Medical Advantage Podcast Ep. 37 “Best of” Episode: Optimizing Risk Adjustment

Full Episode Transcript

Medical Advantage Podcast: Welcome to the Medical Advantage Podcast, where you can hear healthcare professionals, expert consultants, and industry thought leaders discuss the exciting new ideas and technologies that are changing the business of healthcare. Tune in to each episode as we hear from some of the most innovative minds in medicine about the future of healthcare and how your organization can stay profitable, efficient, and on top of healthcare best practices.

Celina DeFigueiredo-Dusseau: Hello everyone and welcome back to the Medical Advantage podcast series. As always, I am your host, Celina Dusseau, and today I am incredibly excited to discuss a very important topic in healthcare, optimizing risk adjustment. Here with us today, Medical Advantage’s very own, Angela Hale and Kyle Enger, consultants with extensive experience in the risk adjustment landscape.

Angela and Kyle, thank you for joining us today. Would you like to start us off by sharing just a bit about yourselves and your roles here at Medical Advantage?

Angela Hale: Sure, thank you so much for the invite, I’m happy to share the knowledge that I have. My name’s Angela Hale and I am one of the healthcare transformation consultants at Medical Advantage. I’m primarily responsible for the in-practice support for the providers to document their patients that are high risk into the chart documentation.

Kyle Enger: Yeah and I’m Kyle, Enger, I’m a senior analytics consultant with Medical Advantage. I primarily analyze healthcare data for physician organizations.

Celina DeFigueiredo-Dusseau: Fantastic. It is so great to have you both here, so let’s jump right in. As we know, risk adjustment optimization is one of the most important areas we work in here at Medical Advantage, but it can be a bit complex to us non-experts to understand at first. I have here a definition one of our own consultants has used to describe risk adjustment optimization that I would like to start with, and that is risk adjustment is the process of accounting for financial risks so that payers and providers are paid more to manage more complex ill and risky populations. And we’ll get into the specifics a bit more in a moment, but broadly, do the two of you agree with that definition? How would you summarize risk adjustment optimization in your own words?

Kyle Enger: Yeah, I agree with that statement. It really covers the basic idea risk adjustment. Really risk adjustment starts at the payer level and its primary use is to properly price health insurance, but payers also apply risk adjustment to value-based care programs in order to reward providers for effectively managing complex patients.

Angela Hale: And I also agree with that, Celina and Kyle. I think that risk adjustment, if you describe it as kind of a score that’s assigned to identified patients, there’s about 72,616 diagnosis codes that are used. And that’s for 2021. Not all of those codes have a risk score assigned to them. About 9,757 of them this year do.

So that score is what’s used to show that risk that we were just talking about by adding all of the codes that have a risk score assigned to them, totaling them together, and you get what’s called a risk adjustment factor, or because we like alphabet soup and healthcare, it’s an RAF. As well as that score can allow some other opportunities for the patient for support services that wouldn’t otherwise be available to that patient.

Celina DeFigueiredo-Dusseau: Wonderful. Thank you both. And before we dive into risk adjustment services proper, I’d like to focus a little bit on the concept of risk scores and Angela, you just started to touch on that a bit. What exactly is a risk score and how is it generated?

Kyle Enger: I can go into some more detail there. So a risk score is a number indicating a patient’s financial risk, and that’s naturally highly related to medical complexity. These risk scores are generated by a statistical model that’s applied to a particular patient population. There are a lot of different organizations. Some of ’em are government, some of ’em are companies that own and maintain these models. But in general, the risk scores, they produce all operate in the same way.

So a risk score with a value of one indicates average financial risk. That means the patient is expected to have roughly average healthcare costs. Higher numbers mean higher costs while lower scores mean lower costs. So for example, if you have a risk score of 1.3, that means we expect the patient to cost 30% more than average.

