RISE Radio

Episode 21: Centauri Health Solutions’ Dawn Carter offers insights into the transition of risk adjustment model V24 to V28

• Ilene MacDonald • Episode 21

Share episode

Dawn Carter,  director of product strategy at Centauri Health Solutions, joins us for the latest episode of RISE Radio, our podcast series that focuses on issues that impact policies, regulations, and challenges faced by health care professionals responsible for quality and revenue, Medicare member acquisition and experience, and/or social determinants of health.

In this 23-minute podcast, recorded on April 18, Carter discusses the shift from the V24 to V28 risk model, the coding changes, and its risk adjustment applications. 


About Dawn Carter 


Dawn Carter, BSBA, CPC, CRC, CPMA, CDEO, CPCO, CSPO,  is a director of product strategy at Centauri Health Solutions. Her career in health care spans 25 years, which most recently includes extensive experience in developing revenue integrity and quality software solutions, with a focus on encounter management and risk adjustment solutions for Medicare Advantage, Medicaid, and Commercial health plans. 

She also provides strategic advisory solutions and consulting services for revenue cycle operations. Prior to that, her experience spans all domains of health care including health plan claims and provider systems administration, and healthcare applications development. Her experience also includes multiple teaching engagements in medical administration, billing, and coding. Carter holds a bachelor’s degree in business administration. She is a passionate and prolific industry speaker, author, blogger and subject matter expert in claims, EDI management, and risk adjustment. 

 About Centauri Health Solutions 

Centauri Health Solutions is a leading provider of technology-enabled analytics and services helping health plans and health systems to manage their variable revenue linked to population health (risk), quality, and eligibility factors. These efforts result directly in better-informed health care delivery, richer benefits, and reduced out-of-pocket healthcare costs for the members and patients they serve. 

 

 

 

Ilene MacDonald:

Hello and welcome to the latest episode of RISE Radio. I'm your host, Ilene MacDonald, the editorial director at RISE. Today we'll be discussing the shift from the V24 to V28 MA risk model, the coding changes, and its risk adjustment implications. I'm delighted to have my guest today be Dawn Carter, director of product strategy at Centauri Health Solutions. Centauri provides services to payers and providers across all health care programs, including Medicare, Medicaid, commercial, and exchange. Welcome Dawn, thanks for joining me today.

Dawn Carter:

Thank you, Ilene, for having me. Always delighted to support, RISE and share with the industry, so thank you.

Ilene MacDonald:

Thank you, and I thought maybe we could start today, for those who might not be familiar or might just need a refresher, if you can explain why CMS is moving away from the V24 model to V28.

Dawn Carter:

Certainly so, as we all know at least those of us who've been in the industry a minute,

Dawn Carter:

we know that from time to time CMS does change the model, and they do this because they perform ongoing analysis year over year, and when they do that, they examine utilization patterns, coding patterns, expenditure trends, all these things that need to be considered to improve the model's predictive accuracy, because this is what feeds the risk-adjusted payments and their accuracy and then having them be able to be in line with what plans expect from a forecasting perspective.

Dawn Carter:

So this time, CMS said that the new model that they are shifting to reflects more recent patterns for utilization, coding and expenditures, and the criteria they generally use when they evaluate this is, they look at conditions that don't accurately predict costs, and we see this reflected in this change. The model coefficients were small. The conditions they represent are uncommon. We did see some of this also with this model shift. O r the conditions did not have well-specified diagnostic coding criteria. So we see there where they're looking at utilization, coding and expenditure patterns, and so now we are making this transition and no doubt, probably in the next couple of years or so, there will be yet another.

Ilene MacDonald:

Always changes.

Ilene MacDonald:

CMS typically takes a phased-in approach to implementation when they make such a significant change. Can you share what the phase-in looks like for V-28?

Dawn Carter:

Yes, so actually the phase-in from, unbeknownst to a lot of folks, the phase-in actually has already begun, and just because of the nature of how risk adjustment works for Medicare Advantage, and it's a three-year phase-in. So this is pretty typical for CMS to phase in the model for a variety of reasons, and so this one's no different and the phase in actually began in calendar year 2023, because blended risk scores are going for calendar year 2023 will be paid in calendar year 2024. And so that phase-in begins as a you know what we know to be like a percentage phased-in approach. So it starts with the first payment year 67% is V24, 33% is V28. Then in 2024, it shifts, it flips around so that the new model is 67% and the old one is 33. And then by the last year it will be fully phased in at 100%, and this is pretty typical for CMS to phase this in. So that's what they've been doing since 2023.

