Transcript of “The Medicare Money Machine and the Risk Adjustment Score Game” Conference Call with Dr. Rick Gilfillan

Jan 28, 2022

On January 12, The Capitol Forum hosted a conference call with Dr. Rick Gilfillan, who served as the Director of the Center for Medicare and Medicaid Innovation from 2010 to 2013, to discuss a two-part series of posts on Health Affairs titled “Medicare Advantage, Direct Contracting, and the Medicare ‘Money Machine.’” The full transcript, which has been modified slightly for accuracy, can be found below.

MS. LISA EPSTEIN: Good morning, everyone. Thank you for joining The Capitol Forum’s conference call on “The Medicare Money Machine and the Risk Adjustment Score Game”. I’m Lisa Epstein, Correspondent at The Capitol Forum. And I am joined today by Dr. Rick Gilfillan.

A quick note before we get underway. The first 20 minutes or so of the call will be an interview with Rick. And then we’ll move to a Q&A format where we will entertain questions from the audience. If you do have questions for us, please email them to That’s And Capitol is spelled with an “o”. Thanks for joining us and let’s drive right in.

So today, as I said, we are pleased to be joined by Dr. Rick Gilfillan. Dr. Gilfillan began his medical career as a family medicine physician in rural Massachusetts. And he has worked in the health care system in for profit companies, nonprofit organizations, and government agencies. In 2010, he was appointed by the Obama administration to serve as the first Director of the Center for Medicare and Medicaid Innovation at CMS, where he developed a new team that worked with payers and providers to develop innovative models for improving patient care and reducing costs.

Prior to and after his appointment at CMMI, Rick served in a variety of senior roles across the industry. Last September, he coauthored a two-part series of posts in Health Affairs that criticized CMS for overpaying Medicare Advantage plans with no demonstrable clinical benefits to patients. The posts also criticized the direct contracting Medicare pilot developed under the previous administration by CMMI, the department at CMS formerly headed by Rick. So Rick, first let’s talk about the high-level takeaway that you want people to get from your posts and your criticisms of CMS.

DR. RICK GILFILLAN: Well, thank you, Lisa. And thanks to The Capitol Forum for having me on this webcast today. First of all, I wouldn’t say it was criticisms leveled against CMS. Just to be clear, I think what we tried to do was reveal to a wider audience the actual underlying drivers of the Medicare Advantage space and the associated direct contracting model.

And the takeaways that we’d like people to get is that (1) Medicare Advantage is actually a subsidized program. That is the subsidies are paid for by taxpayers and by Medicare beneficiaries to the tune now, we think, of about $20 billion a year in extra payments, by some estimates. And that subsidy is funding the rapid growth of Medicare Advantage. Medicare Advantage, we would say, has over thirty-five years never cost the same or less than the traditional Medicare program. And we wanted to just kind of get that information out and make it clear exactly how the Medicare Advantage marketplace works. And with that, make the case this is not a model to build a public option on, we believe. So that was the main takeaway on the Medicare Advantage side.

On the direct contracting side, the takeaway is this model is introducing the same entities that raise costs in Medicare Advantage into the traditional fee for service space, with the potential for actually enrolling even more people—or I should say aligning many more people—with Medicare Advantage plans and direct contracting entities owned by them. And so our concern was to make this clear as a threat, frankly, to the overall Medicare program and the cost of providing Medicare coverage to Americans.

MS. LISA EPSTEIN: Okay. So is that what you meant when you said the Medicare gold rush?

DR. RICK GILFILLAN: The Medicare gold rush actually is related but different. It turns out that Medicare Advantage has been so profitable through the risk-score gaming arbitrage we talk about, that it has made Medicare Advantage plans, and practices that focus on MA very attractive from an investment standpoint. Many of the direct contracting entities are the same firms that have benefited in that Medicare Advantage world.

The gold rush then is actually the rush of private equity, venture capitalists and SPAC entities into this space in pursuit of the opportunity that they think is there because Medicare is growing so rapidly. It’s now almost $1 Trillion in spending annually. It’s moving to about $1.25 trillion annual spending in the next five years. And the size of that pool and the profit opportunities it presents are such that there’s a gold rush of private financing rushing into this space, backing Medicare Advantage plans, backing primary care startup entities and backing analytics firms which are built on the risk coding efforts of Medicare Advantage players. So the gold rush is about private money rushing into this space to try and capture a piece of it.

MS. LISA EPSTEIN: Okay. So part of that, you talk about price per patient life and that there’s disparities between the traditional fee for service Medicare price per patient life and the Medicare Advantage price per patient life. What did you mean by that?

