 It's either good morning or good afternoon. I think we're right on the cusp. So I'm really, really pleased to be here. But I do need to tell you that I am here. This is my disclaimer. I have no conflicts. But because my far more able colleague, Dr. Gervany Rondawa, is visiting his family in India. But we would be thrilled to be able to follow up on whatever happens here. And I cannot overstate my enthusiasm for what you have pulled together here. I think it's critically important, and the time is now. And if there's two big forces framing the context for these discussions, which I'm enjoying immensely. First, and this will be a little bit of what you think depends on where you sit, is the value imperative in healthcare. And the reason that's so important is that for all of the background noise we're having about budgets and a sequester, et cetera, et cetera, it's very clear that an aging population, which is a tribute to the success of biomedical science in part, as well as public health, is what's gonna break the bank. So I can't say our conversations are terribly informative, but at a point where we're gonna be expanding access to people, the vast majority of people who are uninsured, that's exciting stuff. But if we don't actually get to this question of value in healthcare, we're not gonna be able to sustain that opportunity. That's if we get through the political noise, but that's the end of my little commentary there. The other context I think that's very important here is the notion of having a reusable or multi-purpose infrastructure for studies, for various types of research, for assessing and improving quality of care, for rapid dissemination and translation. Notice how that rolled right off my tongue. There's only one small little problem here. And a lot of this coalesces under many, many meetings about a learning healthcare system and so forth. And who doesn't think that's a good idea? It's fantastic. It sounds like the fun parts of school and training, right? To sign me up, I wanna be there. The reason it's not so much fun is it's actually far more efficient to have your own project-specific infrastructure. More expensive in the aggregate, but way, way more efficient. I've got this piece and I control that and I don't have to get bogged down by all the messiness of healthcare. So I'm here to tell you just how messy healthcare is. So our focus in life, whoops, I'm going the wrong way here, is very much about where the rubber meets the road or as my father used to say, what's for supper tonight. So many of the research investments we make are very much focused on how can we improve healthcare right now? What does clinical utility mean right now, given the evidence we got? Not what kind of information are we gonna have in the future, but what can people do today? We have a lot of investments in assessing and improving the safety of care and I will say that one of the emerging priorities in that field right now is diagnostic errors, which includes everything from ordering the wrong tests to the test results getting lost to wrong interpretations and so forth, but I just thought I would flag that since diagnosis is clearly a big theme of what we're discussing. We do a lot of work in patient-centered outcomes research, clinical effectiveness, and we also have the only source of data in the US that actually can provide rigorous information on what we're spending in healthcare for everyone. Not just those who are insured, but literally everyone including stuff that people pay for out of pocket. And we fund grants, training programs, peer reviewed, and so forth, and we support the US Preventive Services Task Force, which makes us famous or infamous depending on how they're rating. But a lot of our work comes down to the huge, huge contrast between efficacy or can it work and effectiveness or will it work in real life? And I won't bug, I won't read this out loud to you, but simply to say that the real world choices and information that people need is often very different than what occurs in the context of a reasonably controlled study setting. Now I mentioned the task force. A law passed in 2008 is what gives CMS the authority to cover for Medicare beneficiaries, services that are rated in A or B by the task force. For those of you who are not official or don't have an official liaison to the Preventive Services Task Force, let me tell you we would love for you to delegate someone to do that. Right now, although the work of the task force, both because of the impact on Medicare coverage as well as the direct link to the Affordable Care Act, which is to say that everything rated A or B will be covered by all plans for zero co-payments means that their work is increasingly important. Right now the task force does not have open meetings. They've evolved a variety of other strategies to get early input when they're considering a new topic from a variety of stakeholders to putting out draft reviews for public comment and so forth, I mean draft recommendations, but the bottom line is we would love more federal partners. And all of their recommendations are undergirded by very rigorous systematic reviews. Now I put this slide up here because this happens to be a particular test that the task force has been struggling with. If you're going to be screening for colorectal cancer, does it have to be a colonoscopy which we've come to think of as the gold standard or is colonography as good? And the only way to really know that is to actually evaluate a