Ilene MacDonald:

So let's talk a little bit about the highlights of the changes of the HCCs and the ICD-10 mappings.

Dawn Carter:

Certainly. They fall into what I see as kind of six main buckets in terms of the high level changes. If people wanted to just have a down and dirty about you know what really happened here and the highlights of the change. First is probably how the HCCs are named and numbered. That's pretty typical. They want them to be descriptive and of course the numbers change.

Dawn Carter:

And this happened, of course, to some of the more common conditions that we're used to seeing, like diabetes and some other things. There also is an expanded number of HCCs. We're moving from 85 HCCs to 115. So, expanded number of HCCs. However, the changes to the ICD-10, to HCC mappings for V24, there are about 9,700-ish ICD-10s that map to those 85 HCCs. But for V28 that number reduces because remember we talked about how they're looking at diagnosis codes specificity and those types of coding issues that draw coding patterns that drive the calculation. So we've now got 7,720 ICD-10s that map to the 115 HCCs. So that's kind of an interesting shift there.

Dawn Carter:

And also they removed, as we see, 2,294 ICD-10s that no longer map to an HCC because they have low predictive accuracy. Because, remember, that's one of the things that CMS is evaluating for when they are making these changes. So they removed quite a few because just low predictive accuracy, and they added 268 codes that did not previously map to an HCC to increase specificity. Now it's important to note that over 40% of these codes represent conditions that are not typically present in the MA population, such as perinatal conditions, things relating to pregnancy, and congenital conditions, because those apply to babies and children a lot. So they've removed a lot of those conditions that aren't typically present for this population and then again to increase the predictive accuracy.

Dawn Carter:

And then, last but not least, they added constraints to the diabetes and congestive heart failure categories. This is one that's been receiving a lot of attention because a lot of Medicare beneficiaries have this. Both have very high prevalence and in V28, the HCCs within these categories will carry the same weight. So at a high level, those are your quick hits for the highlights of the changes.

Ilene MacDonald:

Dawn, how do these changes affect the risk adjustment factor or RAF?

Dawn Carter:

So, there's a number of changes that have been widely discussed. We know that a lot of our colleagues in the industry have published a lot of great content regarding the estimated impacts, and these are folks like Pareto and PricewaterhouseCoopers so if anybody's looking for like a deep dive into the RAF impact, those are some good sources to go to, because they really did a deep dive into this with their actuaries and their finance staff. So the estimated impact lies in a few different things and you know, for example, more of the risk score will be attributable to demographic factors than diagnosis factors due to this recalibration. And these demographic factors, of course, are things like age, sex, where they live, things like that. So that's first and second.

Dawn Carter:

The new model will affect reimbursement most significantly for beneficiaries who are new to Medicare, and so the source I looked at estimated for dual eligibles there would be a pretty good size of RAF impact here at about negative 6.2 percent. Then for the non-dual eligibles it'll be about negative 3.5 percent and then new to Medicare's it'll be a jump to 13.9 percent RAF impact. So that's pretty significant and that's something for health plans to consider, of course, when they are planning their outreach strategies for new members and their campaigns for their welcome to Medicare visits, in-home assessments, chart review campaigns, and things of that nature. So that was pretty significant. Another source said that for different models so utilizing 2022 dates of service for community non-duals, partial duals, and fully dual members, and this was Pareto, they analyzed the differences between the two models for MA enrollees and then they found that, prior to applying coding intensity and normalization factors, the proposed V28 model is resulting in a RAF decrease somewhere in the range of 7 to 22 percent, with an average of 13 percent, and then, after applying the updated fee-for-service normalization factor, the impact of the model is, you know, positive 2 percent to negative 14 percent, with an average of negative 4.

Dawn Carter:

So that's a wide, wide spread. So that was just kind of their take on it. And then, in terms of looking at annual risk-adjusted premiums, they used an average of 11 coverage month and a $750 premium amount. They calculated the RAF decrease, translating into an average $418 per member per year decrease in an annual risk-adjusted premium. And, of course, these numbers change depending upon who's doing the analysis.

