 Great. All right. Our next speaker is from the Centers for Medicare and Medicaid Services. Steve Furrow, please go ahead. Good morning. I'm Steve Furrow. I'm a physician. I work at CMS. And contrary to most people's thinking, that does not stand for Centers for Mayhem and Scalduggery. It is the Centers for Medicare and Medicaid Policy. I work in the hospital and ambulatory policy group. We help decide how much we're going to pay for stuff in that particular group. I was asked to do this, not because I've been doing payment policy a long period of time, actually only about a year. But most of last decade, in a previous life at CMS, I managed the coverage process. And so I know a little bit about the evidence part. It's been about a year working for Carolyn Clancy at AHRQ, a little bit more about evidence in the last couple of years before coming back to CMS, worked in a small nonprofit in Baltimore, who has recently published a guidance document on how best to design clinical trials around molecular diagnostics. And so I have a bit of experience with that. My job today and my task today is to explain to you the Medicare payment system in 15 minutes. And so I'm going to talk to you about how to push this boulder uphill that's part of the Medicare payment system. There are four things that must happen to get paid in the Medicare payment system. One, it's got to be legal. Two, Congress has got to say that you can pay for that particular type of a service. Three, it's got to be reasonable and necessary. We'll discuss with you why that is not safe and effective. Different terminology for different agencies. And then talk about the coding and payment issues that are probably most in your mind. The legality issue is essentially an FDA one if a particular test in the diagnostic world or a drug in the drug world or a device in the device world requires FDA approval. It has to be approved at least for one indication before we can pay for it. That's a bit more complicated in the diagnostic world in that there's sort of two ways to get diagnostics into practice. One, the FDA way and one, the LDT way or homebrew as we called it in the older days. And if FDA has decided that something needs to be approved by FDA then we don't pay for it until it reaches that approval. But in general, not a major issue in most of the diagnostic world. The bigger issue is that Congress has got to have created a class of services for the particular service that is of interest. And my comments are going to focus mainly on molecular pathology tests because as I'll mention in a moment most of the other services get paid without a lot of intervention on CMS's part. There are benefit categories that are large groups. Congress has pay for hospital services, pay for physician services, pay for decade ago drugs in the outpatient setting. In some cases they say pay for a specific service. And that is commonly in the screening and prevention world. There has been a long standing interpretation of our statute that says CMS cannot pay for screening and prevention. Until recently that was unless Congress had passed a specific law on that particular screening or prevention test. So the only reason that CMS pays for cervical cancer screening, mammography, PSAs, a whole litany of screening tests is Congress has passed a specific law on that specific screening or prevention service. Now recently, a couple of years ago, Congress passed a law that says for anything else that we have not passed a specific law on, they call them additional preventive services. You can pay for those if you find evidence to support them and if the USPSDF has greater than A or B. So that's still a pretty high barrier particularly for a number of the molecular pathology issues that we're talking about now. The third obstacle is the coverage obstacle in our statutory language 1861A1A, 1862A1A. And I know that we're mostly federal people here so we love knowing what statutory guidance says. It says that we cannot pay for things that are not reasonable and necessary for diagnosing, treating or improving. And that's where we don't pay for prevention or screening prohibition came from. It's not listed in this sort of Bible for CMS. Reasonable and necessary has never been defined in a regulatory setting. It has been considered in some non-regulatory documents to be adequate evidence to conclude that the item or service improves health outcomes, adequate evidence as appropriate studies. In the diagnostic world, CMS has been fairly clear that reasonable and necessary means there's clinical utility of that particular diagnostic. There are, and here comes the fun, there are various ways that CMS makes these coverage determinations. Contrary to popular public opinion, CMS does not pay any bills. CMS hires contractors. Contractors pay those bills and they pay those bills on a regional basis. Several years ago, 35 regions is down to 16 and will be 10. It's getting closer to 10 now. And those contractors pay bills based upon national guidance. If there is no national guidance, they make their own rules. And so for each of these regions, if there is no national guidance on how to pay for a particular technology or whether evidence exists upon a particular technology which would cause a national coverage decision, then each contractor can have a local coverage decision. And there can be an equal number of coverage decisions that differ based upon the number of regions. They're commonly similar, but they don't need to be. Another important thing about this local is that for lab tests in particular, if you have a single lab in the country that's producing a particular test, Oncotype DX, Genomics Health in California, then there only needs to be one local decision because that local contractor in the region where that test is performed will make a local decision around how that test is performed and that becomes national because all the bills are sent out by that particular lab paid for by various contractors but to that lab and that lab has control over what that coverage determination is. So there are various incentives based upon the Medicare payment system on where you put your lab. There are a lot of labs being developed in Florida and Puerto Rico now. And if you look at some of the pricing