# Navigating Risk Adjustment Data Validation (RADV) Audits with Confidence, and the Costs of Unsupported Diagnoses

- **Show:** Becker's Healthcare Podcast
- **Published:** 28 May 2026
- **Source:** https://podcast.show/beckershealthcarepodcast/episode/153912596/
- **Transcript:** https://podcasts.opennash.com/shows/beckers-healthcare-podcast/navigating-risk-adjustment-data-validation-radv-audits-with-confidence-and-the-c-2e34cd/
- **Provider:** OpenNash Podcasts (machine-generated)

## In brief

Medicare Advantage plans now face bigger financial risk from unsupported diagnosis codes in federal RADV audits, which check whether the diagnoses used to set payments are backed by medical records. Dave DeHommel argues plans should stop relying mainly on after-the-fact chart chasing and instead build year-round clinical data capture, a central records repository, and audit-ready routines.

The tension is between legal uncertainty and operational certainty. A Texas court vacated CMS's extrapolation method after a Humana challenge, and CMS has appealed, so the audit rules may change. Yet CMS is still catching up on audit years, and every plan is expected to be audited. DeHommel says plans must prepare regardless of how the courts rule.

## Key insights

1. **Removing the adjuster raised unsupported-diagnosis repayment exposure** The fee-for-service adjuster acted as a statistical buffer, because plans set bids using unverified traditional Medicare data but audits look for near-perfect accuracy. Without it, plans have less cushion for unsupported codes.
2. **Legal uncertainty does not reduce audit preparation needs** A Texas federal court vacated CMS's methodology after Humana's challenge, and CMS appealed. DeHommel says CMS will still define some method and plans must follow it, with more audit years due in 2026 and 2027.
3. **Plans are shifting from retrospective to concurrent capture** Instead of waiting a year to review charts, plans want to put clinical insights in front of providers during visits, and chase records for 2026 services during 2026.
4. **New rule requires chart review results link to encounters** Starting with 2027 dates of service, coding results from retrospective chart review must be tied to an encounter to affect risk scores. The exception is new members, whose records can be reviewed and submitted unlinked.
5. **A central records repository speeds audit response** Medical records are requested for quality, utilization management, appeals and risk adjustment, often stored separately. Keeping them in one place lets plans reuse records, reduce provider burden and shorten retrieval time.

## Questions this episode answers

### What is a RADV audit and why does it matter to Medicare Advantage plans?

RADV (risk adjustment data validation) is a CMS audit that checks whether the diagnoses a Medicare Advantage plan reported are supported by medical records. Unsupported diagnoses can mean the plan has been overpaid and may owe money back.

### What was the fee-for-service adjuster and why did removing it matter?

According to Dave DeHommel, it was a statistical buffer that accounted for plans' payment rates being built on unverified traditional Medicare data, so plans did not have to be perfect in audits. Removing it increased plans' exposure to repayment.

### Is the new RADV extrapolation method still in effect?

Not currently. DeHommel says a federal district court in Texas vacated it after Humana challenged it, and CMS appealed in November 2025, so the outcome is still awaited.

### Why can't plans just rely on reviewing charts after the fact?

DeHommel says retrospective review will continue, but claims take roughly 90 to 180 days to complete, and CMS is focused on capturing patients' full illness at the time of care. Plans are therefore moving toward capturing information during visits.

### What should a plan leader do in the next 90 days?

DeHommel suggests starting a central clinical data repository, adding contract language that prioritizes audit-related record requests, running mock RADV audits when possible, and deciding whether to build or buy retrieval and coding support.