Dawn Carter:

700; white-space-collapse: collapse;">wn Car</span>lene MacDonald: I noticed that CMS mentions that some of the diagnosis codes were removed due to principle 10. Can you explain what that means and why they are calling it out? Yes, principle 10 is something that's relevant to the risk adjustment methodology, which is part of what we call the guardrails of risk adjustment. They're kind of guidelines, rules that we live by for risk adjustment, and so it's important to understand what that means and why it's significant for them to call out. Now for principle 10, what CMS did here was review conditions where coding variation in Medicare Advantage was higher relative to Medicare fee-for-service, so traditional Medicare and they did this with its clinical experts to identify conditions that had high discretionary coding variation. So this is where different clinicians are coding differently, maybe more specifically, maybe less so, or with other factors, but that's what they're trying to identify is you know where's this discretionary coding variation taking place? And so principle 10 says discretionary diagnostic categories.

Dawn Carter:

These are ones that are quote particularly subject to intentional or unintentional discretionary coding variation or inappropriate coding by health plans or providers, or that are not clinically or empirically credible predictors of future expenditures. They should not increase cost predictions, and they think they say these should be excluded from prospective payment models. That's a mouthful, but it basically, again, is one of those strategies that is used to level the playing field, because they're comparing this with Medicare fee for service and it's sort of like elimination of outliers, so to speak. So if everybody's eyes glazed over with all of that definition, you could say leveling playing field, kind of excluding outliers, and that's kind of where they're going with this.

Dawn Carter:

Now, luckily, there were only a small number of codes that were removed due to this principle, 75 of those 2,294 codes, which is 3. 4 percent, were removed due to this principle. But because it was specifically called out, people might be wondering what does that exactly exactly mean and why do we care about it? Well, we of course would probably care more about it if it was a lot more codes, and you know we would want to, you know, examine that a little further, but it was only 75 of that overall number.

Dawn Carter:

Speaking of that overall number, can you talk about the reasons why CMS removed over 2,000 ICD-10 codes? So most of those removed and we have like 40 percent of them were removed because they were related to sequelae, which are late effects of a disease process, and this was done because ICD-10 guidelines require the actual late effect to be coded separately. So for those coders in our audience we want to give you a shout out saying because they probably understand exactly what that means and know where it is in the coding book, in the coding guidelines, that it says that. So these are the ICD-10s. Again, you know, for the end in S for coders. The next highest percentage, at 28 percent, were removed based on a combination of model principles such as clinical meaningfulness and their tendency towards overprediction. So that's, kind of you know, the biggest two categories that were removed.

Ilene MacDonald:

And can you say which conditions were most impacted by the change?

Dawn Carter:

So most impacted by the change... There were four that were probably the most impacted overall and that would be major depressive disorders, diabetes with chronic conditions, vascular disease and rheumatoid arthritis and inflammatory connective tissue diseases. Those were the four that most sources and myself feel were most impacted by this change and this shift.

Ilene MacDonald:

And what additions and deletions do you consider the most significant?

Dawn Carter:

Oh, there's quite a list, according to most sources, of the most significant adds and things that we haven't seen before, and these are things like anorexia nervosa and bulimia nervosa, alcoholic hepatitis with and without ascites, malignant pleural effusion, obstruction of the bile duct, severe and persistent asthma, toxic liver disease with hepatitis, benign neuroendocrine tumors, sarcoidosis of the skin, post-polio syndrome, birth trauma and maternal use of drugs, and newborn problems and disorders specific to the perinatal period. So those are our high-level categories of codes that were affected.

Ilene MacDonald:

Dawn, what makes this shift different from any of the previous model shifts, other than the number of HCCs and the ICD-10 mappings?

Dawn Carter:

So the first, more obvious reason or obvious difference is we see that CMS, you know, continues to get better at measuring and managing the predictive accuracy of the model, and we can expect, and we can also see, that they do this ultimately to realize financial savings.

Dawn Carter:

We can definitely expect to see more work being done related to how social determinants of health affect the predictive accuracy of the model, because much of the current research indicates that cost is not an appropriate surrogate for predicting risk in vulnerable populations. So we can expect to see continued work on this. We've already seen work from the quality perspective with the Health Equity Index and the Health Equity Summary Score, because CMS recognizes that caring for these populations is more of a challenge. So we'll continue to see a lot of work in that regard. And it will also help incorporating this data into what we're doing helps plans not only promote health equity but also helps them more accurately target the most appropriate interventions to people with the highest clinical risks. So ultimately, as they get better at measuring and managing the predictive accuracy of the model, it helps better inform in the long run, the strategies that we use in risk adjustment for getting ahead of these conditions and also targeting the most appropriate interventions to the people who need them most, and so this also affects quality.

