One prominent researcher said recent findings of steep Medicare losses from ACOs run counter to his published analyses.


April 10—Whether accountable care organizations (ACOs) save money for Medicare has emerged as a key question because it could determine how hard regulators push participating providers to take on downside risk.

A recent Avalere analysis garnered widespread attention after concluding that Medicare ACO net savings have fallen short of initial projections by more than $2 billion.

Specifically, the Congressional Budget Office (CBO) projected that the Medicare Shared Savings Program (MSSP) would produce $1.7 billion in net savings for the federal government from 2013 to 2016. However, over that period the MSSP increased federal spending by $384 million, which was a difference of more than $2 billion, Avalere concluded. (The research was funded by the Pharmaceutical Research and Manufacturers of America [PhRMA].)

But ACO leaders have pushed back on the significance of that conclusion.

Joe Damore, vice president of population health management for Premier, said ACO savings were “optimistically modeled” in 2010 in a way that underestimated the complexity of the care delivery shift and the investments required to transform entire systems of care.

“At the same time, models ignored the difficult reality of operating simultaneously in both a fee-for-service and value-based environment,” Damore said.

He noted Avalere also found that MSSP ACOs produced $1.6 billion in program savings compared to benchmark projections over the life of the program, and those savings increased each year.

Allison Brennan, vice president of policy for the National Association of ACOs (NAACOS), viewed the Avalere analysis as simplistic.

“It’s important to note that CMS [Centers for Medicare & Medicaid Services] benchmarks do not represent the true savings of the program,” Brennan said in an email statement. “The evaluation of ACOs is complex, and oftentimes the desire for an easily produced answer (derived by just looking at benchmarks) unfortunately gets more attention than a more thorough and accurate answer which requires more rigorous analysis.”

J. Michael McWilliams, MD, PhD, a professor of healthcare policy at Harvard Medical School who has studied ACO finances, also criticized the accuracy of the Avalere savings analysis.

“While the savings have been quite modest to date, they do contrast pretty starkly with these calculations that various analysts do with the CMS benchmarks, which are just not useful,” McWilliams said in an interview.

McWilliams’ Medicare ACO analyses include a 2016 examination of 2013 and 2014 results, which found that the MSSP produced net savings of $685 million, or 1.6 percent, in 2014. (That research was funded by the Laura and John Arnold Foundation and the National Institute on Aging, part of of the National Institutes of Health.)

Analyses by Avalere and others don’t pick up on the fact that spending reductions by ACOs indirectly affect Medicare spending, nor do they account for annually lowered benchmarks that reduce spending or for spillover reductions in Medicare Advantage spending, McWilliams said.

As of 2014, “There was evidence of net savings to Medicare and evidence of growth in the savings,” McWilliams said. “Assuming those effects continue to grow, the savings may be even larger in the future because the savings have been quite modest to date.”

Significance of Savings

Conclusions about whether Medicare has obtained savings from MSSP ACOs are critical because CMS is weighing whether to tighten or loosen downside-risk requirements to garner more savings. Although a growing number and share of MSSP ACOs are moving from one-sided to two-sided risk, 82 percent remain in the one-sided, bonus-only model, according to CMS figures. The share was 91 percent in 2017.

“There’s a big debate about how quickly to move ACOs to downside risk,” McWilliams said.

He has urged caution, as justified by his and others’ findings of overall Medicare savings from even ACOs in one-sided risk models.

“The impetus to transition into downside risk may not be that strong if ACOs in one-sided contracts without the downside financial risk are indeed saving Medicare money, which they are according to our evaluations and other research teams,” McWilliams said.

Further, the push to assume more downside risk may not greatly increase Medicare savings, as evidenced by the experience of two-sided Pioneer ACOs. The savings from those ACOs was “quite minimal,” McWilliams said, and less than he found in the MSSP, while the Pioneer program also “suffered massive participation losses.”

ACO advocates have long warned that a broad push of one-sided MSSP ACOs into two-sided risk could lead many to drop out of the program. That concern was borne out by a NAACOS survey of ACOs that found 43 percent would leave the MSSP if CMS required downside risk.

“You can’t have savings from ACOs if you don’t have any ACOs,” McWilliams said.

He also worried that a rush to downside risk will accelerate provider consolidation. That concern stems from the widely held view that larger organizations are better able to absorb potential downside-risk losses and afford ACO setup costs, which average $2 million, according to the NAACOS survey.

Changes Needed

If CMS wants to strengthen incentives for providers in ACOs to find savings, it could use other approaches that would not drive out participants.

For instance, Medicare could increase the shared savings rate in one-sided MSSP ACOs from 50 percent to 100 percent, which would increase Medicare savings through “spillover savings” and the resulting lowering of benchmarks, McWilliams said.

He viewed the 2018 launch of MSSP Track 1+ as a positive step. That model offers the lowest amount of risk of any two-sided Medicare ACO model and qualifies a participating organization’s physicians for 5 percent Medicare payment bonuses through the advanced alternative payment model track of the Medicare Access and CHIP Reauthorization Act.

Fifty-five ACOs were in the inaugural class of Track 1+, and it remains to be seen whether the program can keep ACOs in the MSSP, McWilliams said.

ACO advocates also have urged changes, as well as patience.

“We also can’t forget that it’s a steep learning curve for ACOs—clinical and operational transformations of this magnitude take time,” Brennan said.

Among policy changes that NAACOS has urged of CMS are modifications to ACO benchmarking, such as allowing risk scores to increase and not lowering future benchmarks when an ACO has savings.

“The latter keeps benchmarks lower, which make it harder for ACOs to earn savings,” Brennan said.


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

Publication Date: Thursday, April 12, 2018