And on the flip side, if you have a risk score of 0.7, that means we would expect the patient to cost about 30% less than average. Again, those are averages if you’ve got a patient with a risk score of 1.3, there’s no guarantee that they’re actually going to cost 30% more. And if you’re looking at an in individual patient, they almost certainly won’t.

They could be much higher, they could be much lower. But risk scores are really intended to be applied to populations. So if you had a big group of

patients with a risk score of 1.3, you’d expect that whole group to cost about about 30% more. Anyway, these risk scores are generated by applying the statistical model to each patient’s demographic information, including, for example, age and sex, and also the diagnosis codes from healthcare claims.

Sometimes the risk models also use procedure information or drug information, but the diagnosis codes are the most difficult kind of information to get partly because of the complexity of the ICD-10-CM diagnosis coding system, but also because, in order to code specifically in a manner that benefits risk adjustment, that’s more work and it’s more difficult than just coding enough diagnoses so that the provider gets paid for the claim on a fee for service basis.

So that’s really the reason why when we’re talking about optimizing risk scores, we talk mostly about diagnosis coding. Also it’s really important to realize that risk scores are specific to populations. So if we consider a typical family practice, they probably see both commercial and Medicare Advantage patients that they might see those patients from multiple payers, they probably have traditional Medicare patients, and on top of that, they might have Medicaid patients. And all these different payer subpopulations can have different risk models and different risk scores that have to be interpreted in the context of the population.

So if you remember a risk score of 1.0 means about average financial risk. And in a Medicare population, that means an expected cost of somewhere around $12,000 per member per year. But if you’re looking at a commercial population, you can have the same risk score 1.0, but it means something quite different.

It means an expected cost in the neighborhood of say, $5,000 per member per year, because the commercial population is younger and healthier than the Medicare Advantage population. So it’s pretty common for a practices’ population to be divided among four or more risk models.

That can sound kind of complicated and who can keep track of so many different risk models? But the good news is that if a practice is tracking and improving risk scores in at least one of these models, they’ll tend to improve in all of them because clinicians aren’t going to change their documentation, coding practices, depending on which payer each patient happens to use. And also all these risk models, they all really depend on diagnosis codes, it’s central to risk coding.

Angela Hale: And so also building on that, when we look at some of the large payers like CMS, which is Medicare, they put out a benchmark for the provider so that they’re able to measure the risk scores. Individual and practice groups get these reports, they’re given annually to the providers that care for the Medicare patients.

In large groups of physicians it may be easier for them to track their risk score or how accurately they’re capturing those risk scores. These groups usually have a lot of support around coding education and how to document, and how to code those high risk populations. For those smaller practices these reports are also available to them but they don’t have as much support on how to capture and improve their documentation of those high risk populations.

The reports reflect how the provider performed the previous year. So if the providers are above the benchmark, meaning they cost the payer more than the benchmark, they’re penalized. And then as well, if the provider is below the benchmark, which means they’ve saved the payer money, they’re rewarded.

So the report’s used in many ways to analyze the provider’s success in the value-based care, with only one part being the risk adjustment.

Celina DeFigueiredo-Dusseau: Great. So in a nutshell, a risk score is like the representation of a patient’s likelihood to require more expensive care. I can already imagine how having an accurate risk score would be so important for reimbursements. And i’m interested to hear a little bit more from the two of you on that. Why is having a risk score that accurately represents your patient’s population and complexity so important?

Kyle Enger: Well, it really matters to providers’ bottom line. Really providers will be paid less if their patients risk scores don’t completely measure their patient’s complexity because health insurers can charge more to ensure sicker populations, but they can only prove that the population is sicker if they have the risk scores to back it up.

And because providers are the source of the diagnosis codes that the risk scores use, health insurers risk adjust value-based care programs so that they can effectively reward practices for caring for higher risk patients in an efficient manner. And this also has the secondary effect of rewarding accurate risk reporting, which is data that the health insurer really needs.

