 If not, then I'll go ahead and introduce our next speaker if that's all right. So Dr. Joe Selby is the Executive Director of the Patient-Centered Outcomes Research Institute or PCORI. Before he was doing that, he was the Director of the Division of Research I think at Kaiser Permanente, Northern California, where I knew him and had the joy of working with him in cardiovascular epidemiology studies. And he's going to come and tell us about comparative effectiveness research and how we can work together. Okay. Good afternoon. I want to thank Eric and Terry and NHGRI and all of you for the invitation to be here to introduce you, if you haven't already met us, to a non-federal organization but created by the Affordable Care Act and a substantial funder of research, nothing in comparison and NIH but yet larger than most other non-federal funding agencies. And the point that I bring to this meeting in this audience is that PCORI for a variety of reasons sees a very special interest in funding research of a certain type that is directly related to the concerns you have that of genetic therapies, genetically directed therapies and genetically derived diagnostic tests and risk prediction markers for several reasons. I'll say also that my being here grew out of a meeting that we had at PCORI, Eric and Terry and Jeff Ginsberg visited us in late June and we decided that I should come in and tell you a bit about what we're up to and also they also invited me at that time to a meeting of Ignite, which we attended last week, Jeff, and found a lot of enthusiasm there, a lot of because of the somewhat more clinical end of the spectrum of the work that Ignite does, they found a lot in common with us. Terry, I'm really glad I was here for Terry's talks for two reasons. Number one, I got a much better understanding of the mission of NHGRI and of your out of necessity in some ways, your tendency to study genomics, genetics in the broad view rather than the disease specific view and now I understand the challenge you face with the other whatever it is, 30 plus institutes. And I asked myself for a minute whether we really had that much to talk about because I think we pretty much are going to be pretty much on the one disease at a time focus. But as I followed through Terry's presentation, she came to a number of items, bullets that I said really would be in PCORI's wheelhouse. One thing I learned was the notion of bed back to bench research. I think that's something I say all the time, although I didn't have a phrase like that for it. The notion of being interested in coverage with evidence development, just to give you a clue, that is right at the center of what we'd love to be funding. And the other bullet that you put up was about identifying the outcomes that matter to patients. So because we are the patient centered outcomes research institute, we've sort of hitched our wagon to the notion that patients have questions of their own. So today I want to introduce PCORI, define CER, PCORI style and tell you a bit about PCORI funding mechanisms. I want to then elaborate a bit on our interests in supporting comparative effectiveness research in precision medicine. And finally, PCORI.net got mentioned before, but I want to say a bit more about PCORI.net as a piece of the national infrastructure potentially. So from the Affordable Care Act, our purpose is to assist patients, clinicians, purchasers and policy makers in making informed health decisions. And we do this through conducting research, through synthesizing research and through disseminating research findings. So all three of those activities. But we've come to think of ourselves as the stakeholder driven research institute. The one that stakeholders that can't get their questions answered elsewhere would turn to, we put a lot of time, effort and resources into engaging with the whole spectrum of researchers from patients right up through payers and the research community, policy makers and the research community. To get the questions, just as Terry said, to get the questions and identify the outcomes that matter to those groups. We require them to be involved in the studies that get done. And why might I wind up here rather than at some other broadly directed institute at the NIH? Well, the legislation went on to tell us that our research ought to be designed to look for the potential that there are actually differences in the effectiveness of healthcare, treatment services and other items in various subpopulations defined by age, race, gender, ethnicity, presence or absence of different comorbidities and genetic and molecular subtypes. So we are told right in the legislation to pay attention to what we typically call treatment heterogeneity. That something works in some people and doesn't work in others. Our further thinking has gone beyond that, which would really be an instance, I think, of bedside back to bench observing treatment heterogeneity in clinical populations to helping to identify for individual patients what the best treatment option would be for them. And their genetic testing and genetically derived therapies would be really high on the list of items that patients could use information about. And the interesting, one of the interesting things is that we are constrained to funding a type of research called comparative clinical effectiveness research. And so that means we don't typically fund epidemiologic studies, even studies that look to define causal relationships. We're not the institute to identify the relationship between genotype and phenotype, for example. We are sort of stuck with the paradigm of comparing two or more options in the area of screening, diagnosis, or treatment for outcomes that matter to patients. So the comparisons need to matter to patients. So this has to be something they're currently considering. And the outcomes need to matter to patients. That typically just means we broaden our range of outcomes and sometimes we cut out some of the intermediate outcomes that cost additional money to collect. We go for simpler