DR. RICK GILFILLAN: Let me be clear. So historically, when someone was buying a Medicare Advantage insurer, the price per person life varied $4,000 earlier on to $10,000 more recently.

What’s happened in this gold rush is these new entities are being funded at multiples of that, some as high as $150,000 per life. And the average valuation that we found across the different entities we looked at was about $87,000 per patient life. So marked way up, very much increased valuations being driven by the expected profits, the expected growth of the marketplace, and I think, frankly, some over enthusiasm and lack of understanding of the underlying realities of the MA and direct contracting space.

MS. LISA EPSTEIN: And what is the price per life that you calculated for traditional fee for service Medicare?

DR. RICK GILFILLAN: Well, let’s distinguish the price per life versus the total cost to Medicare for providing care. We don’t price traditional Medicare lives in the same manner because membership is typically not being bought and sold as they are in the MA world. So let me just clarify, when you make that statement, which you’re talking about?

MS. LISA EPSTEIN: Well, you said $87,000 was the average price per life in the Medicare Advantage space. So what do I compare that to as far as the traditional Medicare beneficiary?

DR. RICK GILFILLAN: Well, there is not much to compare it with. You could ask, well, how do ACO lives get valued because they typically have some sort of population of traditional Medicare beneficiaries who are aligned with it. And it’s not as easy to value those lives because there have not been many direct acquisitions of ACOs. But you can see actually in some of the numbers we’ve looked at, the range of valuations of these firms follow their mix of Medicare Advantage lives, which are much more highly valued vs. their ACO lives. Looked at that way I think it’s probably on the order of five times more for MA or something on that order, but it’s not something that’s well-documented. And the real comparison for our purposes we used was the historical valuation of those Medicare Advantage lives.

MS. LISA EPSTEIN: Okay. And then the biggest factor in driving this gold rush is the CMS overpayments because of increased coding of diagnoses, high paying diagnoses, for the risk score gaming?

DR. RICK GILFILLAN: I would say there are several factors. One is the growth of the marketplace. As I mentioned, it’s going to grow tremendously. Two, there’s a variety of arbitrage opportunities, and the risk score gaming is a big one. And in fact, it is guaranteed profit. That’s why we call it arbitrage.

Other arbitrage opportunities that people take advantage of include cherry picking counties to get the best rates, and picking doctors who seem to have lower costs for their population. All these make it a lot easier to be profitable.

And then the third element is the addition of direct contracting as a vehicle for getting access to the other half of Medicare spending. If you think of total Medicare spending, if you said it’s on the order of almost $1 Trillion in 2022, about half of it almost is in Medicare Advantage and the other half is in traditional Medicare. We haven’t seen the numbers yet, but somewhere around 45 percent or maybe even more will be in Medicare Advantage. Well, if I’m an investor and I’m investing in an MA plan, I only see myself as having access to half of that Medicare spend. If I now say the direct contracting program makes it possible for these same entities to get lives and revenue and income from the traditional half of the pie, as it were, now my company looks much more valuable. So it’s that addition of direct contracting to the Medicare Advantage opportunity that is the third major driver of this gold rush.

MS. LISA EPSTEIN: Okay. So let’s back up a little bit. And for listeners who may be unfamiliar with what risk adjustment scoring is and how that can be gamed, how about you just give us some of the basics of how that works and how that’s so lucrative for the Medicare Advantage firm?

DR. RICK GILFILLAN: Okay. Well, let’s take two populations of patients. Let’s think about you have two Medicare Advantage plans. And they each have let’s say 100,000 lives. Well, the illness burden of those populations may be different. And so CMS, when they pay, they want to make sure that they’re paying appropriately for the illness burden of a population. So what has been done historically is they try and adjust the payment rate for the risk of the population. So if an average population is a risk of one and CMS pays $1,000 a month, then my risk adjusted payment to them will be one times $1,000 per person per month or about $12,000 per year.

Now, if I say that population is 20 percent sicker, then their risk score would be 1.2. I would multiply that times that $1,000 a month. And instead of paying $1,000 a month, I would pay them $1,200 a month or about $14,000 a year. So there’s this dramatic potential increase in payments based on how sick the population appears to be.

So then the question becomes, well, how do you measure how sick a population is? And the way CMS has done it since 2006 is, they say, well, I’m going to look at the diagnoses that this population has and create a risk score, for every person in the plan that will predict how much more or less than average it will cost to provide the care they need. The average person will be 1.0. Some will be not as sick as average, so they’ll be less than one. And some will look sicker and they’ll be 1.2 or something like that. Using those scores, I will calculate a payment for each individual by multiplying the risk score times the average cost. The plan receives that total amount for each

patient across their population. We can also see what an overall average for the plan is to compare across plans.