## Facts

- **Dave DeHommel** [0:02:17]: "I think CMS, when they came out with the 2023 ruling, they had estimated $17 billion in annual MA overpayments from unsupported diagnosis codes."
- **Dave DeHommel** [0:02:17]: "that RADV ruling was challenged by Humana. And a federal district court in Texas essentially vacated that ruling."
- **Dave DeHommel** [0:03:00]: "And as we know, CMS did appeal that ruling back in November of 2025."
- **Dave DeHommel** [0:03:00]: "the extrapolation methodology that CMS defined, which excluded that fee-for-service adjuster is no longer applicable."
- **Dave DeHommel** [0:03:46]: "they're doing it based off of traditional fee-for-service data, which is raw and unverified and includes essentially unvalidated diagnosis codes through claims."
- **Dave DeHommel** [0:04:40]: "that fee-for-service adjuster essentially was that buffer, that statistical buffer to support health plans."
- **Dave DeHommel** [0:05:26]: "But catching up with four additional payment year audits in calendar year 2026 and the remaining two in 2027, health plans have to focus and really execute in this area moving forward."
- **Dave DeHommel** [0:07:24]: "with the MA RAD-V and the OIG audits and some of the industry commentary, like from MedPak and others, it's really about, is that full burden of illness capture"
- **Dave DeHommel** [0:11:18]: "I think health plans are challenged because all health plans are being audited as part of this risk adjustment data validation audit, where historically it was maybe I will be selected"
- **Dave DeHommel** [0:13:32]: "there's a very long tail there, anywhere between 90 on a good day to 120 days to 180 days to get the bulk of that information"
- **Dave DeHommel** [0:14:46]: "going after medical records, for example, in calendar year 2026 on 2026 dates of service, as opposed to waiting the next year in 2027"
- **Dave DeHommel** [0:15:31]: "the advance notice just kind of reinforced tying the retrospective chart retrieval encoding documentation to the encounter through their claims linking criteria that was new and came out in the final notice this year."
- **Dave DeHommel** [0:15:31]: "starting on 2027 dates of service, the results of retrospective chart retrieval and coding projects, if it is going to impact the health plan's risk score calculation, it must be linked now to an encounter."
- **Dave DeHommel** [0:16:14]: "the only exception there is if there's a new member that joins the health plan, then they're able to look back and do some retrospective retrieval and coding and submit those unlinked."
- **Dave DeHommel** [0:00:48]: "previously, I was managing director of risk adjustment at Blue Cross Blue Shield of Michigan, spent 16 and a half years at Priority Health"
- **Dave DeHommel** [0:00:48]: "20 years of health plan leadership experience focused on care for seniors and vulnerable populations."
- **Dave DeHommel** [0:19:56]: "Reveleer put together a resource page at reveleer.com slash beckers dash rad v"

## Transcript

**[0:00:00] Erica Spicer-Mason:** Hello, everyone. This is Erica Spicer-Mason with Becker's Healthcare. Thank you so much for tuning into the Becker's Healthcare podcast series. So today we're going to talk about navigating risk adjustment data validation or RADV audits with confidence and the costs of unsupported diagnoses. And joining me for this conversation, we have with us Dave DeHommel, the Senior Vice President and General Manager of Payer Solutions at Reveleer. Dave, thank you so much for being here. Welcome to the podcast.

**[0:00:30] Erica Spicer-Mason:** It's great to have you with us. Thank you, Erica. It's a pleasure to be here. Well, so glad to have you with us. And before we get into the heart of the discussion today, I wanted to give you the chance to just share a little bit more about yourself and the work that you're doing, especially in the Medicare Advantage and the risk adjustment space.

**[0:00:48] Dave DeHommel:** Yeah, absolutely. Just to give a little more of my background, as you said, right, SVP and General Manager of Payer Solutions at Reveleer. I have 20 years of health plan leadership experience focused on care for seniors and vulnerable populations. That experience spans risk adjustment, revenue cycle, stars, quality, analytics, audit, and operations for Medicare Advantage, Medicaid, and ACA. 14 years of direct accountability for risk adjustment across multiple lines of business. And previously, I was managing director of risk adjustment at Blue Cross Blue Shield of Michigan, spent 16 and a half years at Priority Health, which is part of Corwell Health and IDN, and had previous roles at both Horizon Blue Cross Blue Shield of New Jersey and Deloitte.

**[0:01:34] Dave DeHommel:** And so I've grown up on the health plan side, but more recently joined Reveleer and have been at Reveleer for just over a year now. Well, great to learn that about you, Dave. And I'm sure

**[0:01:47] Erica Spicer-Mason:** your expertise in the payer space has served you well in your role at Reveleer. And I want to get right into just kind of setting a bit of an industry scene for our listeners here. I know that CMS's removal of the fee-for-service adjuster has materially changed the stakes of RADV audits. So from your point of view, how has that shift increased repayment exposure for Medicare Advantage plans, especially when diagnoses can't be supported? Yeah, it's a great question.