large number of patients using both tests. So we come from the standpoint that genetic tests are no different than other diagnostic tests in terms of being able to develop appropriate evidence. Now Angelina Jolie is obviously all over the news and I do believe someone pointed out earlier that the task force in 2005 gave an A, B rating rather, to clinicians first having a conversation about family history with patients and if positive recommending that they get the test. This turns out and they're reviewing that right now there will be another recommendation coming out because they routinely update everything that they do. Now this gets into very dicey business which involves some themes we've discussed this morning. The task force is a scientific group and they're really good at assessing science. They are not so great at translating that into insurance ease. So whether that situation will change or not but right now there's still something of a gap between what science says and what insurers need to know if we're gonna be assured that there's equitable coverage and payments across many insurers. Scientists of course wanna stop where the science stops which is a very laudable position. On the other hand, if you take that position then what that means is that you're delegating a huge amount of authority to insurers and health plans which means that no one has a right to appeal and so forth. So that's why that's gonna be a very important issue. But I really put up her picture because I think this is already a huge, huge public discussion. And I think future activities to bring the federal family together we would be very wise as we're thinking about how do we involve payers and others who are part of the private healthcare system since the vast majority of healthcare in this country is delivered in the private sector to include patient groups as well. Now, so what would we say from our standpoints are very, very important in genomic medicine? Clinicians and patients need to know about what are the benefits and harms of genetic testing. We've had a lot of discussion this morning about the urgent need for electronic infrastructure for clinical decision support and very importantly assessing real world penetrance. We haven't done this yet for genetic tests but we have done it for some other types of diagnostic tests constructed a data points series where we're actually beginning to look at for example how many women get preoperative MRI before they, well when they're being evaluated for breast cancer. It turns out if you're in the Medicare age population because that's where we have the best data it's about a third. Now whether that means the oncologists have been under enthusiastic or over you can tell me but the point is we know very, very little for any healthcare service including genetic tests how widely they are being used or not and ultimately what clinicians are gonna be looking to is guidelines and quality measures. So we've had some very thoughtful comments this morning about the urgent need for sort of structured knowledge as the basis for clinical decision support. Well you know where all that comes from is actually the clinical practice guidelines which are just about as squishy as you can possibly get. Some of this I think comes from a need for people with different opinions to develop consensus. So people may not agree on an age cutoff when someone should be tested or not but they can agree on don't do this for older patients. Older being anyone older than I am, right? So absent to really get to an ongoing system of having structured knowledge for clinical decision support that means the clinical guideline developers need to be part of the conversation and we've got a repository of evidence-based clinical practice guidelines at guidelines.gov. And the value question I think is gonna be increasingly important. What is the appropriate use? I will just tell you a story from my family where it was mentioned to my 83 year old aunt, 83 last week, that perhaps she should be tested for BRCA because she was being considered for ovarian cancer and her sister had just been diagnosed with breast cancer. No family history whatsoever. I'm just gonna say this is from a cancer center you would think of as top notch. I'll leave it at that. So we have a lot to learn about appropriate use and I think there's a real opportunity to learn directly from care delivery. If we can get past the messy inefficiencies of a common infrastructure for improving clinical practice as well as doing research, I think there's a huge opportunity both because you've got a built-in platform for rapid dissemination and because you can learn a lot and those are signals that I believe tend to get lost in the healthcare system today. Now I wanted to mention one other project that our colleagues from CDC are gracious enough to say we are working on with them. Actually they're doing all the work. We help them with some of the mechanics up front. When we collectively across HHS, largely NIH and ARC and then we worked with others in the secretary's office because $400 million was delegated to the secretary and she didn't actually have her own research agency. When we were working together, we actually allocated $20 million to the CDC to begin to examine what it would take to upgrade existing cancer registries. So again, even in the notion of multi-use infrastructure, the tempting thought is to say, we'll start from scratch over here and it's much more about driving while you're changing