So how does CMS do that? Well, they’ve developed a sophisticated program called the HCC Risk Adjustment System. HCC stands for Hierarchical Condition Categories. And what happens is all the diagnoses that are submitted by providers get put into a software program. And that software program takes all the diagnoses and computes what HCCs a person may have. If they have diabetes without complications it would be one HCC. Depending on what codes are submitted, they may actually look like they have diabetes with complications. Or maybe they have heart disease, it could be mild heart failure or severe heart failure. CMS has calculated the increase in medical costs associated with each of those categories. Each HCC has a coefficient that predicts how much more it will cost for a patient with that HCC.

So they take all the diagnoses. They calculate the HCC’s that each person has. and then they add up those coefficients to get a risk score for each patient. They multiply that risk score times the plan’s average cost per person. That becomes the amount that they pay to the MA insurer for that patient. Obviously, the coefficients and the payments are more for the complicated categories, hence the incentive to submit more codes.

So here’s the wrinkle that makes this a little bit complicated. The HCC model is built using data from the traditional Medicare fee for service experience. How did they get diagnoses? Well, typically in the Medicare traditional fee for service world, the diagnoses from come from claims that providers submit. They come from the diagnoses they put on claims. Well, doctors don’t put every diagnosis on a claim. When we see a patient, we typically want to put one diagnosis down so we can get paid, and that’s it. And so the result of that is the traditional Medicare database has a lower number of diagnosis codes for the population. And as a result of that, when you compute how many HCCs are in that population for every person, it’s lower because there aren’t as many diagnoses.

And then if you ask how much is each of those HCCs worth? Basically, what they do is they take the total expenditures in Medicare and assign them to HCCs. Simplified, they divide the total dollars by the total number of HCCs in the population and come up with let’s just call it an average HCC value for each of those HCCs. In doing that, because the FFS data is incomplete, the number of HCCs is actually low so the HCC coefficients are inflated.

Now, let’s go to the Medicare Advantage world. Medicare says to me, I’m going to pay you based on how sick your population is. I’m going to judge that based on how many HCCs each person has. And I’m going to decide that based on the diagnoses of your patients. Now MA plan, you can either send me diagnoses from the claims you get from providers, or you can send a nurse out to people’s

homes to do health assessments that identify diagnoses, or you can do chart reviews to find and then report more diagnoses. You just have to document that they’re real and that they were attended to each year.

What does that do? Well, that says to the MA plan find and submit every last diagnosis you can if you want more revenue. I’m going to pursue every way I can to get more diagnoses because the more diagnoses I put in, the more HCCs it will generate, the sicker my population will look and the more Medicare will pay me. Remember, they use those same inflated HCC coefficients that they derive from the Medicare fee for service dataset. So I have more HCCs per person, and each HCC is valued higher than it should be. The more codes, the more revenue and at the same time the patients are not sicker, not needing more services or having higher costs. So risk score gaming is basically the MA plan optimizing its coding and risk scores to get more revenue it can use to be more profitable, offer better products and grow.

MS. LISA EPSTEIN: Okay. So I have some questions on that. So CMS has all of this data. As you said, they run all these diagnoses through their software to generate the HCC codes. So with all the data, can’t CMS tell which MA plans are engaging in overly aggressive coding and then regulate this or prevent those entities from maybe getting direct contracting agreements to move onto that program?

DR. RICK GILFILLAN: Well, here’s the thing. First of all, let me be clear. This is a fundamental business reality. When MA was developed, people kind of knew there would be a potential for gaming. And by gaming, we are not saying this is fraud. If a diagnosis is real, this is not fraud. This is just a loophole, if you will, in the way the program was established. And Congress knew that this was happening and they established a corrective mechanism for it. They said, well, we’re going to create a Coding Intensity Factor (CIF). We know that coding is going up much more rapidly in Medicare Advantage than it is in traditional Medicare. So we’re going to give CMS the power to decrease MA risk scores to compensate for this. And so you can imagine that if they decreased the scores the plans got by the amount that they were going up, then they could correct this. But what happens is CMS has not done that. Congress mandated they had to increase it to a minimum of 5.9 percent. And they did that in 2018. But CMS has not continued to increase that number and meanwhile, scores have increased dramatically more.

