 Sgwrs, mae'n gwybod i'r ffordd i chi'n gweld, fel'i ddechreu y peir yw'r ysgwrthfawr, lle'r llwysawr ar y FNH. Felly, mae'n gweld y ffrifio cyflwyng CMTP a'r SHEI. Mae'r rhai o dda'r ffyrdd i'r ysgwrthfawr yw'r wyf. Felly, mae'n gwybod i'ch gweld i'r ysgwrthfawr, gweithio mewn dfodol i'n bwysig a'r byw yn ddechrau'r bydd hynny yn dweud o'r cyfrifio o'r ddechrau. Felly, y gallai y syniad bwysig hwnnw? Felly, mae'n bwysig hwnnw yn ddweud o ddweud o'r ddweud o'r ddweud o'r ddweud o'r gweithlidol genomig a'n ddweud o'r genomig ar hyn, a'n ddweud yma. As I's been talking about a lot over the past day, this coverage is contingent on demonstrating clinical utility. For a lot of these tests, there is insufficient evidence of clinical utility to persuade payers that these are things that are clearly worth testing. So there are questions about who should be paying for the test, who should be what research needs to be done in order for clinical utility to be demonstrated. i'r ddweud ymwneud, ychydig i'r ddweud ymwneud. A'r ddweud yw'r ddweud yw'r ddweud, mae'n gyntaf i'r hyn o'r cymryd yn berthynasol. Felly, rwy'n meddwl gennymig yma, yn y Cwgogol, ond rwy'n fawr i'r ddweud yw'r ddweud y ddweud yma ychydig i'r hyn o'r ddweud yw'r hyn o'r ddweud i'r ddweud i'r ddweud i'r ddweud. Ie, mae'r cwestiynau o'r cwestiynau yw'r cyfnod o'r wneud, ond ni'n gwneud i ni i'r LHGRI i ddim ni'n ganddo i ddim ni'n ganddo i'r cyfrifiadau o'r ganwraethio'r cyfrifiadau a'r gwneud i ni'n gwneud i LHGRI. Yn ymwysg yw yma yn yr hyn ychydig i chi ddweud yma yng Nghaerfer, What are the endpoints that influence coverage decisions by health insurance issuers? Is it just clinical utility or is other health efficiencies like length of hospitalisation and so on? Given that so many of these tests don't have clear clinical utility, are there innovative ways in which payment is being done and patients are receiving tests in the absence of that clear data? So, after the workshop we had a subsequent call with Sean Tunis and his colleagues at the Centre for Medical Technology Policy, and that was the origins of this workshop that was held in October. So, as I was mentioning, several members of the Genomic Medicine Working Group were there. There were also a number of payers we had representative from Wellpoint, some from Palmetto GBA, which is a Medicare contractor. As someone mentioned yesterday, CMS was also there. A number of test developers were also represented, myriad, genomic health, Everson genetics. Wean Currie was there representing the CDC. Eric and Laura and Jean, a number of NHGRI staff were there, and a key group that I missed off unfortunately on this slide is the Centre for Medical Technology Policy, the group that we work with. So, we decided to focus, to structure the workshop on four test cases. Some of these are a little different in the details, but some of these are instances where there is limited evidence of clinical utility of a test, but where an insurer is nevertheless willing to pay, and patients are getting access to the test. The first one is the one that I think it was Mark mentioned yesterday. In 2009, CMS announced that they were going to, for pharmacogenomic testing for warfarin, they were going to start a coverage with evidence development policy. What that means is that they are willing to cover warfarin testing, but only if a patient is in a clinical trial. Jeff Roche presented from CMS, and then we had a representative of Everson genetics who is funding a clinical trial to look at the impact of warfarin testing on hospitalisations of severe clinical events. They are expecting that research to come out at the end of this year. We also had a couple of representatives from Genomic Health who talked about their performance-based risk sharing agreement they had with United Healthcare with regard to Oncotype DX coverage. They weren't at liberty to give us all the details of the exact agreement between them, but it was related to the collection of the data to see if the use of Oncotype DX would reduce referrals for chemotherapy. Then a couple of former colleagues of MedCo talked about a partnership that they had with the Mayo Clinic, again where they were looking at what the effect was warfarin testing. Lastly, a little bit of a different one, I'm sure many of you will have heard about Palmetto GBA's initiative they started. Last year, the Moldex program, one of the challenges that this Medicare contractor reported as having was that they were getting bills from labs and they really had a limited understanding of what exactly it was they were paying for. So they would get a bill from one lab for $150, another bill from another lab for $650, and to them it looked like they were paying for the same test. So to address this, they started this program where if you want to be reimbursed for a particular test you have to let Palmetto know exactly what it is that you're doing, what the analytical validity is, what the clinical validity is, what the clinical utility is, and then they make a decision as to whether they think it's reasonable and necessary for the Medicare population that they are responsible for. So we considered all these examples and were discussing whether these could be models for broader application, whether there are others that could do similar kinds of things, and to listen to some of the emerging themes. So these were some of the things that came out. One thing that's apparent is that there's little consensus. We can't think of all payers as being the same and the United Health example, they were interested in referral to chemotherapy. I don't know that that would particularly fit with CMS because CMS says they strictly look at patient outcomes, those are the kind of barometers they look at. So in answering the question, what is it that insurers want to see in order to determine whether something has validity or value, it's a complex question to answer. It was apparent that from