tires. I mean that's the only way you're gonna go to. So cancer registries become a very important resource. So this is supporting 15 specialized cancer registries to include genetic information, prognostic biomarkers and so forth. And I don't need to read these out loud to you. What I will say is our colleagues who couldn't be here today volunteered that they've learned quite a bit about getting to standard operating processes that you talked about because the difference between what works in a project versus how you're gonna make this part of standard processes I think is gonna be a big part of our efforts going forward. So how can ARC help in all of this? Our foothold or footprint in any of this is relatively modest and yet questions come up about it all the time. To the extent that it's relevant to task force recommendations, then we are evaluating the value of genetic tests as we would any other diagnostic test. I will say by the way that the task force is independent but we tend to get our pronouns confused because we provide all the scientific and other support for their work. The recommendations are theirs alone however. And we've funded a couple of other studies, one being a smaller study looking at the value of the test that identifies people at the highest risk of bleeding for morphine. And like some earlier studies from NIH, also a smaller scale, found that you could demonstrate a very, very small benefit which has nothing to do with the value of the test and has everything to do with the state of healthcare delivery today. So I would argue for my colleagues from the Genome Institute that although it's really tempting to say keep me away from that mess, you're not gonna be in a position to be able to have that luxury which doesn't mean you have to get submerged instantly and all the time. But I do think that's why I'm so excited about the fact that you're having this meeting today. So I think we invested of the $1.1 billion for CER, almost a third in data infrastructure which includes everything from creating newer, enhancing existing patient registries, distributed data networks and the like, a lot of investment in the methods to make sure that the inferences being gathered and cleaned from those observational data are rigorous and can be considered reliable. We also, working with the Patient-Centered Outcomes Research Institute, and I understand Joe will be here later, we have a unique mandate to disseminate the findings of patient-centered outcomes research. Now the one thing I've learned being on the board of PCORI and funding this work for quite some time is there is no rigorous definition of PCORI and I'm starting to see just how useful that might be. Certainly for this purpose, it's very useful. And our authority gives us 16% of the total of the Patient-Centered Outcomes Research Trust Fund. So in this instance, I'm mentioning an opportunity that has resources with it. And our mandate is not just to disseminate what PCORI funds but actually Patient-Centered Outcomes Research writ large. So look forward to future discussions on that front. Now the Medical Expenditure Panel Survey to give you the shortest description I think has ever been used, it is basically an ongoing household survey. If you ever sign up for this survey, you are just completely lost to any other survey efforts because we ask you questions like crazy for two and a half years. And when that household drops out, they're immediately replaced. So it sort of continues forever. For many, many issues that come up, there is one designated household respondent, but we also add to that and supplement it with self-administered questionnaires. So for example, for many years, we have been asking individuals in the household, not just the respondents, about their attitudes towards health and healthcare. I'm sure you'd all be shocked to find this is a well-validated questionnaire from NORC. I'm sure you'd be shocked to find that young healthy men tend to be the most skeptical of healthcare. Those are all the people we're trying to sign up for the exchanges, by the way. But in any case, there is a platform there. There's a machine that can be used in terms of assessing public opinion, understanding attitudes, and so forth. And again, I think this experience with CDC will be quite valuable. This is really up here for the picture of the smart phone because I think that as we're looking forward in terms of what is gonna be the next venue, four or five years ago, no one thought that actually smart phones would be the major vehicle or device by which we're transmitting evidence-based information. And yet, what's shown here on the slide is the app for the Preventive Services Task Force. This is not just flat information, you actually have to enter in one or two individual-specific references to get the recommendation. And then you can keep clicking to get to or touching, depending on your device, back to what's the underlying evidence. Now, because of the big investments in infrastructure, one of the other investments that we made and will be continuing for the next couple of years is in something called the Electronic Data Methods Forum because we felt that many of these projects working in isolation, we couldn't afford not to have them sharing their insights, lessons, and so forth in a way to continue to rapidly improve the measures. And in Indiana in particular, those of you who aren't familiar with this, Indianapolis, I would have to guess is the city in the US that has been