Ilene MacDonald:

Can we talk about how the shift affects risk adjustment strategy from a member and provider perspective? Can you maybe share what things health plans and other organizations should be thinking about as the shift takes place?

Dawn Carter:

Of course, this is a very timely question, you know, given that we just heard from CMS in the final notice and the final rule, and so, as a result, we know that, more than ever, attention will need to be paid to clinical documentation improvement and diagnosis code capture. Workflow and the technology that supports it is going to make the difference here. For example, understanding the prevalence of conditions affected by this shift are going to inform strategies for in-home assessments, wellness exams and other strategies throughout the year and more attention to the severity of conditions over time. And there are a lot of companies out there doing innovative things with lab testing, for example, in-home lab testing and other services to help plans get ahead of these conditions sooner. And providers will also need to have meaningful business intelligence and education regarding diagnosis code capture and the impacts of the work they do on risk and quality. And so it's not enough just to present them with the whole array of reports that says, hey, you need to do a better job of coding or here are your gaps, please close them. You know, providing them with meaningful tools and meaningful education and seeing so they can see where they compare to other providers is going to make a difference here, you know, fully engaging with them and enabling them. There's a difference between provider engagement and provider enablement. So there are also tools that capture data from unstructured sources, such as medical records, that are going to become more important to work smarter and not harder. So these we hear a lot about these your artificial intelligence, machine learning, and other technologies these are we hear more and more about these as helpful tools to support that. Now, this doesn't mean that things like coding jobs will go away. It just means that there will be more opportunity for complete and accurate capture of data that makes a difference.

Dawn Carter:

And recently at a RISE conference, I actually heard a physician say providers are going to be expected to use these tools to support and augment what they're doing, working smarter and not harder, and those that don't embrace that are going, I think he said, they're going to be left behind or have a hard time, and this was a physician who stood up and said this. So you know, everybody recognizes that there are a lot of helpful tools and technologies out there to help address workflow, help address, especially like coding, you know things of that nature and it's important to kind of keep abreast of what's going on there. So, overall, meaningful business intelligence, making sure that you have a robust clinical documentation improvement program, and those are the things that are going to make the biggest difference in terms of making sure that the right members are getting the right care at the right time. And then influencing that Quintuple Aim. You know it's not just about improving cost, improving quality, the member experience, the provider experience, but now the advancement of health equity, and this is in alignment with CMS' overall strategic plan for the Quintuple Aim and the advancement of health equity that comes. Everybody's got a part to play in that.

Dawn Carter:

And then we can't forget, you know, of course we've got, you know, the OIG out there and we just heard, you know, this week, from Christy Grimm, you know talking about.

Dawn Carter:

You know the increased scrutiny that's happening in terms of the.

Dawn Carter:

fraud, waste and abuse, false claims, and then also earlier in the week we heard about still a lot of attention being paid to prior authorization and making sure that members and Medicare Advantage have the same access to care that they do in traditional Medicare.

Dawn Carter:

So that scrutiny is going to continue and so there's a lot for plans to worry about and it can seem overwhelming, but if you have the right tools and the right support in place, it's going to make it a lot easier to manage. You know all of those moving parts and it is. You know it's a lot when you think about it, and certainly there, but there's a lot of support out there in the industry in terms of resources, in terms of RISE, we've got a lot of great vendors and other entities out there that put out a lot of great content and so, luckily, we have those resources at our fingertips to help us be able to, work smarter and not harder and then making sure that that member is at the center of everything we do, regardless of, you know, a risk model shift or anything else that's going on.

Ilene MacDonald:

Thank you, Dawn, really appreciate everything that you offered today, and if people need to reach you, they can get you at Centauri.

Dawn Carter:

Absolutely can, and everybody knows I have an open door policy. I always welcome questions discourse. In fact I had somebody reach out to me last night and say, hey, can I get on your calendar for a few minutes? I want to ask you something. I'm like absolutely and I always enjoy those opportunities.

Ilene MacDonald:

Thank you have a great day.

Dawn Carter:

Thank you so much.