On Thursday CMS announced its final rule governing the Medicare Shared Savings Program, updating its initial rules from November 2011. CMS highlighted several primary changes and additions to the program:
The creation of a Track 3, which includes higher rates of shared savings, the prospective assignment of beneficiaries, and the opportunity to use new care coordination tools.
Streamlining data sharing between CMS and ACOs, helping ACOs more easily access data on their patients in a secure way.
Refining the policies for resetting ACO benchmarks to help ensure that the program continues to provide strong incentives for ACOs to improve patient care and generate cost savings. CMS also said it intends to propose further improvements to the benchmarking methodology later this year.
Establishing a waiver of the 3-day stay Skilled Nursing Facility rule for beneficiaries that are prospectively assigned to ACOs under Track 3, which would allow payment for SNF services when a beneficiary is admitted to a SNF without a prior three-day inpatient stay.
The new Track 3 allows for a 75 percent rate of shared savings and 75 percent of all losses. In addition, once a beneficiary is assigned to a Track 3 ACO, the beneficiary will not be eligible for assignment to another ACO.
Also announced were final rules relating to Track 2 ACOs and symmetrical minimum shared savings and minimum loss rates. ACOs can choose no minimum shared savings/losses, symmetrical shared savings/losses between 0.5 percent and 2 percent, or shared savings/losses based upon the number of assigned beneficiaries.
Track 1 (one-sided risk ACOs) ACOs will no longer be forced to switch to Track 2 (two-sided risk) after the initial 3-year contract period. After a second 3-year period, they will be asked to advance to Track 2 or otherwise leave the program. In addition, CMS will allow previously dropped program participants to reapply.
There are 406 MSSP ACOs covering more than 7 million Medicare beneficiaries.
Our Take: CMS is taking great steps to ensure that the ACO model is successful. In a letter to previous CMS Administrator Marilyn Tavenner, signed by the American Medical Association and 2 dozen other associations, the physicians comments generally addressed issues of data sharing, patient attribution, risk and benchmarks. CMS adopted into the rules most of their requests.
In our research we’ve found that these four issues are ever-present and were needed to be addressed. One CEO of a large health system that dropped the ACO program expressed her frustration with using CMS data, telling us: “I'll tell you about [working with] CMS data. They changed their formats, they changed their fields, and they didn't give it to us when they said they were going to give it to us. It was very much of a problem. We challenged the data all the time.” Another CEO of a Pioneer told us that they left the program “because the benchmarks were flawed.”
If an ACO startup invests in infrastructure—data systems, quality measurement systems, additional personnel—the breakeven is about 4-5 years. Allowing Track 1 ACOs an additional 3-years under the protection of one-sided risk gives them time to have their investment in the program pay off.
But the federal agency must still address some issues with patient attribution. Once a beneficiary is assigned to an ACO, that person will not be eligible for assignment to a different ACO—even if the beneficiary chooses to receive several primary care services from another ACO during the relevant benchmark year.
It’s easy to see why this is a problem. It isn’t a stretch to imagine the senior who lives in Scottsdale, but spends their summers in the higher altitudes and cooler weather in Flagstaff. Plenty of Chicago seniors spend their winters in Florida. CMS will attribute the care to a Scottsdale or Chicago physician and not take into account care given in Flagstaff or Ft. Lauderdale—which may or may not be the same quality of care. CMS needs to consider patient attribution a priority for future adjustments to its rules.