I’m never a fan of when people have news to tell me and they ask, “Do you want to hear the good news or the bad news first?” I feel that this question puts preconceived notions in my head and if anything, it ruins the good news knowing that bad news is accompanied with it. I’d rather you just tell me what the news is and let me decide if it’s good or bad!
So that’s exactly what I am going to do here. In the past few years, when the Centers for Medicare & Medicaid Services (CMS) has released the data from the previous year’s Medicare ACO (accountable care organization) programs, it was always tricky to figure out what story the information was really telling.
On one hand, if you were to read just the CMS announcements on their ACO results, you would think that the programs are performing flawlessly. Here’s an example from 2016, in which the agency’s announcement stated, “According to the results, over 400 Medicare ACOs generated more than $466 million in total program savings in 2015, accounting for all ACOs’ experiences. Of these, 125 qualified for shared savings payments by meeting quality performance standards and their savings threshold.”
On the other hand, a deeper dive into the data tells a bit of a different story. Using the same year, 2016, Healthcare Informatics reported at the time, “Although more than 400 Medicare ACOs generated more than $466 million in total program savings in 2015, nearly seven in 10 of those ACO organizations (279 out of 404) did not generate enough savings to receive bonuses. So while the total savings of Medicare ACOs did total $466 million [in 2015], it should be noted that many ACOs are not generating savings at all, or are generating some savings but not enough.
As you can see, it’s clear how the same set of numbers can tell two different narratives. The CMS announcement touted the total savings, but did not highlight that the savings were actually only coming from 30 percent of those ACOs—meaning the remaining 70 percent were either not generating enough shared savings to qualify for earnings or were losing money outright. As our Editor-in-Chief Mark Hagland stated in an analysis at the time, it would be helpful for the public if CMS officials were more straightforward about ACOs’ level of progress.
This past fall, CMS this time didn’t release the ACO data with much elaboration at all. The agency issued a news release for the year one results of its Next Generation ACO program, which we covered in October, but for its Medicare Shared Savings Program (MSSP) participants, there was no announcement; data for these ACOs could be viewed online in CMS’ Public Use File.
The 2016 Next Gen ACO data revealed that 61 percent of program participants (11) were able to earn shared savings while the remaining seven ACOs in this model generated losses outside a minimum loss rate and thus owed shared losses. Adding up the 11 ACOs which were able to generate savings ($71 million) and subtracting from the seven ACOs which owed losses ($23 million), the net of all gross savings and losses is about $48.3 million in savings, per the CMS data. What’s more, all 18 ACOs in this model scored 100 percent on quality across 33 measures they were graded on.
As for the MSSP data, as detailed in an analysis from the Advisory Board, an estimated 134—or about 31 percent—of the 432 ACOs that participated in MSSP for the program's fifth performance year generated shared savings in 2016. What’s more, 294 of the 432 MSSP ACOs generated neither savings nor losses in 2016, while four generated losses. In sum, the gross savings for this program in 2016 amounted to about $652 billion. And the data also revealed that about 77 percent—or 330—of the 428 MSSP ACOs subject to pay-for-performance measures for the 2016 reporting period achieved an average quality score of 94 percent. The remaining 98 MSSP ACOs, which were in pay-for-reporting status, earned quality scores of 100, according to the analysis.
Comparing 2016 Medicare ACO results with 2015 results, the one main takeaway that sticks out to me is how Next Gen ACO participants performed in the first year of the program. As a refresher, Next Gen participants have the opportunity to take on higher levels of financial risk—up to 100 percent risk—than ACOs in other current initiatives. While they are at greater financial risk, they also have a greater opportunity to share in more of the model’s savings through better care coordination and care management.
Bearing in mind that this model includes two-sided risk, I for one would have expected worse results in just the first year of the program. Consider this: in 2015, MSSP ACOs, which take on less risk overall, had a 30 percent shared savings rate compared with 61 percent in the Next Gen model. While the number of MSSP ACOs (over 400) greatly exceeds the number of Next Gen ACOs (18), this finding was still quite noteworthy to me.
So the question now becomes, what does all this mean for the future of ACOs and what could we conclude about their progress so far? I recently spoke with ACO leaders at UnityPoint Accountable Care (UAC), a subsidiary of Des Moines, Iowa-based UnityPoint Health, and they noted that having a history of working within innovative value-based arrangements through the Pioneer ACO program and through MSSP Track 1 has positioned them well to have success in the Next Gen program. As it turns out, success in the Next Gen model came quickly for UnityPoint Health, as the ACO was able to generate $10.5 million in shared savings in the first year, which ranked third out of the 18 ACOs that participated.
UAC leaders also spoke to the importance of being data-driven and having the right IT tools in place. They specifically told me about an impressive solution the organization has which creates a predictive model and a “heat map” that predicts the risk of a patient being readmitted. The solution actually drills down to the exact day that the patient is predicted to be readmitted within the 30-day window so that UAC providers have a better idea of what the intervention should look like.
Beyond UnityPoint Health, more ACOs have similar stories of success to share. In just the past month alone, Healthcare Informatics has reported on the following ACOs and how they have been able to flourish in their given programs: Downers Grove, Ill.-based Advocate Health Care; Innovation Care Partners (ICP) in Scottsdale, Ariz.; and Triad HealthCare Network, a subsidiary of Greensboro, N.C.-based health system Cone Health. These stories all describe what it takes from a leadership, physician engagement, health IT, and culture standpoint to not only survive, but also thrive, in a high-pressure ACO environment.
Nonetheless, there will still be doubters. Skeptics will point to the ACOs that are failing to generate savings or how much money it costs to operate a successful one. That’s not to say that these people don’t have good points, though. This 2016 blog from a Harvard health policy researcher explained that based on 2015 data, CMS actually lost $216 million that year since nearly every ACO at that time was in a one-sided risk model meaning they didn’t share losses with the government when they overspent past their benchmarks, but they did share in the gains. As such, in these one-sided risk models, CMS is on the hook for the losses all on its own.
Once again, there will be people on both sides of the spectrum. You can spin the numbers whichever way you prefer to make your stance stronger and you can poke holes in the stories of those who have done well. But one thing seems to be for sure: when it comes to ACO evaluation, there will always be good news and bad news.
Which side of the ACO story are you on? Send thoughts or questions to @RajivLeventhal or comment below.
