Assuming Risk Key to Medicare Savings Under New Payment Models

Accountable care organizations (ACOs) only generate savings for the government if they accept risk, according to an analysis from the Kaiser Family Foundation.

ACOs generated net savings of $47 million in 2016, the Menlo Park, Calif.-based non-profit found in an independent analysis of publicly available ACO data and previously unpublished numbers from the Centers for Medicare & Medicaid Services (CMS). But those savings weren’t uniform among the organizations, which seek to drive down overall costs of care by fostering collaboration between a variety of providers — including skilled nursing providers.

For instance, ACOs that were simply rewarded with bonuses for performance ended up costing Medicare $72 million. Organizations that agreed to assume “downside risk,” or a portion of the shared losses, uniformly saved the government money, with the Medicare Shared Savings Program Track 2 and Track 3 turning in savings of $18 million and $14 million, respectively.

Advertisement

The Pioneer model generated $24 million in savings, while the Next Generation structure resulted in $63 million in net savings for the government.

The news isn’t all good, however.

“Despite the financial incentives for savings in ACOs, there have been no notable differences in quality between traditional Medicare and ACOs,” Kaiser noted in its most recent analysis of the findings.

Advertisement

The ACO data forms part of Kaiser’s new fact sheet about the Center for Medicare and Medicaid Innovation (CMMI), a CMS initiative that looks to find novel ways to cut costs while still maintaining high care quality.

Among the takeaways, the non-profit noted that the CMMI’s funding and goals have not shifted significantly under the Trump administration —  though Kaiser did acknowledge that the agency has made changes to certain models or cancelled them entirely.

Check out the full fact sheet at the Kaiser Family Foundation.

Written by Alex Spanko

Companies featured in this article:

, , , , ,