The 2015 Medicare Accountable Care Organization (ACO) performance results are out and they are a decidedly mixed bag. In a press release, the Centers for Medicare & Medicaid Services (CMS) put a positive spin on the numbers, pointing to the increased cost savings of $466 million (up from $411 million in 2014) that the Medicare Shared Savings Program (MSSP) and Pioneer ACO models collectively generated. CMS also called out the marked improvements in quality. All 12 Pioneer ACOs have improved their quality scores by over 21 percentage points since the first year of the program in 2012. Nine of them were above 90 percent in 2015. The average quality performance of MSSP ACOs improved by over 15 percent between 2014 and 2015 for four measures: screening for risk of future falls, depression screening and follow-up, blood pressure screening and follow-up, and providing pneumonia vaccinations.
The Good, The Bad and The Ugly
The increased cost savings and notable improvements in quality and patient care are absolutely reasons to be optimistic. The $1.29 billion in total Medicare savings that Medicare ACOs have generated since 2012 is nothing to sneeze at. Yet the results also clearly show that ACOs are still struggling. Despite the overall increase in savings, similar to last year, less than one-third (125) generated enough savings to qualify for shared savings bonus payments. The number of ACOs that achieved shared savings in 2015 is higher than in 2014, but not by much (120 in 2015 compared to 92 in 2014). Unfortunately, just under half (48 percent) produced no savings at all. Overall, most ACOs who contributed to Medicare program savings did not receive a portion of those savings. I’d venture to say that improved performance is not yet being rewarded.
Two Key Takeaways
First, I believe that change, particularly of this magnitude, takes time. The MSSP started in 2012, so CMS now has four years of data they can use to compare results. Their analysis shows that the ACOs that have been in the program longer and have more experience are, in general, doing better. Forty-two percent of ACOs that started in 2012 generated savings above their minimum savings rate (MSR), compared to 37 percent of 2013 starters, 22 percent of 2014 starters and 21 percent of 2015 starters.
It makes sense that over time ACOs are gaining a fuller understanding of the intricacies of the program and are learning what works and what doesn’t. This is a good thing. I’m very hopeful that with more experience they’ll continue to drive higher cost savings and improved quality outcomes. Additionally, I’m glad to see that CMS is continuing to make tweaks and improvements to the program in response to feedback from stakeholders and the public. They seem to be committed to continue making these modifications until they find the sweet spot in terms of the incentives and flexibility ACOs need to succeed. In a recent tweetstorm about the 2015 results, Andy Slavitt, acting CMS administrator, wrote this:
Sep 1, 2016
Andy Slavitt ✔@ASlavittReplying to @ASlavitt
At this stage, if an ACO tells me they are e.g. doing post-acute care differently, it is more encouraging to me than if they're saving $. 6
Andy Slavitt ✔@ASlavitt
Current models are like the iPhone 2. They are adapting to realities of pt care & providing fdback. 5 yrs from now will look different. 7
He’s essentially saying that CMS is going to continue evolving ACO models until they find what works – as they should. And no matter what they look like, their goal will always be the same – providing better care to an increasing number of Medicare beneficiaries at a lower cost.
Second, theories abound as to why ACOs are struggling. I believe it can be boiled down to one reason: the lack of good analytics to make the right decisions on care treatment. Since CMS uses risk adjustment to set ACO benchmarks and reimbursement rates, I can’t help but wonder if some of the issues stem from problems in their risk-adjusted coding practices. High-accuracy coding is the key to giving providers a complete snapshot of the true health of their population so they can identify, plan for, and appropriately treat their high-risk patients. I suspect many ACOs’ current coding practices are manual and error-prone, resulting in missed codes, inaccurate risk scores and benchmarks, and unrealized reimbursements. Medical practitioners need easy-to-use tools to “code for care.”
ACOs can improve their HCC coding by instituting new, automated coding practices that leverage technology such as data and text analytics and natural language processing (NLP). With the right tools they can achieve significant improvements without relying on their already time-squeezed providers or hiring an expensive army of trained coders to manually comb through charts. ACOs should make the investment in improving their risk adjustment process for better care planning, more accurate RAF scores, and benchmarks and reimbursement rates that reflect their actual cost of care – all the elements that are essential to their success.