As healthcare continues to move away from fee for service payment and towards value-based payment, risk adjustment becomes ever more important for providers’ bottom lines. And this also means that payers have to audit providers to make sure they’re documenting and reporting diagnoses correctly and the information feeding into risk adjustment models is actually reliable.

Angela Hale: So payers have developed a model that they show what it should take or the financial burden to care for a patient. Payers such as CMS, again, that’s Medicare, do audits on these charts to verify that the patient’s record has the documentation to support those high risk codes for each visit. So each year, the payer requests patient charts to audit for identified patients.

The audits look for the documentation that supports, again, that high risk code, and there’s specific criteria that would support each code. If in an audit, the code is not supported in the documentation, then remember that RAF or risk adjustment factor is not validated. So likewise, if the documentation supports codes that were not included in the claim, then the risk adjustment factor or RAF is still not correct.

So we look at some of the audits and they’re called a RADV audit, and it can be done either by Medicare to the payer or from the payer to the provider. So either way, the same goal and they’re looking at the same patient record, same diagnosis codes that’s on the claim for the data service to support or to validate the documentation.

The only big difference in the typical audit is typically the payer audits the records before the claim gets sent to Medicare. And otherwise, Medicare then does audits after the claim’s been sent to them in process. So this allows the payer to find issues before the claim is sent to CMS.

Celina DeFigueiredo-Dusseau: That’s such a wonderful summary of why it’s so crucial for organizations to ensure that their risk scores are accurate. And of course for many organizations that can be a little bit easier said than done with this being such a complex topic. As we start to transition to talking about optimization techniques for risk adjustment, can the two of you start explaining some of the terms that you’ve mentioned, like HCC’s, RAFs, RADV, and MEAT system.

Kyle Enger: Sure, yeah. Just like the rest of the healthcare system, risk adjustment has its own set of jargon and acronyms. HCC stands for hierarchical condition categories, and these are disease classifications that are used in various CMS risk models. They are groupings of diagnosis codes that relate to particular conditions.

Now, the hierarchical part, that means that more severe manifestations of a condition trump less severe manifestations. For example, one patient, over the course of the year, they might have diagnosis codes on their claims that indicate both uncomplicated diabetes as well as diabetes with complications.

Now, if those are two separate hierarchical conditions in the risk model, diabetes with complications and diabetes without complications, but diabetes with complications gives a larger risk adjustment factor or RAF than uncomplicated diabetes. So in that case, we only use that bigger risk adjustment factor for complicated diabetes.

So Angela was talking a bit about RAFs before, but you can think of these risk adjustment factors as a partial risk score. If a patient has several conditions, each of the condition has its own risk adjustment factor and they’re added up to get the full risk score. There are also a few other risk adjustment factors that pertain to the patient’s age and the patient’s sex and things like that, that also feed into the risk score. Really just each risk adjustment factor is just a part of the total risk score.

Angela Hale: And also MEAT is another acronym that indicates what’s necessary to code a condition. And it stands for monitored, evaluated, assessed, or treated. And at least one piece of this MEAT is what needs to be done in respect to the particular particular condition for it to be then coded on the claim. And then RADV, which I mentioned earlier, stands for the risk adjustment data validation where CMS audits the Medicare Advantage Health plans to verify that each of those HCC codes or diagnosis codes that have those risk assigned to ’em, that those codes are included on the patient visit are supported by the documentation.

Celina DeFigueiredo-Dusseau: Thank you for explaining the acronyms for such important concepts. Of course, we have acronyms for everything in healthcare, and it brings about many questions. What I’m curious to know is what are some of the questions that organizations can ask to assess the accuracy of their own risks scores? What can they do to improve?

Kyle Enger: Okay, this is a tough question. It’s hard to know what a particular population’s risk score should be because we really don’t know what the diagnoses were that should have been coded but weren’t actually coded. But we know that generally practices underreport diagnoses because first of all, it’s extra work to code completely and specifically. But also the payoff for that extra work, it happens at some future time when whatever value-based care program that you’re participating in decides to send you a payment which might be several months or even a year or more down the road.