So yes, there’s a mechanism to correct it. Politically, it has been hard for administrations to do that. Because the Medicare Advantage world is very good at raising the specter of, oh, we’ll have to increase premiums. We’ll have to decrease benefits. People are going to be upset. And when does open enrollment season occur? When do people see what their benefits and premiums are going to be for next year? In November of every year. And every two years, that coincides with an election. So the result is that every administration has been reluctant to do that.

So let me just make one other point. There’s been evidence of fraudulent coding activities in the past. People were prosecuted and the most egregious practices have stopped. And I think that today most of these activities are not fraudulent. They’re codes that are probably accurate in some manner. Although, people probably stretch that to some extent. But most of this activity is not fraudulent. It’s just playing the game that they’re presented with and playing it very hard. And some people are much better at it than others. You can see some firms have higher risk scores than other firms. Because they work it more and they find ways to incent providers to do more coding on their behalf, to submit more of those codes. In some cases, they actually have providers going out and ordering studies that are not indicated to get more diagnosis codes. But I just want to be clear, most of the overpayments are not like fraudulent. It’s just playing the game that they’re presented with and playing with it harder, more or less effectively. The overpayments are not about “bad actors”, it is about how virtually every MA Plan aggressively pursues RSG. Some are just better at it than others.

MS. LISA EPSTEIN: Okay. So that plays into what you were discussing about this coding intensity factor. So let’s just talk about that for a bit. The coding intensity factor, if I understand correctly, means that if you have beneficiaries in Medicare Advantage plans, the Medicare Advantage plan is highly incentivized to find every possible illness or diagnosis that that beneficiary has and document it so that they get more money from CMS. And that’s what you mean by coding intensity?

DR. RICK GILFILLAN: That’s correct.

MS. LISA EPSTEIN: Okay. So then the coding intensity factor or discount is applied and you said about five percent, but it hasn’t changed in many years. So what you’re saying by that discount, just explain that discount or the coding intensity factor and how that actually works.

DR. RICK GILFILLAN: Right. And just so I’m clear, what happens, the way the Medicare Advantage average rate for a plan is set is the result of a bidding process. The Medicare Advantage plans submit a bid that says, CMS, I need this amount of money to cover the benefits I’m suggesting. And by the way, that’s less than what it costs you in FFS. And so Medicare says, well, if you’re saving me money, I will keep 1/3 of the savings, but I’ll give you 2/3 of the difference back as a rebate which you must use to provide lower costs or better benefits plus your profit. And the result of these risk scores going up is that the rebate becomes much larger and the rebate is used to provide better benefits or lower premiums for the Medicare Advantage folks.

So if I submit my Plan risk score of 1.2 instead of one, I have a lot more premium to play with to provide better benefits, and to get more profits, frankly, out of the system. And we describe that in

our blog. So if my score is 1.2, what CMS does is take that down a bit using the coding intensity factor. I’m going to decrease and take your 1.2 and subtract the CIF of .059 from it. So I’m going to take you down to almost 1.14. They use that to calculate the premium and rebates. So you have less money coming to you. But if the reality is that I’m 20 percent above where I should be, just taking it down by or 5.9 percent doesn’t really eliminate the overpayments.

MS. LISA EPSTEIN: And are you proposing or do you have a proposal to suggest CMS increase that? And if so, to what percent do you think would be fair?

DR. RICK GILFILLAN: Well, I think fair would be to make it neutral with traditional Medicare and would be to eliminate the overpayments. Rick Kroenig, who is one of our coauthors on a recent paper, thinks now the difference between traditional Medicare and Medicare Advantage is running in the range of .19 to .2. Meaning after you net out that factor, it’s about a .14 overpayment. So totally neutralizing it would be to put it at .14. Could CMS do that tomorrow? Yes, they have the power to increase the factor. But would they do it? No, because it’s a political decision, as I mentioned, people are worried about the potential for it causing MA plans to increase their premiums and decrease their benefits.

So unlikely to be done all in one fell swoop like that. It needs to be done gradually over time. We have a proposal, actually, that was in our Health Affairs blog on Monday, which I sent to you, which basically says what they should do at a minimum right now is just freeze the differential at the current level and index that coding intensity factor to the annual change in MA risk scores versus traditional risk scores. And that was running about one percent. Now it’s running about two percent a year.

So our proposal was that they index it and thereby freeze the overpayments at the current level, but not allow them to increase going forward. So that’s our current proposal. And our longer term proposal is that basically they change the risk adjusted methodology, not leave it in the hands of plans or providers to create those risk scores, but rather create an objective, independent way of determining the relative risk of the different populations. And we have a proposal in our paper to do that.