issues, there's reasons for that based upon the contractor who manages that particular region right now. So the complications of the Medicare payment system make it difficult first to know how successful you're going to be in your particular item and it may somewhat assist in the decision making is where you're going to do your particular work. CMS has only made one national decision around genomics and that's in the war for insensitivity arena where it has a national decision that says that CMS will pay for the evaluation of war for insensitivity if performed in a specific study. And that was the study that Terry was mentioning earlier that only recently has been begun. It took about three years after the decision was made before a study was put together and so essentially there was no coverage for this particular war for insensitivity test until that study began and there's no coverage except within that study for that particular test. That's something that we may discuss at length if you wish later on. And that coverage was evidence development concept that war for insensitivity testing is a part of. So let's get to the important part, coding and payment that most people are interested in. How much are you going to pay for my particular service? And again Medicare bases its payments on fee schedules. There are 16 different fee schedules. There's an inpatient, there's hospital outpatients, there's physicians, there's hospice, there's skill nursing facility, DME, they go on and on. And each of those fee schedules establish once a year payment prices for that particular year except for the lab fee schedule. We'll talk about that a bit more in just a moment. What is important, most important I think about this slide is the fourth bullet and that is if CMS pays for things that have a code and have a price against that code. So if you have a code for any service and that service has been priced, it is difficult to not pay for it. It requires significant work, a lot of editing and payment systems, a lot of national and sometimes local decision making to not pay for that service. And so there are a number of things that Medicare pays for for which there is, as one of our members mentioned earlier, very little evidence of benefit and that's because it has a code and it has a price. And there's very little way for CMS to know about what it is and is it paid for other than through those particular codes and prices and without going through some significant work to not pay for those. Lab tests in general will focus the rest of the conversation just on lab tests as an aside for other types of genomic work, for consultations, for decisions on what tests to order. There really are no limitations and no decisions at a national level on what you do and don't do. The very first line of Title 18, Title 18 is the section of the Social Security Act that governs Medicare. I wrote it down so I get it right. The very first line in that title says nothing in this title shall be construed to authorize any federal officer and employee to extra size any supervision or control over the practice of medicine. Pretty far-ranging limitations on the kinds of guidelines that Medicare can put out. Now, a lot of people say we do a lot of control. Well, the controls are not on how medicine is practiced. The control is whether you're going to pay you for that particular practice or not. You may call that control. CMS has different words for that. There are two ways that lab tests are paid for. They're paid for on the physician fee schedule. In some instances, mostly on the clinical laboratory fee schedule. The physician fee schedule is what pays all physicians bills. It pays practitioners who are certified by Congress to be paid. Physicians, nurses, practitioners, physician assistants, clinical nurse specialists. Different practitioners have some ability to be paid by CMS. Medical geneticists are not part of that list. So if you're a geneticist, a non-physician geneticist, you cannot be separately paid for your work, either counseling or in evaluating the test. In general, the physician fee schedule is only used when there is specific physician work involved in that particular test. So it's mostly for pathology kinds of things and pathology specimen kinds of things and diagnostic imaging. Most lab tests are paid on the clinical laboratory fee schedule, and it's a bit different than most of the other fee schedules. There are three ways that the clinical lab fee schedule ends up being priced. The contractor pricing is typical, and that can be through two methods known as crosswalking or gap filling. I know this is exciting. Crosswalking says we're going to price it similar to another test. Gap filling says there's not a similar test. We're going to go out and find the cost data around a particular test, and that's what we'll decide to pay for it. And contractors do that within each of their regions. When a new test comes out, they have a specified period of time to come up with that information and generalists based upon bills that are sent to them for that particular test. And then they publish what they believe to be the appropriate cost. They send that to us at central office. We review that, and later in the year we put out what is known as the national limitation amount. And that is a value that we arrive at that's somewhere approximating the median, though not always the median, of those various contractors' price. And so the contractors still do the bill adjudication, and they will pay the lowest of what's on the bill, their price, or the national limitation amount. The national limitation amount is the most that will be paid. And the next to the last bullet is an important bullet. There are no, CMS does not make changes to these prices on an annual basis as every other fee schedule. So once a price is set, unless Congress changes it, some other very unusual things, it remains that price forever. So we're paying the same thing for a lab panel, basic chemistry panel that we paid for in 1969. And they're far cheaper today than they were in 1969, but we pay the same amount. Those are the rules that we have to live by. So you can see that while there's a bit of concern at the agency about $3,000 test, if you establish a price for $3,000 