Whereas if you’re billing fee for service you get your claim paid probably within a few weeks, and you know exactly what happens and you know exactly how much you got paid for it. The reimbursement for value-based care is, it’s in the future and it’s more uncertain but organizations that audit their documentation and coding that can help practices and providers assess where they can improve their documentation and coding and thus improve the risk scores.

And it’s also really important to know whether your billers and coders understand risk adjustment and take that into account when they’re doing billing. When they’re doing billing, billers and coders really should have good training. The American Association of Professional Coders or AAPC has a number of coding certifications. Probably the most popular one is Certified Professional Coder, which I think really any professional bill or coder ought to have. But there’s also a certified risk coder certification, which specifically gets into risk adjustment concepts. Angela and I have both earned that. As well as several other of our colleagues at Medical Advantage have earned the CRC certification too.

It’s also really important for there to be clear and open communication between physicians and coders that lets them check each other and cooperate. Both to improve documentation by the physician as well as coding by the coder. And sometimes if you have your billing services subcontracted out, it’s happening outside the practice, sometimes that communication can break down. But it’s really important, it can be very useful for improving documentation and coding.

Angela Hale: And, and also as Kyle mentioned the audit program that the practice has is extremely important and they’ve gotta make sure that the program has feedback to the provider so if there are opportunities for improvement, those are identified and the provider then is giving the education or training, whatever’s required to make sure that those improvements then are made moving forward.

Kyle Enger: Yeah and if you can, tracking your risk scores over time is beneficial too. There are a lot of EMRs that have tools that can help with this, some of them might require an extra fee to enable that. You can also look at reports from the value-based programs that practices participate in, because those often include your risk score and some of them even include risk scores for other participants in the program or for the whole patient population in the program. And that can help you understand where your risk reporting is relative to your peers.

Also one other thing that helps me when thinking about risk reporting is to think of it as a chain of events, where each event must succeed in order for a condition to be included in the risk score. So if we take it all back to the beginning, the patient has to actually seek care for their condition, and then the provider who sees the patient has to document it, then the coder has to code it from the documentation. And as all coders know, if it wasn’t documented, if it wasn’t written down, it didn’t happen.

And the coder is never gonna guess on documentation. They’re going to just pay attention to the strict letter of the documentation. So if it’s not documented properly, the coder isn’t gonna code it. And finally the claim has to get billed to the insurance company. We all know that providers sometimes get behind on billing. Sometimes care happens, it was useful to the patient, it was important care, but sometimes it just never gets billed. And if the billing doesn’t happen, you kind of lose twice. You lose on that fee for service payment for the procedure and you also lose on risk reporting and your value-based care payments might go down as a result.

So if any of these four links in the chain fails, then risk reporting fails and the patient looks healthier than they really are. And this illustrates that even though we often think of documentation and coding of diagnoses as the most important issues for risk reporting, you can see here that patient outreach and scheduling visits and prompt billing are also important components of risk reporting.

Angela Hale: And also we wanna make it clear this isn’t about increasing the risk score. We wanna make sure that it’s just accurate documentation. The goal is not to just bump up those numbers and that risk adjustment factor. It’s to capture the patient diagnosis in a way that makes it accurate documentation that supports each of those adjustment codes that are entered on that claim for that visit.

Kyle Enger: Yeah, that’s a great point. The documentation and coding needs to be as accurate as possible. And sometimes diagnosis codes are used that are more complex than the patient’s actual conditions, that’s called upcoding.

It’s bad practice, it’s unethical, it can even harm the patient. As we’re increasingly sharing medical information among more providers, wrong information in a patient’s medical record could lead to unnecessary tests and procedures, and that leads to unnecessary risks and expenses for the patient.

So this is obviously a problem and that means that health plans are highly motivated to catch and correct upcoding when they audit providers. One other thing to remember is that for risk reporting is that diagnoses have to be reported every year in order to count for risk adjustment.

