An administration executive said if MSSP ACO providers decide to leave the program, they might be interested in moving into either direct-provider contracting or an expanded bundle program.


Sept. 11—A new analysis joins other research in reaching new conclusions about the extent of savings that accountable care organizations (ACOs)—especially upside-only models—provide Medicare. It’s a data battle that could determine whether many ACOs continue.

The National Association of ACOs (NAACOS) commissioned the firm of Dobson|Davanzo to conduct a statistical analysis of Medicare Shared Savings Program (MSSP) ACO savings from 2013 through 2015. Those ACOs—the vast majority of all Medicare ACOs—provided more than $1.8 billion in gross savings, or nearly double the $954 million savings previously calculated by the Centers for Medicare & Medicaid Services (CMS).

The Dobson|Davanzo study used difference-in-differences regression analysis to compare spending on Medicare beneficiaries inside and outside of ACOs.

The analysis found $542 million in net Medicare savings after shared savings payments were distributed. In contrast, CMS’s analyses—based on a comparison to ACOs’ benchmarks—concluded that MSSP ACOs cost Medicare a net $344 million during those three years. Similarly, a March analysis by Avalere concluded that by 2018, ACO net savings had fallen short of initial Congressional Budget Office projections by more than $2 billion.

“The [Dobson] analysis should put to rest claims that shared-savings-only ACOs do not save Medicare money,” Clif Gaus, president and CEO of NAACOS, said in a release.

Christopher Stanley, MD, a director at Navigant, agreed that the difference-in-difference approach is a more accurate method than CMS used to tally the effect of MSSP ACOs on medical costs.

“The results are compelling—and do demonstrate significant cost savings at the CMS program level,” Stanley said.  

The analysis also was in line with other research, such as a recent study in the New England Journal of Medicine (NEJM), which found ACOs provided $145 million in net Medicare savings in 2015. That study—funded by the National Institute on Aging—also used difference-in-differences analyses to compare changes in Medicare spending for patients in ACOs before and after a participating provider’s entry into the MSSP with concurrent changes in spending for local patients served by nonparticipating providers.

J. Michael McWilliams, MD, PhD, a professor at Harvard Medical School and an author of the NEJM study, said “the debate over the MSSP has unfortunately been derailed by misguided use of the benchmarks as valid counterfactuals for program evaluation. People really have to stop using the benchmarks to assess program performance.”

The latest analysis also followed the August release of the latest raw data on the existing 472 MSSP ACOs. Early analyses of that data found that between 2016 and 2017, those entities went from costing Medicare nearly $40 million to saving a net of $314 million, out of $95 billion in Medicare spending—according to the measurement method preferred by CMS.

“The savings are modest as a proportion of total spending, but as we demonstrate, they have been growing over time and are quite meaningful among the physician group ACOs,” McWilliams said in an interview. “There are no home runs in health care. This is what progress looks like.”

Another ACO researcher, John Hsu, MD, an associate professor in the Department of Medicine and the Department of Health Care Policy at Harvard Medical School, said the inherent challenges in measuring ACO savings indicate the need for ongoing studies.

“Ideally, we compare the observed findings to those that would have occurred in the absence of the MSSP program, but this is very difficult to do well, including with a DD design,” Hsu said in an interview, referring to difference-in-differences analysis. “So, the more evidence that we have, and the greater consistency of findings across studies from a range of investigative groups and methods, the better.”

Overhaul Looms

All of this data comes as CMS considers a major overhaul of the MSSP, which would require the 80 percent of participating ACOs that operate in upside-only Track 1 to move to downside risk within two years. Track 1 ACOs have a total of six years to take on downside risk under current rules.

Trump administration leaders have said such a change is needed because ACOs exist to move providers toward savings and quality improvements, and that requires risk.

“From my personal experience as a risk-based group, you need accountability,” Adam Boehler, director of the Center for Medicare & Medicaid Innovation (CMMI), said at recent media briefing. “If everything is an upside option in value, you don’t really invest.”

ACO advocates warn that the CMS proposal would reduce the number of ACOs participating in the MSSP and also slow the growth of new entrants. CMS appeared to concur when the proposed rule estimated that 109 ACOs would drop out.

“If this ends up meaning somewhat fewer ACOs, so be it,” Alex Azar, secretary of the U.S. Department of Health and Human Services (HHS), said Sept. 7 at an American Heart Association event.

It’s not clear that the latest data will end up changing the administration’s mind, but it may bolster the view of other policymakers that the MSSP overhaul at least needs softening.

For instance, leaders of the House Ways and Means Committee wrote CMS on Sept. 4 to urge the agency to reconsider its plan to reduce the share of savings that ACOs could garner from 50 percent to 25 percent (ACO advocates likewise support retaining the higher figure). A reduction in potential savings “may also hinder an ACO from being financially able to take higher levels of risk,” the letter stated.

If the proposed rule is not altered, Stanley expects poorly performing ACOs to exit the program and focus on volume-driven delivery of health care while high performers would move to downside risk.

“We do not believe that this new study will influence any ACO’s decision to stay in or exit the program—their own results and likelihood of future success will determine their strategy,” Stanley said.

Post-ACO Options

If MSSP ACOs decide to leave the program, Boehler suggested they might be interested in moving into either direct-provider contracting or an expanded bundle program. Hospitals in ACOs, specifically, may be better fits for bundled payment programs than for ACOs, he said, because ACOs track patient care longitudinally over several years, while bundles carry short-term, episodic risk.

“So you may have some hospitals that say, ‘Listen, do I really have a relationship over a period of years? Or do I say my relationship with the patient is when they come into the hospital and 90 days thereafter?’” Boehler said. “That is a very good avenue for somebody to get into the risk game too, where they are matching where they want to take risk with their relationship.”

Medicare is accepting applications from providers that want to enroll in the voluntary Bundled Payments for Care Improvement Advanced program. CMS will re-open it to new provider enrollees in 2020.

Also this summer, CMS announced it was putting together a direct primary care model, based on some of the 1,000 stakeholder responses it received after asking about potential new directions for payment models.

“The risk of assuming two-sided risk is important, but also critical would be the payment options outside of MSSPs or other APMs,” Hsu said. “In other words, all future paths involve some level of organizational risk. Knowing the potential risks associated with one path alone is insufficient to predict future behavior.”


Rich Daly is a senior writer/editor in HFMA’s Washington, D.C., office. Follow Rich on Twitter: @rdalyhealthcare

Publication Date: Wednesday, September 12, 2018