test, it's always going to be a $3,000 test regardless of what happens, unless there's some other intervention. So let me talk just a bit about MOPATH and where things stand. Before 2013, molecular diagnostic tests were paid under a stacking code system. You see CPT codes in that first column there, and here are the definitions of those CPT codes. Those are what you did to arrive at that code, they're not code specific, and you would bill a certain number of codes at the price for that code and add up to some price. No idea what that test was, this is what you would get. So for instance, for cystic fibrosis gene, we paid, if you submitted these stacking codes, $3,500 for a cystic fibrosis test. For this particular one, it's $2,100. Not a very good way to know, in fact, what's being done, and not really a way to know whether what's the appropriate number of these particular services ought to be involved in a particular test. And if you look at the different labs for that, you could see that for this particular genomic test, there were three labs that had vastly different values for them. CPT created some CPT codes for molecular diagnostics tests. Most of them made available this last year, we chose not to use them. They have improved at some point in that there are 105 specific diagnoses to be made by genetic tests. BRCA, KRAS, cystic fibrosis, there's 105 of those. And they have specific values that will be established for that specific test. But for all the new genetic tests, they established 10 more codes, and those codes are resource-based. So level one, the name of that code is level one. And that is, and it has 27 genes listed under that code. And those 27 genes supposedly need less resources than the other codes. And so for each of those tests, there are a number of genes assigned to that particular level. And so if your particular gene is not one of those 105, they fall into one of those categories. And I hit the wrong button. And so what does that mean? If what's becoming, as you know better than I, a popular pastime of coming up with panels of genetic tests, screening that are diagnosis-specific that a number of labs have developed and we've seen bills for, here's not an uncommon one, a panel for hypertrophic cardiomyopathy, which has these 18 particular genes as part of those tests. None of those genes have a tier one, part of that 105 test. They are under one of those particular six levels of codes. And so we end up getting a bill for 18 level two codes, stacking codes, essentially is still what we have. So CMS would see a bill and still don't know the particular tests that are being done because all we see is the name of the CPT code. We don't see the name of the gene that's being tested. So what's happening now? We're in this process for all of these new molecular pathology tests of creating prices through gap filling. The contractors have submitted those results. These are just some examples for the cystic fibrosis test I showed earlier. The gap fill prices vary between 800 and 1300, far different than the 3,500 that was being paid by stacking codes. And even more interesting and more, as some would say, detrimental to the development of genomic testing is that many of these tests will be defined as screening tests. Cystic fibrosis and a patient who doesn't have cystic fibrosis is a screening test by CMS definition. So it will not be paid for by CMS. And so the price will not be between 800 and 1300, it will be zero. So a number of these tests that have been paid for previously by CMS under stacking codes because there is now a specific code, or at least the genes are under a specific code, those particular genes will not be paid for. And this is the other example that I showed. So what do we need? What does CMS need? First of all, if people are unhappy with the way the payment system works at CMS, you need to talk to our board of directors, known as the U.S. Congress. We change our payment policies, our payment rules based upon statutory guidance. There's some wiggle room around that. We wiggle as much as we can. But our, you know, most of your federal agencies, you know how your general counsels dislike wiggling. They want you to be pretty close to what the statute has to say. We also need some help in convincing the people who come up with these codes that there may be a better way to code for this other than the way that's currently coded for this. Because if you've got 47 genes that fall under one CPT code, some of those will be screening codes. Some of them will be not. How do you decide how you're going to pay for that particular code? Is there a better grouping that could be done? Could we convert all these to specific disease codes? Is there some other way to do that? You're smart people. Hopefully you can help with that. As I mentioned earlier, we pay for things when there's clinical utility. Our contractors look at clinical utility. If you've had some discussion of that, the more clinical utility we have, the better. And finally, pricing these tests based upon cost. So all these gap-filling prices that I talked about were based upon information that our contractors were able to gather about the cost of test. And in general, they were able to gather none. All they had were bills that were sent in. Because we're not looking at what are you charged. What are the costs? And most of the companies, most of the academic centers are not interested in telling CMS what your costs are. And so if we don't have cost data, then we're going to have to use something else. And in general, that something else is not something that you're going to be happy with as far as payment goes. So that's CMS in a nutshell. That's payment. I know it was long and tiring and more bureaucracy than you like, but I thought it was helpful to understand. Very helpful, Steve. Thank you very much. So maybe somebody other than Mark first, if we... Yes, Pearl. Anything to go before Mark? Let me think of a question now. Did I catch you right? You said that if you, like, are testing for cystic fibrosis and then come up without it, it now becomes a screening test and you don't get paid. No, if you're testing a patient who you don't know has cystic fibrosis. But you have a high suspicion. If you don't... It's the definition of