studies, but in general we have probably a broader range of outcomes than many trials would have. Our research needs to be conducted in real world populations and settings. And as we've already said, it needs to attend to differences in effectiveness and preferences across patient subgroups. One important point, though, is that when we're comparing these two or more options, some people think we actually have to have two different active options, like screening with this test versus screening with that. No, it can very clearly, in our view, be screening versus usual care. It could be screening with a genetic test versus the usual clinical prediction algorithm, for example, that would be another one. But it can sometimes be screening or diagnosing or treating, treating compared to not screening, diagnosing, or treating. So we have three mechanisms right now. And I think this is what we're gonna have for the next four or five years. The broad announcements are like R01s at NIH. They're about $500,000 in direct costs a year. They tend to have a limit of three years. We have a little flexibility there. Any clinical area that an investigator brings to us, as long as they persuade us that it's important to patients, that it is CER, that they've got patients and stakeholders engaged, and that it's likely to change practice, we'll take it up. We've funded about 275 of these studies to date. And so these are our most clearly investigator initiated line of research. The next are targeted. So this is sort of at the other extreme, really, where there's a single clinical area, really, with some fairly narrow questions. There may be two or three questions, and we may fund a cluster of three or four studies. Much larger amounts are put on the table. We've put as much as, I think, right now we've put as much as $30 million, but we are preparing to put as much as $50 million onto the table for a single set of projects. So big funding, I'll show you the targeted funding questions to date. And then there's something that's a bit in between called the pragmatic clinical studies. So in the process of identifying our targeted announcements, we identify other high priority topics that we don't issue targeted announcements on. Rather, we put those topics on a list, and that list is about 25 topics long now, and we issue a pragmatic announcement, pragmatic clinical studies announcement, every six months, and fund grand awards that are up to about $10 million in direct costs, so they can be large as well. This I just want to show you, it's not the ideal slide, I couldn't find the ideal slide this morning, but this just shows you our revenues. Our revenues began in 2011, and so the first column, first bars, the green is revenues, and the blue is expenditures. Okay, well I'll just keep coughing then, that's all right. So you'll see that we have taken in about $836 million and committed most of that, and just spent about $210 through 2014. 2015 we took in, oh thank you so much, we bring in about $400 to $500 million per year, and our spending necessarily lags a bit because we can make multi-year commitments, so we don't have to spend every penny we bring in each year. The spending is estimated to peak in 2017, and you'll see that the money stops coming in right now in 2019, so we aim to have all of that money committed, but there will be a tail of spending out through 2014, 2024 I mean, but the main message here, I wish I could show you the commitments, the commitments is what new monies will allocate per year, you know over the next several years it'll be in the range of $300 to $500 million per year, tapering in 2018 and 2019, but enough that it's worth our conversation. This is the way we get to targeted announcements and pragmatic clinical studies, we take stakeholder input, and I want to just say that in the area of genetic testing and genetic therapies, we've had a lot of stakeholder input from patient organizations from clinical specialties, and from payers, a lot from payers expressing concern about what, you know, the submerged part of the iceberg that they're expecting to run into shortly. We screen them to make sure these questions really are CER and they're not already answered or not already funded into substantial degree by others. We put them on a list and we take them to our board, the board can approve them for the development, the board is the scientific oversight committee of the board, and they can approve them for development of topic briefs, we bring them back to the SOC, they approve them for advancement to the advisory panel. We have multi-stakeholder advisory panels with everybody from patients to payers, a number of researchers, clinical specialists, etc., and they prioritize them. We take that prioritized list back once more to the SOC and they approve the refinement, and right now we are here. So we've been through the advisory panels, we are in the process of refining precision medicine topics right now, so we are essentially poised to make a large funding announcement, but I will admit that we are still looking for some really good ideas. Now I may have come to the wrong place just in that you're not disease specific and I think a lot of these will probably be smaller disease specific studies, but some of the things Terry mentioned like for example whole genome sequencing versus more targeted sequencing makes sense. So that's how we prioritize topics and precision medicine has done very well. I think I've already told you about the pragmatic studies. Two cycles a year, six to nine awards per cycle, up to 90 million dollars per cycle or 10 million dollars per study, five years duration typically, and the maximum direct cost per project 10 million. The only thing, the only other thing we really emphasize is that these researchers need to work closely with major stakeholder organizations as they're doing the study because we want it to be actually understood, agreed in advance that the study is being done well, the right question, so that when the results are available we will have partners to