the examples and from the discussion that payers are willing to consider evidence other than randomised clinical trials. The Palmetto representative Becky Turner was clarifying that they certainly, that's the gold standard, but they are willing to consider other evidence and whether a test has clinical utility or not. We also discussed Medicare statutory limitations. So Medicare can't pay for preventive services. Medicare can't pay for confirmation of a diagnosis as opposed to making a diagnosis. And I think this is something we need to think about because if Medicare is often used by other payers as a reference point, and if they're not paying for genomics-based services because it doesn't make sense for their population or it doesn't fit within the statute within which they operate, then there might be something we might want to think about how other payers may not also adopt for that reason. There was a discussion about who is responsible for funding this evidence base. It was quite clear from some of the payers that they don't think that that's their job. You bring them the evidence when you think you can justify this is something they need to pay for. And also, as was alluded to yesterday, it was apparent that there are some complications when it comes to sharing of patient data between a provider of health insurance and a test developer. So we came up with a few action items, some of which are no underway. The first was to write a white paper summarising some of these case studies thinking it might not be apparent to everyone that these are options, these are ways in which things are being done to get around the clinical utility problem. And also to summarise some of those emerging themes. We're going to conduct a policy analysis on data sharing what the issues are. Of course, Hipper is an issue Pearl mentioned yesterday about some state laws. There may be other things, so that'll be, we're going to get going on that next month. Mark's going to be leading an effort on research on physician ordering of tests. So one of the points that came up was that no matter how good the clinical utility is, if it's not being ordered appropriately or interpreted appropriately by physicians, then of course that clinical utility is very much limited. And then another point that was made, there were so many of these tests with so little clinical utility perhaps you ought to think about how to prioritise these. And lastly, there was the discussion about how much of an investment it is to do clinical utility research. So there was some kind of an idea of, if there might be some way in which provides some kind of infrastructure by which this could happen. So the white paper is an early draft, is ongoing that the second and the third were just starting working on those, the fourth and the fifth haven't started yet. So if you want to, I've tried to summarise everything that was in the workshop, but if you want to read in more detail, you go to that URL and it'll give you a more in-depth summary of the workshop. So I was also asked to come up with one or two ideas as to what might be next steps for things to think about. So these are completely unveted, they're just things under my head, so take them as you will. But it struck me that perhaps one of the things we want to do is to talk to some of the decision makers who really are responsible for determining what gets covered, who gets covered and how, whether it's actuaries, underwriters, to truly try to understand things from their point of view. Something else that struck me is that we often talk about the lack of coverage, but we rarely do we specify which tests we think really ought to be covered right now and which to have clear clinical utility but where health insurance issuers aren't picking them up. So I think that it would be interesting to discuss which tests should be covered right now and which tests might be appropriate for this coverage with evidence development kind of model and I use that broadly. And then of that it would certainly be interesting to look at what is covered by, what currently is covered. And lastly I think one of the longer term goals of what HGRI is, think about how we could either fund research or partner with others in research. Lastly I just want to flag a number of other issues that are out there which weren't really addressed by the workshop but are certainly within the realm of coverage and reimbursement and the question is whether NHGRI would want to take a lead role on this, whether we want to partner, whether it's just something we want to monitor and be aware of. The first is that Jeff Roche of CMS said quite clearly that if there was an application for instance of whole genome sequencing that fit their definition of reasonable and necessary for the Medicare population that's something they would definitely entertain so it might be worth thinking about whether there's a case that could be made for something like that. I don't understand things from their point of view, recognising of course, as I was saying before, the central role of Medicare. Then there are some of these other things that are going on, some of the interest to the AMAs and CMS is the adoption of new CPT codes that's going to allow better tracking of which tests are being done. And there was discussion at the workshop in parallel to that. There's a lot of enthusiasm for the Moldex initiative and there are questions whether this is just something that's happening within the California and the other states within the purview of parallel to GBA in that Medicare contract area where this is something that should be a national programme. For the genetic councillors, the most important coverage and reimbursement issue is direct billing of course. Something that I thought might be interesting to look at. We know the FDA now has over 100 drugs where on the label it has some kind of genetic, some kind of pharmacogenomic indication. And given that