MS. LISA EPSTEIN: But could CMS implement that piece on its own through rulemaking?


MS. LISA EPSTEIN: Or would that require an act of Congress?

DR. RICK GILFILLAN: Actually, it could be done through rulemaking. Again, it would be a political decision. And I think frankly, there’s enough conversation about this right now, enough attention being paid to this, that I’m hopeful that we are going to find our way to a different kind of risk adjustment system going forward.

MS. LISA EPSTEIN: One that would eliminate the gaming tactics and practices?

DR. RICK GILFILLAN: Yes, one that would make it impossible to continue these kinds of gaming risk scores and giving plans the ability to control their revenue. If you look at the OIG, the Office of the Inspector General, they did a review of 2017. They documented that about $9.2 billion in spending that year went out based on diagnoses that were not documented any place other than in chart reviews done by plans or nursing visits to the home for health risk assessments that were done by plans. And by the way, the vast majority of that was done by 20 plans out of, I think, 150 or so total number of plans. So the big part of this is being driven by those that are most aggressive coders.

MS. LISA EPSTEIN: Okay. And then how did this leak onto the fee for service beneficiaries, implicating the direct contracting program? How does that play into this?

DR. RICK GILFILLAN: Well, let me make one more comment on MA. For every dollar in excess payment to these plans, Medicare Part B beneficiaries, everybody who has Medicare Part B, actually paid for about 15 percent of those extra payments. And let me explain why. About 58 percent of all Medicare payments come from Part B, which is the outpatient side of Medicare, vs. Part A which is the inpatient side. Everybody who has Medicare gets Part A for free, but they have to pay premiums every month to get Medicare Part D coverage. And last year, they were about $150 a month.

Well, every dollar that Medicare allocates to Part B ends up being part of the calculation in terms of what that premium should be. So, all of the Medicare Advantage spending, every extra dollar, 58 percent of that gets allocated into the Part B fund and premium for patients for beneficiaries has to cover 25 percent of that expenditure. So simply put, 25 percent times 58 percent is about 14.5 percent of every extra dollar that’s actually paid out of the pocket of Medicare beneficiaries, whether they have Medicare Advantage or whether they have traditional. Medicare beneficiaries are actually paying for the subsidization of Medicare Advantage.

The direct contracting program provides some, but not all, of the risk coding opportunity of Medicare Advantage. And we estimate that there’s probably like 20 to 40 percent of the same opportunity to be gotten in direct contracting. And we think that will ultimately cause an increase in spending on the traditional side for coverage of direct contracting beneficiaries.

MS. LISA EPSTEIN: Okay. So basically, that boils down to everyone on Medicare that has either a Medicare Advantage plan or both Part A and Part B, they are paying for the money to flow to CMS and then to the Medicare Advantage plans. Or as far as direct contracting for those direct contracting entities. That’s what it boils down to.

DR. RICK GILFILLAN: Yeah, every dollar that goes out in overpayments, Medicare Part B beneficiaries – 57 million people—pay 14.5% out of their pocket. Taxpayers, of course, pay the rest.


DR. RICK GILFILLAN: And these overpayments now, I think, ballpark our estimate—Rick Kroenig’s estimate, as I said—is about $20 billion a year. That’s going to increase over the next eight years. That number will get close to $100 billion annually, in his estimation, if we do not do something to stop the escalation in risk scores.

MS. LISA EPSTEIN: And didn’t, in 2022, the Part B premium went up kind of significantly, right?

DR. RICK GILFILLAN: Yes it went up to about $170. Of course, a significant part of that, some of that, is related to this risk coding. But some of it is also—a large part of it—was related to the anticipated new medications for Alzheimer’s Disease that now look like they’re going to be more limited in coverage. And so I think the administration has asked CMS to go back and take another look at what that Part B premium should be in 2022.

MS. LISA EPSTEIN: So besides Medicare Advantage plans and the direct contracting program, are there other types of CMS programs, maybe the PACE program or any other programs, that tailor CMS payments based on risk scores?

DR. RICK GILFILLAN: Yes, the PACE program is, I consider, part of this whole thing. And while we didn’t write about it specifically, it provides a similar risk score arbitrage opportunity. And I think that’s a significant part of the reason why for-profit PACE has become a much more active space from an investor standpoint.

MS. LISA EPSTEIN: Okay. Let’s talk about the direct contracting piece. Do you see that pilot program playing out for the next, I think, four or five years and then terminating? Or do you think that’s an open question? Do you think that CMI might tweak the program after getting input from people like yourself with concerns and suggestions for reform? What are your thoughts on that?