help disseminate the findings. These are the targeted awards and I won't read them all but I'll just let you, you see that in the later ones the hepatitis C, the novel oral anticoagulants, treatment resistant depression, we're getting more and more specific. The questions are getting tighter. The first ones are like treatment options for African Americans and Hispanic Latinos with uncontrolled asthma was a much broader kind of an announcement. But I want to draw your attention to a central message here and that is that two of them with the double stars, that is the multifactorial fall injury prevention strategy in older persons, it's called STRIDE and the hypertension disparities reduction awards in African American and rural populations were are co-administered by with an Institute of NIH, NIA for the falls in the elderly, NHLBI and NINDS for the hypertension disparities reduction awards. We have had great success, okay? We have had great success in these two collaborative studies with the NIH. We also showed one there with the Agency for Healthcare Research and Quality. But in the STRIDE study, the prevention of falls in the elderly, PCORI put $30 million into this and a five-year award was made in early 2014 for $30 million to a multi-center consortium and they put together a multi-site cluster randomized clinical trial complete with a one-year pilot stage. The pilot has been completed and the main trial is now underway. Now what NIA brought to this initially was expertise both in the literature of the falls prevention but also in mounting large-scale interventional trials like this. So they ran it, they solicited the research and ran it but we participate with them in every aspect of it and sit on the steering committee of the study. Really a wonderful collaboration. The other one with the heart, lung and blood and NINDS, we think it is going equally well. So the message here is that we are very interested and we see the value of a collaboration with other funding agencies particularly with institutes here. So that hopefully will come up a bit in the discussion but I think when Eric heard it he was interested and we'd love to discuss that further. So a bit about our view of precision medicine that the products of precision medicine in our probably naive view include new therapies based on discovery of the impact of inherited or acquired genetic variation in disease susceptibility and also on markers for predicting disease risk or treatment response that can help guide treatment choices. But we also certainly adhere to the notion that precision medicine has to include things other than genetic factors such as age and life expectancy, gender, socioeconomic status, literacy, comorbid conditions and also this curious notion of patient preferences that really to get down to what works best for individuals one needs to be able to factor in patient preferences as well. So this is my notion of bedside back to bench and the complementarity of CER and precision medicine. So on the left is basic science, early clinical development and testing and maybe phase three trials even for the medications and at a certain point it comes into the real world of clinical care and it may be in the form of genetic tests or targeted medications and then comparative effectiveness comes into play. Comparative effectiveness can evaluate both genetic markers and it can also evaluate the effectiveness of new therapies but in addition if you do comparative effectiveness the way we prescribe with attention to treatment heterogeneity it may identify some of those clues that lead one back to the laboratory. So our role would be rather limited in some ways that most new diagnostics and new targeted therapies ought to ultimately prove their value to patients compared with current practices. I paused for a minute when I saw the the marker for Stevens-Johnson syndrome that 44 of 44 people had the allele and I said would I really stick to my guns that we needed an effectiveness trial there and I think I wouldn't but we know that most genetic tests don't end up quite to be quite that predictive but comparative effectiveness research is typically needed when new therapies appear and proposed to replace current standards. The benefits may differ from what was done pre-approval the adverse effects may have been too rare or distant in time to be detected pre-approval and ongoing efforts are needed to identify those patient subgroups who do benefit those who don't. Predictability is also not the same as effectiveness I sometimes run into people that don't buy into this and my caveat about Stevens-Johnson syndrome holds here but inaccurate prediction can certainly cause harm and the utility of the new test might need to be compared with current prediction tools. Although I did at the ignite meeting you'll be pleased to know that I was corrected I suggested that sometimes clinical prediction rules are just as good or the genetic tests don't add anything to clinical prediction rules and somebody pointed out to me that that might not be quite the right question that if the genetic test was simpler to administer than a complex clinical prediction rule you might I might have asked the wrong question or used the wrong gold standard and evaluated the new genetic test so I learned from that exchange. We've conducted a landscape about whether a landscape review about whether it's time for comparative effectiveness research in precision medicine. We started with engaging Catherine Phillips from UCSF and an assistant at Bakori and they did structured interviews with nine precision medicine experts. I don't know somebody in this room might have been interviewed I don't have the list of interviewees here. Questions like should Bakori move forward with precision medicine and initiative what's your rationale for having us to go forward etc and I want to show you some of the findings from that. Should we move forward almost unanimous agreement that it's time that we ought to move forward the rationale was that clinicians are starting to use precision