This event, set at the Fairmont Hotel in Chicago, was one of the most powerful and productive risk adjustment events I have attended:
The presentations were strong and meaty with content
The scope of the topics was geared perfectly to the target, including both Medicare and Health Insurance Exchange domains
The questions and answers were often vigorous and plumbed deeper and more keenly than expected
What Was Learned?
Of course, this depends on what you already knew and what you paid attention to. I can only offer my perspective:
1. Risk Adjustment for the Health Insurance Exchanges vs. Medicare
I thought I understood the differences before I came to the conference. But I found out just how superficial my grasp of it really was. And while multiple speakers primed the audiences with slides that offered to draw out the distinctions, I was surprised how many more differences emerged, depending on the particular speaker’s vantage point. The truth is that, by the time you factor in risk corridors, reinsurance, and the concurrent risk-adjustment methodology of HIX, predicting the financial performance of a marketplace plan is extremely difficult. Once you layer in the unknowns of the health status of the new buyers, you are dealing with a large degree of uncertainty.
Despite these significant concerns, I understand that the risk adjustment technique for the commercial marketplace will ultimately be more predictive than the Medicare model, once there is adequate data. Also, the smaller percentage of the population that is well-served by in-home assessments, added to the lower monthly revenues compared to Medicare Advantage, means that the whole discipline of risk adjustment will have to ratchet up in its efficiency and effectiveness in order to generate a ROI for the effort and investment. Consequently, powerful analytics will move to the forefront as the indispensable tools, and labor-intensive techniques that were sustainable only under Medicare Advantage will necessarily get less play under the commercial marketplace environment.
2. RADV Gets “Real”
Perhaps a bit hyperbolic, but when CMS finalizes this round of Medicare Advantage RADV audits, there will be price tags associated with error rates above a certain level for the first time. The big question, of course, is what will the FFS benchmark be and how far off will the healthplan error rates be in comparison?
Again, this is another key point in the difference in the way the commercial HIX market will work. First, there will be 100% RADV audits in this space, compared to the current 10% level in Medicare (potentially ramping up to 30%). However, the “take-back” penalties under HIX rules is accomplished in the next year’s payment rate reductions through an offset. It makes you go, “Hmmm”. How well is this going to work for all the at-risk players?
Again, this is a story that leads us back to data analytics. In order to build out a bullet-proof RADV exposure, a robust quality oversight and improvement program needs to be implemented. The mainstay of this program is going to have to be data-driven: under-investment in these tools can only be done at serious peril. Being on the wrong end of the risk adjustment game under a zero-sum rule is taking chances when the swings in one issuer’s risk profile against the overall market can be highly leveraged. In other words, leverage means outsized pain or gain.
3. The Feds Are Bringing Out the Big Sticks
The scariest creature in the woods may not be RADV. Once the attorneys introduce you to the False Claims Act application to the risk adjustment game, you will agree. One speaker pointed out the significant error rate of human coding at the practicing physician level, which is understandable, but exposes us to staggering financial risks, depending on how the feds decide to prosecute using this big stick.
4. Only 10% of the Risk Codes Are Going to Come from Claims
90% of the information, according to Apixio, is going to be textually-based. At first shocking, it makes us think about how we are going to find all the coders we will need and how to train them. We think in terms of coder productivity being something like 4 charts per hour.
But again we return to the crucial role of highly specialized data analytics. The strategy recommended here is to leverage the machine-learning of contemporary analytics engines to recognize and capture information that, in turn, is pushed in a highly efficient manner to coders. If this is indeed the future for marketplace HIX risk adjustment coding, it only makes sense that it would also transform the way we approach Medicare Advantage risk adjustment, and for that matter, Medicaid, as well.
5. There Is a Long Way to Go on Capturing and Implementing Best Practices on Coding
Dr. Richard Bernstein provided some real-world examples of physician coding errors alluded to above. There are methods of filtering and isolating a good number of these kinds of errors using advanced analytics. At the same time, Lynda Dilts-Benson engaged the audience in a demonstration that the best practices of coding and documentation require a lot more work, training and education, to really alter the quality of coding and, hence, risk exposure under both RADV and False Claims Act prosecution.
6. ICD-9 and ICD-10: Multiplying the Scale of Coding and Accuracy Concerns
Unless you have lived in a cave for the past couple years, you already know this. Perhaps you were hoping that the gems (general equivalency mappings) crosswalk would be the easy solution: mapping one-for-one from ICD-9 to ICD-10. It’s not going to be that easy. This is another case for thorough quality management and data analytic tools. Failing on this conversion does not look pretty, and speakers recommended that no one take the year off in preparing just because the feds blinked and postponed implementation until October 2015.
What Are the Bigger Picture Take-Aways from the Conference?
Several things struck me as I tried to grapple with the implications of what I was coming to understand.
Data analytics is king – to state the obvious from the foregoing comments. This is not only true of the stratification and targeting tools for risk adjustment. It applies to financial modeling, healthcare economics, member engagement and retention, as well as population health management. Silo data shops with subsets of information on specialized applications are swimming against the current of where things need to go.
First to the market – the old rule states that whoever comes out first and takes share becomes nearly impossible to catch. If that is true of the health insurance marketplace, there are plenty of frontrunners in many of the markets across the country. But what about the risks of stumbling when the risk profiles of buyers are only just beginning to emerge? I thought about the strategy that one Blue plan took in their market: creating a new “non-Blue” brand for this market so as to not grow too fast at first, and secondly, to pave a safer exit way in case the whole thing turned bad. Indeed, I am aware of plenty of players that decided on “go slow” as their motto. Time will tell which strategy prevails, but you only get one crack at the opening of a new market.
Running the Risk Adjustment Shop – With all this complexity and overlays of Medicare and HIX, the job of running a risk adjustment shop is only getting more challenging. The skills and competencies required go way beyond technical competence. Indeed, with the movement towards risk-adjusted payment streams for more and more business segments, the CFO is increasingly a key stakeholder. The revenue picture is elevated significantly beyond the back office visibility that RA had a few years ago. This is C-Suite material. There are highly competent technicians running shops across the healthplans in the country, but how is their boardroom game? I wonder. And how do we prepare these folks for success in their roles?
If you attended the conference and have things to add, or you want to comment on my remarks, please launch me a message so I can weave in your ideas and reactions, as well.
RISE Contact: Kevin Mowll firstname.lastname@example.org 831-465-2283