there is great concern about how physicians are misusing the tests or misinterpreting the tests, it might be interesting to look at some of those and to ask whether we think that they are really clear as to what the use of the test and what the use in helping administer the drug is. Lastly, a couple of things with the Affordable Care Act, the exchanges kick in in 2014 and one of the pieces of that is that each state is coming up with defining what are the essential health care benefits that need to be covered in order to be in the exchanges. The question is, do we think there's anything within genomics that meets that bar? And similarly, the recommendations of the US preventive services task force have been elevated, I think, to a new height with the passage of the Affordable Care Act. And so there are requirements if there's, if there's an A or B recommendation from the task force, there are requirements for health insurers to follow that recommendation and coverage decisions. So the question is whether we think there's anything within genomics that we would want them to consider. So those are just some ideas. So I'd like to thank Mark and Jean, who are on endless planning calls for the workshop. I know Mark enjoyed them particularly, as well as Laura and then Laura Coons, who is instrumentals now, who's a fellow in our office currently in Louise Slaughter's office and gone to the hill now, as well as the CMTP staff who are instrumental. So with that, hopefully I've tried to give an overview, but there are many of you there. So Mark, Jean, or whoever else might want to mention one or two things that I've skipped over or missed. Just to make the observation that those four examples dealt with really serious clinical problems. Two with Warfren. If a patient started on Warfren, it's my understanding that a quarter of them are in an ER within a period of time with either clots or bleeding. So that's a serious problem. Another one is whether a woman with breast cancer has to get chemotherapy or not. And then the other one had to do with the correct approach to lung cancer with targeted therapy. So somehow out of the mix, they chose really important clinical questions. So let's start with Jeff and then Mark and then Bruce. I have two questions. The first is could you clarify whether there was an explicit commitment from the payers to actually fund research on clinical utility, or is that an aspirational goal for the group? I would say it was more correct to say there was a specific commitment not to do so. We only had a few representatives there, but I think from the few that we had, I think they don't see it as their job. If you want to go to them and justify why you think your test needs to be covered, then they will consider it and they will look at the literature, but they are not going to a priori go out there and do the research themselves in order to determine whether they ought to cover the test. That was certainly what I heard anyway. But I think one of the things that was discussed was the idea of could there be innovations in coverage design that could somehow facilitate a collection of the data around certain key questions that the payers wanted answered. So for example, the United Healthcare Scenario, they were reasonably convinced that Oncotype DX was useful and had utility based on the evidence, but they were not convinced that clinicians would actually use the information to change care. What they were concerned about was that they were going to pay several hundred dollars for a test, and everybody was going to get chemotherapy anyway. So the way they set that up was to say, we will give you a target and if you can demonstrate to us that this does in fact change the care in some predefined, and we didn't get all the details about how this was set up because that was done under a proprietary situation, we will pay you for the test. So they worked directly with the test provider to collect data, so the test provider actually contacted all the clinicians that ordered the test after the test results were given and then collected data on whether or not chemotherapy was given. And they actually then hit the target that United had set in conjunction with them and they were reimbursed for the test. So it was that sort of idea which is that no, they don't want to pay for research because that's not really what their job is, but they're not necessarily averse to thinking about different ways of using their benefit structure to encourage collection of data that they think is key to making decisions. And I mentioned yesterday, I did follow up since well points headquartered in Chicago, I did follow up with well point representative and they are very anxious in partnering with us to provide data that may be useful. So I think as a first step we should really make sure we take them up on that offer. Yeah, so the data, one of the things that we found particularly in the Mayo MedCo study was the idea that there are data sources out there, in this case a pharmacy benefit manager that has a lot of data that could potentially be used to answer research questions and that there was willingness, I think under certain conditions to be able to share some of those data to be able to again answer specific questions. So it's just for us to think about other places to look for the data that we're interested in. Yeah, that could be incredibly valuable outcomes measures for some of our initiatives. So my second question may be, it's a brief answer, but did you have a chance to explore how they are thinking about covering clinical genome sequencing? We've heard. We didn't talk about it. I mean, I think that's the fair thing. We really talked about these four use cases and innovative designs so we did not get into the space of saying, here's what's coming down the road and really spend much time talking about that. I think oncology is an excellent area to look at and first of all in regard the genomic