the recipient of the most federal money for health information technology over the years, probably because they had the vision way, way ahead of almost everyone else. But they are actually adding to their very high-functioning health information exchange, the capacity to collect and store genetic and other types of biological specimens. But this is just a screenshot from something called the Healthcare Innovations Exchange. We're constantly interested in learning from people who are not researchers. What are they doing to improve care? Because sometimes actually they're doing much, much more than translating research into practice is getting us. And while ready, we're seeing tools about getting good family histories and other discussions about genomics and so forth. So again, I just wanna close by saying that with two points, one is to again congratulate you on pulling this meeting together. The other to Steve Furrow's point about how you get payment policy change. First, I would predict that we're gonna see payment policy changes, not this year for sure. Over the short to medium term future, because how we pay for services writ large makes no sense at all. Right now we pay for the more expensive version of radiation therapy for selected types of cancer, whether or not it's better. I'm not sure we're gonna be able to continue to afford that luxury. Obviously that means getting right to the root of the statute that Steve told you about. But more importantly, I actually did see a group of surgeons once go right to the hill and they got policy changed. This was in the context of a specific study. It was something like coverage with evidence development. In other words, you could only get the surgical procedure. This was lung volume reduction surgery for people with end stage emphysema. You could only get that covered if your center was part of the study and people were randomized and so forth. And the surgeons were quite concerned that the usual co-payments might actually bias who it was that would be enrolled in the trial. And someone very, very patiently from CMS explained that this was written into the statute and nothing could be done well after lunch. They all marched on down to the hill and within days that was changed. They got a waiver. So just know that I believe there will be payment policies changes coming anyway, but also that it can be done. I've actually seen it. So thank you for your attention. Happy to answer any questions. Thank you very much, Carolyn. It was terrific. So Dan first. I'm delighted. I think you said that you looked at Warfarin genetic testing and there was small value. And I'm delighted to hear that because I think Warfarin is the poster child that everybody looks at. And it has these warts that make it possible that the poster child won't perform when other things will perform. But I want to just pose the question of whenever I talk about this, I say, well, Warfarin genetics is really interesting. And for 80% of people, it doesn't make a hell of a difference at all. And it's only a small subset of people in whom getting the genetic testing makes any difference at all. You don't know who those people are up front. So you have to genetically test everybody. And we're moving to this idea of multiplexed pharmacogenetic testing. But how do you or how does CMS or how does FDA or how does anybody or Mwing better yet, how do you think about the problem of doing it? In a small number of people, there will be value for a genetic test. But when you start to do large randomized clinical trials, you might still not capture enough people to actually demonstrate a benefit across large populations. If you do it gene by gene or if you do it drug by drug, and then you have to demonstrate it sort of with many genes or many drugs. So it's a sort of conceptual problem because the gold standard is the randomized clinical trial or the observational trial in a large population. And yet, for this particular application, that just won't work. Well, let me be clear, though. I was talking about clinical benefit, not value. So in theory, if you can adequately and accurately identify the people most at risk of bleeding, those are the folks that you would be having coming in for more frequent testing. I don't know, you might actually go to your member of Congress said that CMS could expand. I guess you have expanded coverage for the home testing kits and so forth. And in the aggregate, you may or may not spend less money if you're not testing the 80% as often. Because right now, not knowing upfront, we just drive everyone crazy. But the issue isn't, again, whether the test is valuable. The question is, and this is effectiveness versus efficacy, the test works, right? The question is, when deployed in the wild, as you might say, can you actually demonstrate a value either clinically, economically, or both? And in this case, you could, but it was modest. In everybody, when actually there's only a small segment of people who really derive benefits, very reassuring. This was across a population. OK, yes. You understood that, and that's great. Because, yeah, we've got to ask you a question. Dress me by name here, I feel. OK, so JAMA had a nice editorial published online yesterday about the difference between evidence-based guidelines versus personalized medicine. I think this is sort of the crux of the discussion here. And that's a research thing that NIH can invest in, whether it's clinical trials, whether it's patient-centered outcome research, whether