DR. RICK GILFILLAN: Yes, I think the common wisdom is that the current team is looking at the direct contracting program and thinking about what they want that to be in the future. And I think there’s a need to have some sort of an advanced ACO program for ACOs that are attempting to do, in the traditional Medicare space, attempting to provide high value care. So the next generation ACO model stopped at the end of last year and those folks need a place to go. And right now some of them are going into the direct contracting model in 2022. We don’t know how many. But the guess is, I think, that they will modify the current program so that there’s an opportunity for those folks to continue in some sort of advanced track. What that will look like, nobody knows at this point. And our recommendation was that they stop it for 2023 and create a new ACO program that would actually adhere more closely to some of the other historical approaches in the ACO world, but also provide more advanced opportunities and possible capitation for ACOs.

DR. RICK GILFILLAN: The irony of this is what happens as a result of the DCE program is that both DCs and ACOs are reliant on or dependent on primary care doctors to participate, to generate membership, to have beneficiaries be aligned with them.

MS. LISA EPSTEIN: Just have to jump in real quick, ACO is the Accountable Care Organization. And that’s just a primary care provider practice or manager of those types of practices?

DR. RICK GILFILLAN: Yes, basically an ACO is a coming together of providers, many of them, most of them often, being primary care doctors. And CMS takes their claims experiences and says, well, based on visits to you in our claims, it looks like all these patients are seeing you regularly. So we’re going to “attribute” these to you as your patients and your population. It’s a population-based program in that CMS creates a medical expense target and the ACO tries to provide excellent care, decrease costs and beat that medical expense target. If they do that and improve quality they can get additional incentive payments.

And there’s been, I would say, clear savings. Although, not as much as we all anticipated. There’s about 10 million people who are covered under the ACO program today in Medicare and many more in the commercial and Medicaid spaces. So that’s a kind of growing initiative as well. And Medicare has seen a very low trend over the past eight years in the traditional book. We think some of that is attributable, not all, but some, to ACOs. So ACOs are kind of like competing with direct contracting entities to get primary care doctors to participate so they can get membership.

And the irony of this is Medicare went out and basically said, let’s get all those folks who are increasing costs in the Medicare Advantage world and bring them into the traditional FFS Medicare using the Direct Contracting model, in a world where we have ACOs, which actually have saved almost $10 billion over their first eight years. So we have one program that cost $140 billion over the last 12 years more than fee for service, and we have one program that saves about $10 billion

over the last eight years. And the prior administration, bringing all those people that are costing more into traditional Medicare. Both programs rely on PCPs who have to choose, they can only be in one or the other. That’s one of the reasons for our concern with the direct contracting program.

MS. LISA EPSTEIN: So one complaint that I’ve heard about, which I assume is about the ACO model, is beneficiaries who are in an ACO, they may not understand that they’re in an ACO. But they see that a doctor is participating in their Medicare Advantage plan. But when they try to go see that doctor, their primary prefers to give them a referral to a doctor in the ACO organization. And then the beneficiaries don’t really understand why they can’t see another participating provider, and they’re being kind of steered toward, let’s say, the cardiologist participating in the specific ACO. Have you heard of issues like that? And what are your comments on it?

DR. RICK GILFILLAN: Well, I think of it this way. In the Medicare Advantage world, a patient sees a network and says, okay, that’s going to be my network. If I sign up for that plan, that HMO plan, for instance, in MA, I know there’s a limited network and I’m only going to have coverage if I go to that network. So the doctor referred me within that network. In the ACO world, the patient continues to have the same Medicare benefits that are available to them, regardless of which doctor they go to. As long as they’re participating Medicare doctor, they can go to any one of them.

Now, the whole point of an ACO is to say, gee, can we coordinate this person’s care, and thereby coordinating the care eliminates some of the fragmentation of care we see in the traditional space if we do a better job we can deliver better quality, lower costs, et cetera. So yes, there’s an opportunity, there’s an effort, to create networks to say, look, I think Mrs. Jones, I’m referring you to this cardiologist. I think she’s good. She works with me closely. We’re both on the same medical records. So she can see your information. I can see your information, et cetera.

But in the ACO world Mrs. Jones can still go to any other physician that is a Medicare participating physician. So I haven’t heard a whole lot of concern about access in the ACO world. It’s possible that happens. And I should be clear, also, the DC program, the direct contracting program, is kind of a hybrid of MA and ACOs. And under the direct contracting program, the patient does still have the opportunity to go wherever they want.