screening that was on one of the slides. The third bullet, if you don't meet that definition, that it doesn't get paid for. So I need a symptom. Okay. So if I have a big family history, that's not enough. I have no symptoms, so... Correct. Wow, interesting. Would you pay for an EKG if it's normal? You know, let's say, you know, the hypertrophic cardiomyopathy as an example. Other, quote, screening things other than a genetic test. Are you always praying for abnormal findings? This applies... This is not just for genetic testing. This applies to all screening tests unless Congress has said specifically you can pay for it. And Congress has said you can pay for an EKG for an annual physical. Okay. Whether it's good or bad or indifferent, Congress said to do it. Just like it said, to pay for a PSA every year. Good, bad or indifferent, you know, does what Congress tell us to do. And we generally do that. Thanks for the clarification. Okay, so I'm curious about the 1862DDD1. I read rags all the time. Is the reasonable and necessary for prevention or early detection? Is that only...is that for specified tests? Yes. I don't understand how that works. So BRCA must be one of them. BRCA is one of those... BRCA as a screening test is not paid for. BRCA is paid for is a diagnostic test. It... Sorry, I'm just telling you what the rules are. Yeah, I should comment. We didn't invite Steve here to defend the rules. No, I actually wasn't at wanting... Yes. Yes, and so if there is not a specific statutory basis for paying for a test and that test is a screening test based upon this diagnosis, this definition, and has not been reviewed by USPSTF, it ain't paid for. And so women who have family members with breast cancer who are BRCA, one or two positive, we don't pay for their screening tests. Okay, we're a bit over time. We have time for two quick questions. So, Mark. Yeah, so just an additional clarification specifically for BRCA. There's another one where a local contractor has made a coverage decision related to that. And so the definition of coverage is... came from the local contractor that covers Utah where the testing laboratory... SoulSource testing laboratory is located. So that is currently what is happening with that particular test. There's one other thing I just wanted to come back to related to the genetic personnel and coverage. Genetic counselors are also not listed as specific billable entities. And so that is one other issue about a provider that cannot independently bill CMS for services provided because they are not represented in the statute. Which does lead again to some internal conflict because 1862 DDD-1 states that you have to cover things that are graded A or B by USPSTF and related to family history of breast cancer. The specific recommendation is genetic counseling. Although it doesn't specify who provides the genetic counseling. And that's a B recommendation. Most medical geneticists who are attempting to get paid by Medicare do it through physicians. There's rules that allow physicians to pay other people to do the work. A physician can get paid for medical geneticists doing the counseling. It's called the incident to benefit. Actually, that's not true for genetic counselors. Unless CMS has changed their tune on that because when we investigated this when I was chairing the economics committee for the American College of Medical Genetics and Genomics we were specifically instructed by representatives of CMS that because genetic counselors were not represented as entities defined by statute that they could not be used in incident to billing for physicians. That's not correct. Okay, great. Could I get that in writing, please? Okay, again, let's not beat up on Steve. Yeah, so we had Dan and then maybe Deborah and then we really do have to... Oh, sorry. Oh, okay, thank you. Hi, thanks very much. In the spirit of us trying to integrate our knowledge bases and resources could you just briefly describe the process by which the contractors assess clinical utility and are there resources that could be used around this room to help inform those decisions? The guidance that's provided by CMS to the contractors to use in determining clinical utility is broad. It is not specific and they don't have a standard process for determining what that is. On a national level, CMS, while it says we want clinical utility to make an RNN decision on a diagnostic, have defined that in different manners. So, for instance, there's a PET scan, national coverage decision that says clinical utility is demonstration of change in management. So they define clinical utility as change in management. In other cases, they define clinical utility as change in outcomes. And so there isn't a single definition of clinical utility. And contractors all have different processes for deciding how they're going to do their local coverage determinations, whether it's clinical utility or something like that. Dan, did you? No? Okay. Then you have the last one and we really do have to go on. No, Amy? No. So it may be helpful to understand how things are done in the Blue Cross and Blue Shield Technology Evaluation Center, which is a resource for clinical decision making for 38 plans with 100 million beneficiaries, and it is different than CMS, but quite influential. And there are consistent definitions of clinical utility. That definition is to improve a health outcome, but there are also different ways of getting at that evidence. There can be direct evidence. Direct evidence is pretty hard to come by quite honestly. In the area of diagnostic studies, that would be something like Taylor X, which we don't have the results of yet, though some years ago we did look at Oncotype DX in the setting of adjuvant chemotherapy and said, yes, this does provide clinical utility. The other is an indirect chain of evidence where you try to understand what is that performance of that test? What is the clinical setting? What are the decision makings? How would that information not only change the decision, but improve the outcome? And on our website, bcbs.com, slash tech, there are examples of our recent assessments, and I'm glad to go into this in more detail, but I just wanted to give you the scheme of that. Great. Thank you very much. Thank you very much, Steve.