Sometimes when we explain this to providers, they don’t like this because obviously conditions like COPD don’t go away, amputated limbs don’t grow back, et cetera. So why do we need to keep reporting them year after year? The explanation for that feeds back into MEAT. Remember, a condition has to be monitored, evaluated, assessed, or treated in order to be counted on claims. And if none of that happens for a certain condition, even if the condition technically exists, then probably it’s mild, well-managed, or sometimes it might even be cured. And then it makes sense that it shouldn’t count toward the risk score.

So to build on the example of a patient with an amputation. If the patient is doing well with a prosthesis and doesn’t need any help with it, then MEAT probably didn’t happen and amputation status, therefore shouldn’t be coded.

Celina DeFigueiredo-Dusseau: Great and of course these are the questions that each of you are asking clients every day on the job, and it all lends to the need for really efficient and effective operations and processes across the board in practices. So I’m curious to know what are some of the other risk adjustment optimization solutions that your team offers at here at Medical Advantage?

Kyle Enger: Yeah in many of our standard reports for providers, we include risk score information, and this includes calculating risk scores, using the most up to date claims on our data warehouse so that providers can have a clear and current picture of their population’s risk scores at any point during the year.

And this is really helpful because final risk scores in reports from payers can be reported months or years after the fact. So having risk scores calculated on the fly from claims data is much more timely. And lets practices see how they’re doing over the course of the year on risk reporting and, and correct during the year as as necessary. Also, our electronic health record consultants can help you get the most out of the risk adjustment features in a variety of different EHRs.

Angela Hale: And just to elaborate a little bit, you know, our team’s made up of many experts in healthcare and we offer a wide range of services because we have so many years of experience among the team.

And that ranges, these services range from educational webinars, the HCC coding tips, chart audit for many of the different users such as the RAVD audits that we spoke of earlier. And of course, we have the actual boots on the ground, the in-practice support. We have experts in analytics, nursing care management billing and coding, provider documentation. We have our EHR optimization team and really everything in between. We really pride ourselves in being able to meet practices where they live. This means that we have a solution that fits the client.

We’ve got very large physician groups that we work with, as well as the small solo practitioner. And we can make it work with those large groups to develop a culture of value-based care and programs such as the MACRA and MIPS programs. And then we also can assist with the newer telehealth problems that have come around because of the pandemic and COVID and the ongoing risk payment pitfalls. We have a vast team with a lot of knowledge and we can help where there’s limited staff to meet the ever-changing needs of staying on top of the healthcare market. And we can develop solutions that work even in the smaller offices as well.

Celina DeFigueiredo-Dusseau: Fantastic. Thank you both so much for sharing your expertise and teaching us a little bit more about risk adjustment. I can’t agree with you more. We have such an incredible team of experts and professionals all across the board and risk adjustment falls right in there. Really making sure that medical practice operations are effective and efficient and really work for the providers and the patients alike.

This has been so incredibly informative and of course for our listeners, if anybody would like more information, I would like to invite you as our audience to find us on LinkedIn and Facebook or our website medicaladvantage.com, where you will find other webinars, podcasts, and blog articles on risk adjustment optimization and many other topics in healthcare.

And should you have any questions or wish to find out more from Kyle or Angela please feel free to reach out to us directly at info@medicaladvantage.com and we would be happy to answer any questions or additional information that you might like. So thank you both for coming on today and we hope to have you back as guests soon.

Angela Hale: Thank you, Celina.

Kyle Enger: Thanks very much. I was glad to do it.

Medical Advantage Podcast: Thanks for joining us this week on the Medical Advantage Podcast, where we discuss the ideas and technologies changing healthcare and what they mean to your organization. For more information, visit us at medicaladvantage.com and make sure to subscribe to the podcast on iTunes, Spotify, or wherever you get your podcasts, so you never miss a show.