MS. LISA EPSTEIN: Okay. So going back to the risk score, you said that these scores aren’t necessarily fraudulent. It’s just coding intensity. We’ve done some reporting on some questionable practices that the codes may not actually be reflective of a beneficiary’s true health condition. Or maybe they inflate the condition to a more serious form. And in those cases, what impact would those diagnoses have on beneficiaries as far as maybe getting life insurance in the future? Or should the beneficiary wish to switch back to traditional Medicare with a supplement and a premium for

the supplement, the Medigap policy, could those diagnoses impact the rates or ability to qualify for a Medigap policy?

DR. RICK GILFILLAN: Yes, that’s a great point, Lisa. I’m not sure how relevant life insurance is quite honestly in this population, but you’re certainly right in terms of Medicare supplemental insurance. With Medicare Supplemental, in most states, you have one shot at getting into Medicare supplemental coverage without underwriting. That’s when you become eligible for Medicare. A small percentage of MA members, less than 5%, try to go back to traditional FFS Medicare because, among other reasons, in most states you actually can be medically underwritten if you want to return to traditional Medicare. That is, they can deny coverage, exclude preexisting conditions, or adjust your cost depending on your historical medical experience. And so preexisting conditions come into play. And if my chart now, instead of having one diagnosis of high blood pressure and nothing else, and now suddenly my chart has emphysema, COPD. I have congestive heart failure. I have diabetes with complications or something. Yes, I’m going to look very different, frankly, to those folks who are doing the pricing for that Medicare supplemental insurance plan. Some of them would be accurate. Some of them, as I’ve said, there are gray areas where people will, as you found in some of your investigations I think, there will be folks who have diagnoses that they’re not even aware of.

MS. LISA EPSTEIN: Right, right. I think that’s a concern that really hasn’t been talked about much, that I wonder of the impact on the individual level.

DR. RICK GILFILLAN: Well, what happens is, in all likelihood, for all but the most well to do, that will end up driving them back into Medicare Advantage coverage. And truth is for all but the very well to do, choosing Medicare Advantage is a one-way street in most places.

MS. LISA EPSTEIN: Once you’re on Medicare Advantage, you’re on Medicare Advantage period.

DR. RICK GILFILLAN: It’s hard to go back. I mean, go back the other way. But you have to have the funds to afford Medicare Supplemental. And you also in many states, not all states, but in many states, you also are subject to, as I’ve said, the preexisting condition evaluation that would end up increasing your premiums significantly. Or just denying you getting coverage.

MS. LISA EPSTEIN: So, several federal agencies have raised concerns or written reports or had a settlement with firms based on risk score activity, risk score gaming activity, CMS, HHS OIG and the DOJ. Do you see these type of regulatory investigations and actions increasing? And could that cause the industry to kind of pull back on some of their practices?

DR. RICK GILFILLAN: Well, I think the OIG and others have called upon CMS to change the regulations and say, well, you should not count home nursing visit diagnoses in your risk scores. Or you should not include chart reviews. You should just rely on submitted claims from providers. So it’s possible over time that those kinds of changes could be put in place. I think I guess that’s—maybe I’m being overly optimistic here—but I think it’s also likely that we will get to different kinds of risk score programs, different ways of calculating risk scores, that will be gradually introduced and eliminate the potential for folks creating these elevated risk scores just by gaming it and looking for more and more diagnoses.

MS. LISA EPSTEIN: Okay. So in one of your posts that I’m referring to from last September, you mentioned coding shops. And do you see the business model of coding shops starting to be eroded as these gradual changes to risk or methodology are introduced, should they be introduced.

DR. RICK GILFILLAN: I think so. But also, we did describe what we call the money machine, the Medicare Advantage money machine. And basically, if you think about this coding opportunity, as I’ve described it, if I’m a managed care plan or Medicare Advantage plan, I go to a provider and I say I’m going to do “value based care deal” with you. Although, they’re not really value based care because they increase costs. But they call it that and they say, I’m going to give you a deal here, doctors. I’m going to give you a medical expense target. And all you have to do is beat that and you can get the savings. Oh, and I’m going to create that target using the premium that I get for your members. And I’m going to set the target as a percentage of that premium. If you’re getting $1,000 today, I’m going to set a target of 85 percent and your medical expense target is $850.

But by the way, if you increase your premium and increase your risk score, thereby increasing your premium, I’m going to increase that target. So if you save a little money on medical expense, if your target is $850 and save $20, well, that’s $20 per person per month extra. But if you increase your revenue by $100, that is you increase your risk by .1, by giving me, the payer, more codes, my target will go up by 85 percent of that or $85 per person per month. So you can increase your incentive from $20 to $85 plus $20 or $105, five times by just increasing your risk score.