it's comparative effectiveness research, whether it's sort of evidence synthesis, all of the above. And I think the scenarios will vary. There are instances where you really need clinical trials, and there are instances where things will go on in practice and you will learn as you go along. I don't think we can make the case a priori that, oh, it's just too expensive to do clinical trials. Therefore, we're going to deploy things in practice and then we'll figure out how they work later. So as part of the implementation of genomic medicine, evaluation should be part and parcel of that. I think I'm going to make that case again this afternoon. Jeff and then Mark. I just wanted to echo what Muin said, and thank you, Carolyn, for being here. Because as NHGRI moves into this area of outcomes research and as we heard from military groups that are trying, and the VA, thinking about how to prove the value of genomics, which just seems like the real opportunity is to engage with groups like ARC and think about, take advantage of all the experience and methodology that you've developed to capture the kinds of data from real-world health care systems that will help us with those evaluations. So I hope we can really make that connection as we move forward. Thank you. Oh, one other thing I should just note. And we've had this conversation with Dr. Manolio a couple of times in the past. We do have some what I would call real-world laboratories. One is a direct connection to primary care practice-based research networks. These are networks of mostly small physician practices, which always makes it sound like the doctors are small. But I mean, it's the number of people in the practice who are small in the number. Yes, so it's less than five docs. The other is an ongoing network that represents a series of partnerships between what we call integrated delivery systems. And we take a very, very pluralistic view of that term linked with a research partner. So it is a way to actually get a toehold into health care. The other unique feature that we have is our statute gives us what I believe are unusual and possibly unique protections for patients, obviously, but also for providers who participate in studies. And I know that sometimes that's something of a barrier, because people are very afraid that under some freedom of information request or something along those lines that their identities could be revealed, which has kept some providers from wanting to participate in studies, say, of improving the accuracy of reading mammograms, right? What if they know my center isn't in this or me? I don't think so, but we actually have protections for that. So a couple of questions. One is fairly straightforward. The cancer registry project that you mentioned, I'm just curious to know if the data relating to the genomics and other information is being captured as structured data, discrete data that is minable using electronic methods. That's the easy question. The harder question relates to the value proposition, which I strongly endorse and agree with. However, at least for some groups, PCORI in particular, they have been statutorily prohibited from capturing economic data as part of their comparative effectiveness. And so I'd be interested in your perspective on whether or not that is something that ARC is able to do and how much we might be able to explore capturing some of the economic outcomes, which we obviously need to have if we're really going to define value. So for the first one, I'd have to refer you back to my CDC colleagues, and I'd be happy to get your contact information and make that connection. I think it's sort of self-evident that you would want to capture it in structured format if you could, but a lot of that is going to depend on the capacities of that local registry and so forth. Another challenge there, by the way, particularly if you'd like to know the impact of genetic tests, ultimately, both on the treatment and on the outcomes, is that so far, most studies looking at cancer registries in terms of longitudinal follow-up, the results have been very, very mixed, which is disappointing, fabulous intake of data at inception, but then the follow-up problem comes into play. PCORI is not actually explicitly forbidden from doing economic studies. What they are forbidden in doing is cost-effectiveness studies, and this is a statutory language you wouldn't want to parse in or spend too much time parsing, and I might just ask Naomi to jump in if I say something inappropriate or inaccurate here, because she serves on the methodology committee. The very specific prohibition refers to the use of the metric quality-adjusted life years, and I would say there's some rather confusing language in the statute, which I actually think is wrong, but it doesn't matter, because it's, I mean, it's scientifically incorrect, but it is in statute. I even know who wrote it, and I'm still not sure what he was getting at it besides rationing. Do you want to add anything? No, not to the parsing, but more general observation if appropriate, which is we've been talking about value and talking about payment, and all of this is going to change, and I don't know exactly how it's going to change, but it will change. So, you know, historically around value, I'd say certain pharma companies have had a lock on the term, so that value is, a new cancer treatment is always benchmarked at about $100,000 a year because that's what the original benchmark is. But I think as payers struggle more