medicine now little is known about whether or which precision medicine approach is best and how much precision medicine contribute over and above conventional approaches so that was what the interviewee said of what approach should we take it should we should do a scoping review and limited outreach to stakeholders avoid systematic reviews and other time wait time intensive activities get on with it issue a relatively targeted funding announcement we call ours no this is the usual we usually call ours PFAs but in issue a relatively large targeted funding announcement not completely open although there was not much consensus about what the studies were that we should pursue which types of studies a slim majority favored CER study of specific interventions there was some support for infrastructure development and quite a bit of support for studies of supporting patient decision making about these little support for method studies in this area and little agreement on the specific topics one top one person's top priority appeared to be another's lowest some did some support for concentrating on chronic conditions other than cancer on population screening cancer got back in pharmacogenetics and infectious diseases we were advised to stay engaged closely with NIH's with the precision medicine initiative but pursue our own agenda and so that's just trying to convey the notion that we are working hard to get to the point where we have a precision medicine initiative that we feel could be complementary to work that you do and complementary to the work of the precision medicine initiative as it comes out I look forward to your comments a few words in closing on picornet which is a very large investment it's now made by PCORI to the tune of over 260 million dollars over the last beginning about two years ago now and extending three years into the future we started with 11 clinical data research networks we're now up to 13 the principal investigator of one of those 11 clinical data research networks is here Lucila ono machado heads the cdrn called p scanner which brings in the five uc five of the uc university california medical center systems the va and a number of community partners as well clinical data research networks by definition draw data from electronic health records and other sources on at least a million persons most of them including p scanner go well above a million vanderbilt is also a key player now in partnership with duke and a very large community-based health care delivery system the mail clinic kaiser permanente a very large federal federally qualified health centers network the city of new york all the academic medical centers in new york are banded together the city of chicago all the academic centers in chicago are banded together data has now been standardized at least put into a standard data model on more than 70 million persons the key currency at this point is the electronic health record data and we're working hard to bring in claims and patient reported information as well also in picornetta now 21 now 20 patient powered research networks groups of patients with single conditions half of them rare conditions half of them common conditions their job is to in in insert a strong focus on patients and the patient's perspective in everything that the core net does their job is also to grow and to partner within the cdr in so that we build larger and larger communities of research ready patients with single conditions and a coordinator center which which supports these networks works hard at standardizing the data and provides the logistical assistance for the meetings the goals of picornetta are are distinct on the one hand it's to improve the nation's capacity to conduct large-scale clinical research more efficiently both randomized trials and large observational studies are on the on the docket and this is one of the few networks that aims to be able to do both and at the same time to support a learning u.s. health care system to support the use of that data right in the health care delivery systems from which the data is coming so with the patients the clinicians the delivery system leaders and payers as well involved really bend the the data and the system in the network toward answering questions about the efficiencies of health care delivery so a very ambitious set of goals we're linked closely with sentinel which is an fda and supported insurance claims-based network so we have the same coordinating center we have overlapping populations and we are working closely with sentinel to try an efficient way of bringing claims data to the table we also work very closely with the ctsas because we share an interest in being able to do a clinical trials and to use electronic health record data within academic medical centers that that triumvirate I think is really essential to making sure that we have one national large-scale population-based network rather than three and I think the leadership of all three is very committed to try to make that a reality having put all of this together we really see ourselves as good partners for a wide range of registries because there will be overlap of any registry with this very large population-based set of data PCORnet sentinel and ctsas together would bring the electronic health record data the registries though may bring more patient reported data this is our operational structure in the middle are the networks over at the left is the rest of the steering committee which includes the networks of course but it also includes the patient council it includes funding agencies the NIH FDA our CDC are all represented there Josie Briggs from let's see I'm not going to remember the acronym that NCAM now has but Josie Briggs I'm sure you all know represents NIH we also include CMS the Office of the National Coordinator and ASPE because they control access to a large amount of data that's crucial representatives of medical product and device manufacturers are on the steering committee as is PCOR itself in the coordinating center on the right are task