These percentage of premium contracts create what we call a money machine. This is just more codes, more payment, more profits for the plan, more profits for the provider. And it just goes round and round and drives more and more. Those kinds of deals are what is driving the gold rush on the primary care side. Virtually all the entities that are MA only operate under such contracts with an opportunity to drive that kind of excess payment. That’s what’s driving the whole valuation in the primary care sector of the gold rush. I think that, in addition to trying to change the risk adjustment model, we also suggest that CMS should eliminate the ability of MA plans to do those kinds of contracts with primary care practices.

MS. LISA EPSTEIN: Okay. And what you’re describing is almost a description of the direct contracting program, if I understand correctly.

DR. RICK GILFILLAN: No, there’s not—

MS. LISA EPSTEIN: Percentage of premium contracts, okay.

DR. RICK GILFILLAN: Yeah. The thing that’s different is the risk or opportunity. The opportunity to drive the risk scores higher is more limited in direct contracting. In MA this has doctors ordering ultrasounds of the carotid artery, arteries in your neck, trying to find evidence of a plaque so they can submit it and increase their payments by $2,800 per person per year. Just by demonstrating that there’s a plaque in the carotid artery. These tests are not recommended. In fact, it’s recommended you don’t do them. But if they do those tests and demonstrate that plaque, they submit the code. The plan gets more money. The payer, the provider, gets more money as a result.

MS. LISA EPSTEIN: Okay. And then I have one question from the audience, and then we will let you go and thank you so much for all your time. Is there any way to find out who these primary practices are that have the percentage of premium contracts? Do they need to be filed with CMS? Do the MA programs have to disclose that information?

DR. RICK GILFILLAN: They don’t have to disclose the information to patients. I am not sure if they have to report them to CMS. I don’t think so. I don’t believe that’s the case, but I honestly don’t know that for sure. Here’s one guidepost. If you read about a primary care group and it says it only does Medicare Advantage or it primarily does Medicare Advantage, you can bet that they’re operating under that kind of a contract. Every time you read focused on Medicare Advantage, you can assume that, generally speaking, they’re going to be operating under a contract that’s similar to what I’ve described. And by the way, if you look at the investor roadshows for all these primary care firms that went public over the past two years, you’ll see this story is told loudly and very clearly and directly in the slide deck.

MS. LISA EPSTEIN: Okay. And as those firms see the patients, gather diagnoses, bring in more money for the Medicare Advantage plan, they get to share a percentage of that higher payment from CMS, that capitation payment.

DR. RICK GILFILLAN: Yes, ballpark 85 percent of it. And so here is one more part of the story. Look at that money we are talking about. These incentive payments could be $100 or up to $200 per member per month in profits. This is how you get the billions of dollars in valuations or MA focused providers. That’s a lot of money. So if you’re a payer, you’re a Medicare Advantage plan, and you do this deal with some providers and you say, oh my goodness, my profits are $50, $60

per person per month, and that provider is making $120, $150, $200 per person per month. The provider’s making more than me, the payer. What am I going to do?

Well, what did UnitedHealth do? Go out and buy the provider. That’s how Optum becomes the largest or one of the largest employers of primary care practices in the country. They go out and buy these groups. And now they have the opportunity to benefit from both the insurance profits and the provider profits under the money machine. And by the way, in doing so, it’s quite possible that they’re not meeting that 85 percent loss ratio required under Medicare Advantage. If they include the incentive payment to their primary care doctors they own as a medical expense in their medical loss calculation, they may look like they actually have an 85 percent loss ratio. But if you really think about where those dollars should be, if those dollars are actually translating into insurer profits, they should not be in the top in the medical expense numerator. They should be in the denominator in fact. And you would have a medical loss ratio that’s much lower than 85 percent.

MS. LISA EPSTEIN: For insurer owned primary care practices?

DR. RICK GILFILLAN: Yes, correct. And for their MA plans if it’s a big enough number. Now, we don’t know that they’re doing that. But MedPAC raised this issue in their last report. And I think at this point, it’s a little bit early to know exactly what is going on in this space. But that is a concern that has been raised.

MS. LISA EPSTEIN: Got it. Well, I want to thank you so much for your time, Rick. I know this is very interesting to me and I’m sure to our subscribers and listeners. And I want to thank you for your time. And that’s it. Have a nice day. And everyone listening, have a nice day.

DR. RICK GILFILLAN: Thank you for the opportunity, Lisa.