**[0:02:17] Dave DeHommel:** and a great topic that a lot of health plans and industry leaders are discussing. But yeah, with that removal of the fee-for-service adjuster, that essentially created more exposure for unsupported diagnosis codes. I think CMS, when they came out with the 2023 ruling, they had estimated $17 billion in annual MA overpayments from unsupported diagnosis codes. That's essentially what they were indicating would be left on the table if they don't perform these audits. But as many of us know in the industry, that RADV ruling was challenged by Humana.

**[0:03:00] Dave DeHommel:** And a federal district court in Texas essentially vacated that ruling. So the extrapolation methodology that CMS defined, which excluded that fee-for-service adjuster is no longer applicable. And as we know, CMS did appeal that ruling back in November of 2025. And so the industry and the health plans are really waiting for what that's going to look like in the future. But one of the main concerns, right, with that removal of that fee-for-service adjuster was it goes against kind of CMS's own requirements to make sure, you know, Medicare advantage organizations are actually a little sound, right?

**[0:03:46] Dave DeHommel:** And there's actually equivalence between traditional Medicare advantage fee-for-service versus Medicare advantage organizations. So when those health plans are required to perform their bids and set their premiums, they're doing it based off of traditional fee-for-service data, which is raw and unverified and includes essentially unvalidated diagnosis codes through claims. So there is opportunities for underpayments to occur in that data. And because that data is a part of setting the bid in the premium process, that fee-for-service adjuster was essentially a buffer to allow for that difference between setting your rates with fee-for-service traditional Medicare Advantage data.

**[0:04:40] Dave DeHommel:** And then when you have the MA-RAD-V audit, which is really looking for 100 % accuracy, that fee-for-service adjuster essentially was that buffer, that statistical buffer to support health plans. So they did not have to be perfect in the RADV audits. So now that that's been removed, right, exposure is much more significant. And CMS is still moving forward with really catching up with all of the payment years for Medicare Advantage. So health plans have to focus on this regardless of what the extrapolation methodology is going to be because CMS is going to define one and health plans are going to have to adhere to it.

**[0:05:26] Dave DeHommel:** But catching up with four additional payment year audits in calendar year 2026 and the remaining two in 2027, health plans have to focus and really execute in this area moving forward. And it's a critical initiative and needs to be a part of just their standard

**[0:05:50] Erica Spicer-Mason:** operating procedures moving forward. Yeah, Dave, I appreciate the way that you painted the landscape for us. I mean, it sounds like there's a lot of change that payer organizations are navigating and still some uncertainty as to how this is all going to shake out and what it will look like in the future and going forward. But appreciate the overview. And I want to talk a little bit more about what the current state is looking like at plans, especially when we're talking about radvi audits here i know that many plans still rely on things like chasing charts and retrospectively validating uh data can you touch on why this model is becoming harder to sustain and also to defend and what really

**[0:06:32] Erica Spicer-Mason:** breaks down when teams try to reconstruct those encounters months or even years

**[0:06:37] Dave DeHommel:** later yeah it's it you know it's interesting it's less about defensibility now and it's it's more about that operational and precision execution difficulty. Health plans are still going to perform retrospective chart reviews moving forward, but really what is shifting is CMS's lens and perspective. The Medicare Advantage model has always been developed and created to make sure that health plans are reimbursed appropriately and accurately to cover the cost of care for the enrollees, the patients, and the members.

**[0:07:24] Dave DeHommel:** And really, the focus is, you know, with the MA RAD-V and the OIG audits and some of the industry commentary, like from MedPak and others, it's really about, is that full burden of illness capture and the appropriate reimbursement that health plans are receiving, not only covering the cost of care, but driving improved outcomes for the patients and members. And so that's where I see kind of this additional scrutiny coming from CMS. And when you think about, right, that ability to go back and tie medical records with encounters and kind of constructing that data and information, it's really what's happening now on the health plan side is like a shift, right?

