage of dual eligibles is significantly higher. “Over 83% of our population is dually eligible, and for the typical ACO I think it's a little less than 7%, so there is a big variance there,” she added. “Our population is about 81 years old, on average, and the portion of our population over age 85 is about 34% and for a standard MSSP ACO, that's less than 11%.”

Also, a significant portion of LTC ACO’s population passes away each year. The death rate is over 22% vs. less than 3% in the standard MSSP ACO. “That creates different challenges, and you have a shorter window to work with the population to improve the cost and quality, and it represents unique challenges for us,” she explained. 

Areas of focus

Krzyzewski explained how these patient demographics impact the ACO’s areas of focus. “In care delivery we're really focused on preventing unplanned hospitalizations. We don't want our patients to have to go to the hospital unless that's the absolute right place for them to be. But in the nursing home setting, if the primary care isn't available and the facility staff members have a question or concern, they're going to send the patients out,” she said. “Weekends and after hours, the patient's going to go to 911, to the hospital, and that will typically result in a hospitalization and create a lot of risk. We see that as an adverse quality outcome, so our participating providers are focused on preventing those hospitalizations, and now that they're accountable for the total cost of care, they're really engaged in trying to minimize that risk. We work with predictive analytics models to identify who's at greatest risk for going to the hospital for an admission, and we arm our providers with that information so they know within a high-risk population, who's the highest, highest risk.”

LTC ACO providers are offering intensified primary care to those folks who are at greatest risk. “We’re also focused on transitions of care. While we try to prevent hospitalizations, we can't prevent them all,” Krzyzewski said. “They do go to the hospital, and when they return to their home in the nursing facility, they're at greater risk, like all of our populations, for readmission. So there's a ramp-up in intensity of the primary care delivered.”

Another interesting component is the complexity of pharmacy and Part D and the utilization of drugs. “We have initiatives to to de-prescribe, and that's important to help folks remain stable in the facility,” she said. “We also are really focused on mortality risk. It is an end-of-life population. Who's at greatest risk for short-term death? We are very focused on advanced care planning, working with the families and the beneficiaries to understand what their goals of care are, with our primary care clinicians documenting that so that those wishes are honored at the time that there is stress and decisions have to be made about whether they should go to the hospital or not.”

Also, in this population, hospice is a key driver of costs, and LTC ACO providers are sometimes surprised to see that some of their beneficiaries have been on hospice for two years. There are some incentives at the facility for hospice utilization, because it helps staffing, Krzyzewski  noted. “So we're constantly looking at what's the appropriate use of hospice care, and can palliative care be offered to provide some of the same benefits?” 
LTC ACO also faces challenges in attribution and satisfying the physician visit. “Even though we have a lot of nurse practitioners who are the primary clinicians, they can’t gain attribution unless there's a physician visit, Krzyzewski explained. “So there are a lot of gymnastics that we have to do to make sure that our providers can get that physician visit simply to satisfy the attribution criteria. That really doesn't make sense. ACO REACH doesn't have that. We have talked about how we would like that to change within MSSP, certainly for our population.”

Mismatched quality measures

Krzyzewski noted that although LTC ACO is committed to quality, by the nature of its population and the quality measures used in the program design, they are at a disadvantage. She also said CMS is going in the wrong direction with the Universal Foundation set of measures because it includes measures that aren't relevant for their population. She used breast cancer screening as an example. “If these measures are going to be increasingly oriented toward a community population that isn't at end of life, you need to have a bonus of some sort if you're treating a population at end of life to add points back so that we can can be measured sort of equally with our peers that are focused on the community population.”