Industry report: ACOs saved twice as much as CMS estimates

Accountable care organizations generated $1.84 billion in savings for Medicare in 2013–2015, nearly twice as much as the $954 million estimated by CMS, according to a new report from the National Association of ACOs (NAACOS).

After accounting for shared savings payments, ACOs in the Medicare Shared Savings Program (MSSP) reduced federal spending by $542 million, in stark contrast with CMS’s estimates that the organizations actually increased Medicare spending by $344 million over the three-year period.

The large discrepancy in savings is due to the methods that CMS and NAACOS used to calculate their figures. And the new report underscores that understating the savings can have policy impacts, as CMS has pointed to the results to make changes to the program.

“The analysis should put to rest claims that shared savings only ACOs do not save Medicare money,” Clif Gaus, ScD, president and CEO of NAACOS, said in a statement. “The findings confirm the wisdom of giving ACOs adequate time to build the care coordination, information technology, and data analytics capabilities needed to manage financial risk successfully.”

CMS uses benchmarking methodology to calculate savings, where actual spending is compared with targets based on an ACO’s historical spending. These estimates have been the basis for mounting criticism from the agency’s leaders over the ability of ACOs to improve quality of care while saving Medicare money, particularly as the vast majority of ACOs do not take on downside financial risk in the program.

According to researchers, the benchmarking method “systematically understates the actual savings” generated by ACOs in the shared savings program.

“The CMS benchmarking methodology addresses the question ‘How has ACO spending changed compared to prior years’ spending?’” the study reads. “While this may be an appropriate way to set performance benchmarks, it produces a biased estimate of program savings when compared to what may have occurred if the ACO program had not been in place.”

The savings calculations have led CMS to propose changes to the MSSP that would require ACOs to take on downside risk after only two years. However, using the same benchmarking methodology, CMS recently reported that ACOs saved Medicare $314 million in 2017.

NAACOS commissioned Dobson | DaVanzo & Associates to conduct an independent study of MSSP ACO cost savings. The study used a difference-in-differences regression analysis to evaluate the program and found that the actual savings were on par with other independent studies–and much higher than what CMS found.

This evaluation of the program asks the question, “How have ACOs changed expenditures compared to providers not participating in the ACO program?”

The MSSP is the largest value-based model in CMS, with 561 ACOs covering 10.5 million Medicare beneficiaries in 2018. The newly proposed rule would change the current regulations that allow eligible ACOs to participate in shared savings with Medicare for up to six years without taking on downside risk.

The change to two years could result in a mass exodus of some ACOs out of the voluntary program, as the organizations tend to do better with savings over time, likely more than two years. The proposal also lowers shared savings from 50 percent to 25 percent.

“It takes time and money to transform entrenched care delivery practices in local communities and build the critical mass to successfully integrate care, manage risk, and improve quality while reducing spending growth,” Stephen Nuckolls, CEO of Coastal Carolina Quality Care in New Bern, North Carolina, which includes 63 providers caring for 11,000 Medicare beneficiaries, said in a statement. “Unfortunately, the proposed changes will hold up the move to value-based care by significantly undermining the business case to voluntarily form new Medicare ACOs.”  

To that end, HHS Secretary Alex Azar maintained the move would be worth it.

“If this ends up meaning somewhat fewer ACOs, so be it,” Azar said in a recent speech.