**[0:08:17] Dave DeHommel:** We're still going to need retro. There's going to be the need for retrieval and coding. What we're seeing now is a little bit more of a shift to move that to prospective initiatives or prospective programs that at the point of care, there's opportunity to put information information in front of providers to capture that full burden of illness. And instead of relying so much on retrospective, some of that is shifting to more of a concurrent opportunity and lens to get that information in front of the providers and try to capture as much information accurately and completely during the visit between a clinician and a provider and essentially their patient.

**[0:09:06] Erica Spicer-Mason:** Yeah, and I appreciate how you've nicely summarized this need for kind of a shift into more perspective strategies. Is there any example from your work? I know you closely collaborate with organizations across the country. So any examples that come to mind that really illustrate that tension or that challenge in shifting more upstream?

**[0:09:28] Dave DeHommel:** Yeah, I think health plans are definitely looking at the opportunity to ensure that they're giving clinical insights to the providers at the point of care. there's greater scrutiny and valuation around the return on investments for the different programs that health plans and risk adjustment teams have in place in trying to ensure that they're capturing that full burden of illness and really looking at centralized data repository, or a centralized image repository. There is a shift in a focus there because oftentimes health plans haven't really thought about all the different use case in the departments that are pulling medical records within the organization and where those medical records are stored.

**[0:10:26] Dave DeHommel:** So that kind of clinical data repository is a key where if you are focused there and start putting all of your images and data in one place, you can make sure that you're reducing provider abrasion, you're reusing the data and the medical records that you have already from prospective, from your quality programs, not just risk adjustment programs, from utilization management, grievance and appeals, right? Think about all the different use cases within a health plan where they're requesting medical records and how they're storing that data, there's definitely a shift in a purpose of how they're gathering that information and storing that information too as well.

**[0:11:18] Dave DeHommel:** And I think health plans too are looking at their provider contracts and the language in their contracts to say, can we put information and language in there in a way that when we have these CMS federal audits come up, we can ensure that our providers understand the urgency and the importance behind this and get us those medical records much more quickly than they have historically. I think health plans are challenged because all health plans are being audited as part of this risk adjustment data validation audit, where historically it was maybe I will be selected and it wasn't a muscle that the health plans really developed.

**[0:12:04] Dave DeHommel:** So therefore, they're really focusing on how do we make this be just a part of the standard operating procedure moving forward.

**[0:12:14] Erica Spicer-Mason:** Yeah. Yeah, Dave, I really appreciate the examples. It's very helpful to hear your perspective on what this is starting to look like on the ground. And it actually leads me nicely into another trend I wanted to touch on, which is that we're seeing some plans shift toward maintaining year-round clinical data capture. I know you mentioned those data repositories in your last response as well. But operationally, when we think about that year-round capture, what does it look like day-to-day and how does it change a plan's

**[0:12:44] Dave DeHommel:** readiness when the audit notification arrives? Yeah, from a day-to-day perspective, it's kind of a combination of perspective, chart-level data capture, and point-of-care clinical intelligence, right? So providers have what they need during that visit and at that encounter. It's just ensuring that that information is being shared as real-time as possible to the providers. And that's always been a challenge in the health plan kind of ecosystem because a lot of the times that data comes from claims and claims doesn't, you know, complete from like a incurred but not recorded perspective that IBNR.

**[0:13:32] Dave DeHommel:** So to the time of the visit to when a health plan is actually billed and that claim finally completes, there's a very long tail there, anywhere between 90 on a good day to 120 days to 180 days to get the bulk of that information so that you know which gaps are closed and then what do you put in front of that provider so they have the most accurate and clinical information to act on. But a part of that day-to-day work too is member engagement, right, or patient engagement.

**[0:14:05] Dave DeHommel:** How are health plans or providers getting their patients and members into that provider setting to have the visits to ensure that they're managing their conditions and they have that opportunity to both diagnose, document, and then bill that full burden of illness to the health plan. When you think about some of those day-to-day operations, it can be a challenge because a lot of the information is still stored in the medical record and the documentation deficit isn't really a challenge because it's in the medical record.

**[0:14:46] Dave DeHommel:** It's how do we get that information to the health plan? So then the health plan can accurately reflect to CMS the full burden of illness of their patient. And that day-to-day is shifting more to that prospective clinical intelligence point of care opportunities. And some of that retrospective work is still going to exist, but maybe moving it up for Medicare Advantage to be a little bit more concurrent. So going after medical records, for example, in calendar year 2026 on 2026 dates of service, as opposed to waiting the next year in 2027 and doing the larger projects for 2026 states of service medical retrieval and coding.

