It’s been quite the eventful ACO (accountable care organization) month for healthcare folks. First, on Aug. 9, the Centers for Medicare & Medicaid Services released a proposed regulation on the future of MSSP (Medicare Shared Savings Program) ACOs, with the broad intent to push providers into taking on more financial risk. Then, on the 24th, CMS released a detailed evaluation on how its Next Generation ACO model performed in 2016. Finally, on the 30th of the month, the federal agency made public the 2017 data for MSSP ACO participants.
On these three dates, three stories were told, but within those three stories are even more layers to what exactly is going on with federal ACO programs and just how much the government is comfortable with disclosing.
As I always like to ask, what’s the story behind the story? Let’s do some digging.
Backtracking a bit, last October, CMS made public the data from year one of its Next Generation ACO model, which debuted in January 2016. The results showed that 11 of the 18 ACOs in this model were able to achieve shared savings in year one of the program, which was quite an impressive feat considering the amount of financial risk that Next Generation participants take on. According to CMS, “The Next Generation ACO Model involves the highest levels of risk in any ACO initiative offered by CMS.”
Fast forwarding to August 2018, as noted above, CMS released a 111-page report on the Next Generation ACO model while also issuing a press release on the model’s 2016 results. But to the sharp-eyed observer, it’s strange that CMS again released 2016 data for the Next Gen program, rather than 2017 data. When asked, a CMS official noted that ACOs participating in the Next Generation ACO model have received their embargoed individual financial and quality results for the 2017 performance year, adding that ACOs get time to review their results and report any issues with their data to the agency. But still, the “re-release” of 2016 data is the key, since it feels as if CMS wants to highlight the Next Generation model over the MSSP model.
Enter the MSSP model, as CMS decided to not issue a press release on the program’s 2017 data, instead opting to release the data on the 472 MSSP ACOs in raw Excel format. Again, this raised my interest level, especially considering the massive changes the government is proposing for this program. While the 2017 performance data for MSSP ACOs was certainly encouraging—in sum, the 472 ACOs achieved $314 million in net savings to Medicare in 2017 after accounting for bonuses paid from the government—it’s quite possible, if not probable, that CMS chose not to make a big to-do about these results since its leadership has been exceedingly critical about MSSP ACOs of late.
Indeed, CMS Administrator Seema Verma has been saying for months that one-sided risk ACOs—which comprise about 82 percent of the MSSP model—are costing Medicare money and have not nearly been effective enough. In a press call announcing the August 9th proposed regulation from CMS that looks to reshape the MSSP structure, Verma bluntly stated that “[Upside-only] ACOs have no incentive, at all, to reduce healthcare costs while improving outcomes, as they were intended. Thus, the program has not lived up to the accountability part of their name,” she said.
As such, in the proposal, CMS is suggesting to shorten the glide path for new ACOs to assume financial risk, reducing time in a one-sided risk model from the current six years to two years. This proposal, coupled with CMS’s recommendations to cut potential shared savings in half—from 50 percent to 25 percent for one-sided risk ACOs—will certainly deter new entrants to the MSSP ACO program. It’s a proposal that has so far been met with mixed reaction as stakeholders weigh the broad intent of CMS to push providers into two-sided risk models.
However, as the 2017 MSSP data began to be analyzed late last week by industry experts, the consensus seems to be that one-sided risk ACOs in the MSSP model are doing better than Verma and others are giving them credit for. As Farzad Mostashari, M.D., former National Coordinator for Health IT and current CEO of ACO company Aledade, pointed out on Twitter, “Track 1 ACOs more than held their own [in 2017].”
So, the question that needs to be asked is, if MSSP ACOs—largely including one-sided risk ACOs—are performing as well as it seems they are, does CMS need to be so aggressive in its suggested overhauls to the program, with the agency’s primary logic being that one-sided risk models are not generating savings for Medicare? At the very least, it would be hard to argue that the timing is not fascinating—that just a few weeks after CMS released its proposed changes to the MSSP, the program’s results from 2017 disclosed terrific results.
What’s more, the fact that CMS did not formalize a press release around the MSSP data is also telling. One would be fair to ask, if the MSSP 2017 data did not shine one-sided risk ACOs in a positive light, would the government have taken the opportunity to formally publicize the undesirable results? Of course, these questions are merely hypothetical, but they are nonetheless important in the larger picture of ACO evaluation.
Regarding the question of how much one-sided risk ACOs are saving Medicare, on a recent Healthcare Informatics podcast, Rita Numerof, Ph.D., co-founder and president of St. Louis-based consulting firm Numerof & Associates, noted that one key factor that sometimes isn’t calculated in an ACO’s cost is the amount of administrative overhead that’s required to operate such a model.
Nevertheless, there are different outlooks on ACO savings depending on where one might look. This research in Health Affairs found that in 2016, Medicare saw a net loss of $39 million for the [MSSP] program, after accounting for bonus payments made to those ACOs earning shared savings. Specifically, the research found, CMS paid out more than $600 million in bonus payments to 134 ACOs in Track 1, while only collecting $9.3 million in penalty payments from four ACOs in Track 3. Meanwhile, the 22 ACOs participating in Tracks 2 and 3 generated approximately $33 million in net savings to CMS in 2016. “In other words, while the program as a whole is reducing per-patient Medicare spending, the amount paid out by CMS in total bonuses is more than the amount taken in by CMS from ACOs in two-sided risk that spent more than expected. Thus, according to CMS’s benchmark methodology, Medicare is not, on net, saving money,” the researchers from KPMG concluded.
Similarly, an analysis from consulting firm Avalere found that from 2013 to 2016, MSSP ACOs fell short of initial Congressional Budget Office (CBO) by more than $2 billion. But on the other hand, also as noted by the KPMG researchers, in Health Affairs, “Some have argued that ACOs are indeed saving Medicare money that cannot be fully captured measuring performance against each individual ACO’s financial benchmark—a more accurate assessment of impact would consider the ‘counterfactual’ of Medicare spending in the absence of ACOs.” To this end, the Medicare Payment Advisory Commission (MedPAC) estimated that ACOs may have saved Medicare from 1 percent to 2 percent more than reflected by their financial performance relative to benchmarks. In other words, the benchmark savings methodology may underestimate the amount of money that program participants are saving Medicare, the researchers noted.
In the end, it’s clear that there is a large disconnect as to just how much money one-sided risk ACOs are saving Medicare. That’s why the purpose of this blog is to request a bit more transparency from CMS, as the agency should release ACO program results in an equal opportunity fashion. If this were to happen, perhaps there would be far fewer questions—and much more clarity—on the ACO landscape as providers and payers continue to make critical decisions as they move toward a value-based care future.