forces and this is a little outdated we're doing things a bit differently now but we do have a number of the CDRNs and a couple of the PPRNs have biorepositories we have not put a lot of effort or time to put a lot of effort yet into assessing these but there's a lot of interest within PCORnet obviously within ctsas as well in harnessing biorepositories and being able to link our repository data with electronic health record data and claims data also a lot of work going in in partnership with ctsas on data privacy issues and on IRB oversight these are crucial areas that Terry mentioned that as well so in closing and this is the over to you slide we do have a natural interest in studying the products of precision medicine and we think it is a virtuous cycle that we'd have something to offer to precision medicine as well that PCOR is very interested in collaborations with other funding agencies particularly when that collaborative partner brings specialized expertise in a content area as NHGRI most certainly would and PCORnet and similar population based EMR driven networks with strong patient involvement in governance are promising and exciting places to do the kinds of research that that interest in NHGRI so I thank you and would hope that there are some comments advice correction suggestions questions or in moderate if you think I see a question yes so I understand that PCOR is not supposed to ultimately consider costs do you monitor costs we tend to monitor resource utilization and not costs costs are a very difficult thing to get in any objective way out of a delivery system some delivery systems have a database that has costs in it but what those costs mean and how that compares to costs in another setting are almost they're predictably inconsistent in in in extreme ways so if one collects all the resources that are used and maybe adds to that an assessment of patients for the out-of-pocket expenses you can then apply standard costs and really compare the total consequences of a treatment decision in one system versus another so for for a variety of reasons we don't measure costs but one the biggest one really honestly is that the data are not particularly useful but you monitor but we definitely monitor resource utilization in theory one could back out costs well that's what we recommend that in fact you you standardize you apply standard costs so that you can really you know so that you really have a meaningful comparison between how things are done at UC San Diego and how things are done in Kaiser Permanente I can tell you that Kaiser Permanente's cost database it you know from third from 27 years there it really doesn't mean anything in terms of that you would be interested in knowing it doesn't mean anything about what it really cost we had initially and I'm not sure the message got to Dan and Bob we thought maybe you guys might want to be commenting on this so I might ask you to comment when you have a chance but I might also ask while you're thinking of that one of the major cultural icons of of an HGRI so we're very strongly I was just curious what approach you guys use to data sharing so we are in in fact working with that at the board level now so not just about the cornet but about PCORI funded studies in general we want to have a we want to position ourselves near the front near the front of the open science movement in general you know we also are a small agency which is good in some ways we don't have tens of thousands of awards like the NIH has so we don't have to go back and talk to tens of thousands of awardees or or break the news to a thousand per cycle but you know we we want to do it in step with other agencies but we will we are working on it now we will have an open science policy and it very definitely will have data sharing as an expectation not always brought to fruition you know it will be basically data sharing if data are requested we we couldn't afford to capture and store all the data from every study but we want people to be prepared and we provide some resources for awardees to prepare their data for sharing which means get the protocols ready we we really actually do want to i'm quite sure part of our policy will involve making protocols publicly available and then sharing the data upon request and in selected studies and that will then PCORnet is also working on its policy but it will be consistent with PCORI's. Lucila? Well perhaps is new maybe it's why you didn't mention the targeted sequencing effectiveness trial from UCSF on breast cancer i think that's one that most relates to this group. Why would you would you like to describe it or you want me to you might do a better job than i? Well essentially they are comparing targeted sequencing the value on outcomes so they have a control trial there okay so so i'll jump in now they're they're comparing targeted sequencing as as one piece of evidence to calculate a woman's risk for breast cancer to guide the decision about how often i have mammography so this is a 55 000 person randomized trial one arm gets a recommendation from their delivery system to have mammography every year from the age of 40 to 74 75 the other arm gets talked to and counseled about their risk and gets the genetic test the targeted sequencing tests and gets an assessment of their preferences and then they make their decision so it's really like taking the us preventive services task force recommendation and putting into the test which which says talk to your doctor assess your risk and your preferences and decide whether you're going to get screened that's what they do and the outcomes are incidents of advanced disease and ultimately mortality from breast cancer so it's a very exciting study and thank you for mentioning it and is that only brca one in two testing or are they doing panels i think it's more than brc one in two i lucidly you may know that yeah and that's already underway or being it's just getting started laura esserman at the cancer center at ucsf as a principal investigator excellent so the the um one of the questions