health, it's several thousand dollars, not several hundred. That sounds like a lot, but it's not because oncology payers are reimbursing ten thousand dollars a month or more for treatment regimens and back you could say diagnostics are a drop in the bucket of the national problem regarding oncology costs and so I think it would be very useful to analyze the entire oncology model. The other thing to keep in mind is that you may be talking to CMS regarding national policy, but the local coverage decisions are rampant and there are, I won't say hundreds, but there are many companies out there that have seduced the local provider into making a local coverage decision for a laboratory which then allows them to, of course, receive samples from anywhere in the country and bypass national policy. So I think you need to really look carefully not just at national policy, but at all the local coverage decisions regarding oncology testing, not just genomics, but there's a lot of companies out there selling chemotherapy testing, things like that and often these tests are ordered by providers who don't know what to do because the patient has progressed on all evidence-based regimens and so they're looking for some justification to give more treatment and add more costs and so they order a test of no proven value and then they give more therapies that add more costs of no value and so this is a real mess and I'd be happy to work with you guys on this. Yeah, we, actually that did come up and we did talk about that and one of the use cases was that a demonstration project that CMS funded to have a clearinghouse of all laboratory tests to try and get around some of these local coverage decision issues, but that is a topic that we are going to be discussing in the workgroup that I'm going to be running and I'll sign you up. Okay, so that teaches you to say the same thing around here. Bruce. I think one of the issues that distresses some payers and hospitals like is the tendency to encourage what I guess could be described as shotgun testing or bundling of tests. Don't include in this genomic sequencing but there are many examples where a clinical question is phrased in a sufficiently vague way that where a $500 test could have answered the question instead a $10,000 test gets sent looking at every possible gene that could ever be imagined to be involved in that phenotype even though some are much more probable than others and I think that actually at least in my institution has actually given genetic testing a bad name in general because the perception is genetic tests cost $10,000 a shot without realizing that in fact if you break them down and use clinical judgment you can often do them for much less so I guess in a way it broadly comes under the heading of clinical utility but I think it is a kind of a mountain within that area that needs to be addressed specifically. This is the utilization management issue and again that will be under the purview of our group the appropriate ordering of tests and as I mentioned yesterday the ARUP white paper that came out showed that in their portfolio $30,000 to $36,000 a month of testing was being ordered inappropriately and there is no disagreement around anyone that there's any worth to doing inappropriate testing but the challenges Derek alluded to in the presentation was that because of coding issues and other things it's almost impossible for the payers to actually do effective utilization review either on the front or the back end and so this is going to be the area about could we in fact develop some solutions that would be able to address this problem so that is high in everybody's priority list because we all recognize that this is just pure and simple waste. There's no impact on patient care except to the possibility that we're going to lead to more expenditures and as Mark pointed out inappropriate therapy so yeah, we want to get rid of that. I've got several in the queue here and we're relatively short so I think I'll just kind of keep going but I'll add you to the list. Gail. So several issues that are big obstacles clinically right now are pre-authorizations and the need to test relatives in order to efficiently care for the patient in front of you and those relatives may not benefit from that test or they may be deceased and we may be working with a repository or a tissue sample and I didn't see you see those on your list of other and I wondered if either of those was getting bandwidth I know we had one insurer who used to if a medical geneticist ordered a test that we didn't need to pre-auth that and they actually got rid of that policy. No, that's not something I had thought about as I was generating the list but I did realize as I put the list together I started thinking of more so I think there were certainly that wasn't supposed to be an exhaustive list it was more like jotting down some of the other things that perhaps we should start to think about perhaps that should be, we should add that one. Yeah and that falls within I'm not trying to characterize the workgroup that I'm going to be leading as the be all and end all of the questions to this point have directly impacted on what the meeting actually flagged as a huge problem which is the appropriate utilization so that will be a major focus. Cathy. Was there any discussion about the first of all I had two questions any discussion about coverage of interpretation or was it really about specifically the test itself? We didn't really get too much into coverage of interpretation it was more about whether tests, whether specifically because of the nature of the case studies it was more about specifically whether the tests are being covered but obviously that's an an additional issue. Although the United Genomic Health use case really was about interpretation because it was asking the question are clinicians using the information to avoid chemotherapy so even though the interpretation was