and more with this whole problem of how are we going to get access and as you know, Massachusetts has been kind of a pilot or an experiment for this, and they have been graded access, but cost is really tough. Definitely the move is toward value-based payment, and I think that's probably going to trend toward, it's certainly going to trend against our procedure by procedure, CPT code-driven payment structure that exists now. I'm not saying it's going to go away fast, but I don't think it can possibly be the future because it is unaffordable. I think as best as possible, there will be an effort to pay for outcomes and less interested in how you get there, but more interested in what you manage to do. I think that's really hard. I don't think anybody has the magic formula, but I do think it will be important for our thinking because I hear the concern around payment now, and I appreciate some of it, but all of this is taking place on the ground that's really shifting, and we need to think of where it's going. I would agree. Wayne, you're coming. Thank you. If I could sort of add on to what you just stated, we all are living in this age of austerity and limited resources, both in terms of money as well as human resources. One of the concerns that I have if you will allow this awkward metaphor is that with the sequencing having taken place, we still are just in the fertilization stage. Fertilization has occurred in this particular area, but we still are somewhere between the embryo and fetal stage in the development of this whole scientific area. I'm not sure I'd use that in a political arena, but that's just me. Thank you, I understand. I thought I was among friends here. Yeah, I'm kidding, I'm kidding. In any event, my concern has to do with the resources that would be necessary to update the information scheme. In other words, when an outcomes research study is completed and a determinant is made and published, my concern is that in the past, sometimes with the standards that have been put in place, even by AHRQ, new data has come forward. It really should be among those who are on the ground, the troops on the ground moving forward. There should be an update in the standard published, but sometimes it seems like it takes too long for that new information to be published so that people recognize that there is a new standard. And if I'd sort of give an illustration, I'm an endocrinologist. So when you look at hemoglobin A1C, hemoglobin A1C standards for community care were published. However, in the endocrine community, we recognized within five years after that publication that the standard that was placed was two orders of magnitude too high and should have been reduced even further in order to achieve even greater protection from complications. But it has taken a long time for that to be entered into the goals or the standards. Therefore, we see hospitals sticking their chest out saying that they've met the goal, but we recognize as practitioners that there's more information, validated information that should be published. And so this is difficult. As I started out, we have limited resources, so it takes a lot to, as you pointed out, redo studies, update studies, and then publish again that once something fills a slot or a cell, it's difficult to go back and say we should work on that again when you still have so many other cells that need to be filled. But that's something in the business plan that needs to be thought of because we have so much more information that's gonna come forward over the next decade and beyond that we have to be ready to go back and revise what we have established as a standard. But I totally agree with you. And to say that our processes for keeping these things updated and so forth, I mean, the task force is probably as good as it gets. I mean, in a standard consistent fashion because they've at least got a benchmark of every five years unless other evidence emerges and so forth. But even that is just fairly pathetic. They're just better than others. And, you know, frankly, historically, we've delegated a lot of this to professional organizations and so forth. But I mean, Naomi, I think, stated the imperative very clearly. The ground is gonna be shifting a lot more focus on value-based payments. The measures and metrics we got now are pretty clunky to put it mildly. So that's where the notion of a common infrastructure coming at this from the side of healthcare now becomes incredibly appealing because you've got a form or space to be able to have not only a systematic approach to research but also in terms of patient care. Those of you who live in this area and have heard the ads for Kaiser Permanente on the radio will know exactly what I mean. Do I think anyone has done that yet? No, maybe Intermountain might be getting close. I think VA in many ways is on the cusp of it or can see it over the horizon. But even there, there's still a lot of parts to be cleared out. Everyone else, boy, it's a mess. But at the same time, you're seeing more and more docs, hospitals and so forth adopting electronic records. So it's messy, but it ain't gonna stay that messy for very long, I don't believe. Additional comments? So I think the sessions this morning really have outlined a number of agendas that need to be addressed and how best to address them, I think, is something that we're gonna try to tackle a little bit more this afternoon. We do have a few other agencies to hear from and delighted that they'll be joining us after lunch.