**[0:15:31] Dave DeHommel:** And the only other thing that I'll add there too, right, is the advance notice just kind of reinforced tying the retrospective chart retrieval encoding documentation to the encounter through their claims linking criteria that was new and came out in the final notice this year. So moving forward, starting on 2027 dates of service, the results of retrospective chart retrieval and coding projects, if it is going to impact the health plan's risk score calculation, it must be linked now to an encounter.

**[0:16:14] Dave DeHommel:** And the only exception there is if there's a new member that joins the health plan, then they're able to look back and do some retrospective retrieval and coding and submit those unlinked. But that is the only exception. Moving forward, all other members have to have those coding results linked to an encounter for it to impact the risk score and the reimbursement that the health plans receive. So health plans are shifting to make sure that they can do that moving forward and have a little bit of time to put that in place internally.

**[0:16:50] Erica Spicer-Mason:** Well, Dave, this conversation has been so helpful. I mean, your expertise is really shining through here. We're hitting at a level of granularity that I know is going to be very helpful for listeners. And yet we're moving through the conversation quickly, but I do want to make sure that we have time for just some closing thoughts and actionable steps that we can leave listeners with. For any leaders at pair organizations right now who are listening in and they're recognizing that they have some of these gaps that we've talked about in their documentation infrastructure, where should they start in the next 90 days to reduce audit exposure

**[0:17:24] Dave DeHommel:** before the next cycle? Yeah, I would say, as we kind of discussed previously, start with that clinical data repository, right? That central repository to be able to reuse medical records and the evidence the plan already has. It's the fastest way to reduce provider abrasion and really shorten those retrieval cycles, review provider contract language when agreements come up for renewal, or if there's an opportunity to, you know, add an addendum and language, you know, look to add terms that escalate and prioritize audit-related record requests, you know, above just your standard retro and concurrent requests.

**[0:18:08] Dave DeHommel:** The other one, as I think through this, right, Mock Rad V is going to be a little bit of a challenge. Some plans have done that to kind of build that muscle and rhythm. But because a lot of the health plans are obviously all of the health plans are going to be under audit this calendar year. when there is an opportunity to do that, just kind of create those mock RADV audits and standards just so everyone is fully aware of the workflow internally and how that works when your selection and sample arrives.

**[0:18:41] Dave DeHommel:** That's one of the best practices that we've seen. That might be something that's easier deployed once CMS catches up with all of these MA RADV audits. And then I think a couple others is the build versus buy decision on the health plan side, right? Those are conversations that leaders are having every single day. So just look honestly at internal resources and then evaluate partners that can lean in on retrieval and coding and validation and support in MA-RAD-V.

**[0:19:12] Dave DeHommel:** and Reveleer is one of those for sure. And really, when I think about MA-RAD-V and just risk adjustment as a whole, right? It's not really just a department that is accountable and responsible. When you think about MA-RAD-V and the financial stakes that are created as a result of these audits, it really is making sure that the health plans are pulling in all of the appropriate teams and departments, executives, and leaders to really drive a very successful MA-RADV audit protocol and lean in with some of those leaders and executives when escalations and issues come up.

**[0:19:56] Dave DeHommel:** And I would just say, you know, kind of in closing, you know, for folks listening who are thinking about and thinking through their own RADV strategy, Reveleer put together a resource page at reveleer.com slash beckers dash rad v and that is r-e -v-e -l-e -e-r dot com slash b-e -c-k -e-r -s dash r-a -d-v it walks through our ma rad v audit solution and our newly released clinical data repository which is that centralized evidence piece we've been talking about and i'm always happy to be a resource if this conversation hits home to anyone that's listening.

**[0:20:39] Erica Spicer-Mason:** Dave, appreciate all of the information and truly all the insights you've given throughout our conversation today. Thanks for making the time for Beckers. I know this will benefit everyone listening. Thank you. Thanks, Erika. It's my pleasure. And we also want to thank the podcast sponsor for today, Reveleer. Listeners, be sure to tune into more podcasts from Beckers by visiting our podcast page at beckershospitalreview.com.