not formally measured it was indirectly measured looking at utilization. Okay and then my second question was there any further discussion about clinical utility and the lines between that and personal utility or social utility in the sense or was it and what the value can be of those genetic tests based on those definitions or was it strictly discussion around clear cut clinical well I'm not sure what that is either but clinical utility To the extent that we all agree on exactly what clinical utility is it was more on the latter there was little discussion of personal utility what benefits someone might get out of personally of understanding stuff about their genome and again it was obviously focused on what health insurance issuers are going to be prepared to pay for. I just want to say that in addition to coverage there are benefits packages that you know so I think having the employers the large employer groups involved they may be another source of data and I think their opinions and their understanding of this would be helpful and you know they're they're often driving what the insurers are are doing. I also just wanted to comment I am the ACMG representative to the Blue Cross Blue Shield Association Medical Advisory Panel and I have you know we're allowed to submit you know ideas for future topics so I have submitted you know whole genome or I shouldn't say whole but genome exon sequencing. You can say whatever you want. So I'm sure that will be discussed probably more as an informational piece and then maybe ultimately to kind of look at whatever evidence does exist so in our deliberations it usually is completely evidence based because we usually focus on technologies that are controversial and evidence is key. I think the employer piece is a good one it was something that definitely came up the challenge of course is that it's harder to find the you know what might be considered a representative employer. I have one I mean I think the person you might consider is Helen Darling She represents I can't remember her official title but she's on the Medical Advisory Panel at Blue Cross Blue Shield. Helen Darling If it's any of your contact information I'll invite her to be a part. That would be great. Thank you. Kate. As of today Medicare just released a document for public comment about reimbursement for multi analytical panels for cancer testing which includes cancer sequencing, type arrays and all of that they're actually looking for public comment in the next six to eight weeks we were contacted by the head of Medicare for our 14 states or 12 states or something like that and they're so we're coordinating actually an effort it's a weird sort of set of states but nonetheless we're recording a set of efforts that are actually going to come back with documents to support multi analytical testing for cancer biomarkers so I just want to make sure that everybody's aware that in fact now is the time to look at what the Medicare guidelines are for somatic panel and sequencing because they're just available for public comment and they want people to get together in each of the Medicare regions and comment from the institutions Kate could you make arrangements to send that link around to the group you could send it either to Ian or to Richard and we'll distribute it so that everybody has a chance to get that that would be great David I wanted to follow up on what Bruce had said there was the discussion of the coordinating committee or the society of societies I know we're not allowed to use that term anymore this would be a very good topic for that coordinating committee because if they're talking about guidelines or the concept of developing guidelines in order to convince insurers that you're acting responsibly if the societies started telling their members these are good ways to use molecular testing that sort of benefits everybody it's better medical care per Bruce and it's less money per the insurance companies and it's responsible for the societies thank you I think we've got time for one more I think you have the last word God makes me nervous two thoughts one I would like to endorse Derek's idea of really trying to put some data behind what tests it is we think should be covered right now and what should be considered with more data and somehow marry that to yesterday all of the surveys I was impressed by the fact that there was this universal the orderers not having a clue as to what is covered and it would be interesting to know what are they ordering are they ordering in the category of what we think should be covered or are they ordering the wackos so I think somehow to get more feather out more data in both of those categories may inform the process okay thank you that's a good point and we'll see if we might be able to to do it the challenge as I've mentioned before is that because of the current coding system the way that we would traditionally collect data is difficult to do because all we see are the procedural CPT codes for molecular diagnostics and we can't tell exactly what test was done and so whether you're in an individual institution and want to know what your doc's ordering or whether you're looking at it on a more broad perspective it's very difficult and even though there are some now modifiers that have been added the use of those has been sporadic although the demonstration project that was part of the presentations there for their region there was a requirement that the bill as submitted had to have the modifiers so they actually were able to use that to understand specifically what test was ordered and do some utilization but that was really the first group that had larger, broader data and that was still pretty early in its stages so this has been a problem that we've been discussing in the coverage and reimbursement realm before I was on SACGHS and working actually back to SACGT and we still haven't come up with a great solution. Great. Terry are you