that always comes up at least in my mind is is what is it that the quarry is funding so are you funding the sequencing for example in that particular is that part of the budget or is the sequencing viewed as i've always been i've been under the impression perhaps mistakenly that that what you're particularly interested in the comparative effectiveness research domain are are comparing two reasonable treatment strategies both of which are sort of out there and available for for for physicians to use and so you sort of help us figure out which is the right one but is the sequencing for example being paid for by the quarry or is that being paid for by carriers and and that because that's an i am almost sure that that's being paid for outside of picori i'm almost certain of it and we may but we may actually be for example talking with cms about they're covering it we have another study of they don't have anything to do with genetics particularly but a proton beam therapy a very expensive treatment for breast cancer and we're talking with cms and ensures about covering its use in a clinical trial i mean one of the and so in general but but you're pointing to an in really interesting um um delimited picori faces um you'd think that perhaps we really are studying only two approved treatments both of which are already covered by uh insurers but that really doesn't answer many of the most burning questions and we're really driven and we're pushed towards questions a hepatitis c was a big area um so um we we need some post-marketing outcome studies of hepatitis c both uh one drug versus another and also screening versus watchful waiting or testing a treating versus watchful waiting so we are pushed towards a type of comparative effectiveness research which really is effectiveness it's a real world but it's not necessarily two treatments both of which are already standard practice i mean if i can if i can give you some sort of free advice that you can take or leave one of the barriers that we identify in the june with medicine working group is this is this issue of who's going to pay for all this at the end of the day and and the fallback position often is well if we could figure out a way of at least getting the the trial funded through cms and there was a period for example when warfarin testing would be paid for if you were part of a study and not if you're not part of the study so it seems to me that you have the opportunity to to help us figure out what's going to be effective and not by negotiating with cms or other payers to to help us uh compare to effect to treat to treatments it also that's one comment take it or leave it the other the other uh question i had was in that breast cancer study what's usual treatment what's the usual care arm yeah is there a usual care the usual care arm like the both of those were recommendation to have annual no i think they annual annual yes and um the the um that's not you know i think that because they're because they're and and i'm not familiar with it enough to tell you the ins and outs but but my suspicion has been that their fear was that um if they did it every two years uh and and they showed no difference somebody'd say yes but certain women should have it every year and so other comments oh maybe i misunderstood quite get this but uh you mentioned as a potential barrier with uh collaboration with n h g r i is the fact that you usually focus on single disorders well we we we typically focus on uh clinical questions and clinical questions often express themselves uh in the form of um you know endocrinologists ask about a question in patients with diabetes etc oncologist etc um yeah well i just wondered there what the kind of the definition of a single disease was in light of the the genetic heterogeneity of most disorder and and that would seem to be a great area where n h g r i and uh picori could interact well i i mean i'm persuaded and i want to and i want to be persuaded and i am that there that there are areas um where we could find uh overlap of interests um beyond just single disease studies or single heterogeneous disease study and then the other thing in in my experience of writing grants to different uh n i h institutes is one thing that they don't seem to like to pay for very much at least in my experience is is uh phenotyping of of patients and different cohorts and i wondered if uh if picori has interest in that which would greatly help some of the genetic heterogeneity issues and how they relate to it well go Karen it it is a good question i don't have a a um an answer for that's very different than the previous one which is that typically we tend not to pay for diagnostic tests etc um if it's particularly if it's part of a clinical treatment algorithm or diagnosis workup but um you know i would not say never because sometimes it may be just essential and it's just very clear that we're not going to be able to get it funded otherwise i mean we've tended to say if this is close enough to clinical to being out there and used in clinical practice even if it's not covered delivery systems or payers ought to be willing to cover it in a study because it's very soon going to be and then it'll be too late um but you know there may they may and we could actually explore particular um questions where that might be a part of the question could be and we'd have to we'd have to work with you on it yeah i think what you know the point that you raise val about about and that you did this as well joe about disease specificity i mean we struggle with that around this table whenever we try to get into into a clinical paradigm it's almost always for specific disease there there is very little that clinicians do this general health you know and so uh you know maybe newborn screening is the is the one example you know the one exception but but it is a big channel so i've been looking at your the PCORI website and there's a task force on biorepositories and so i'm wondering about the biorepository work that's going on within PCORI is that just for in-network groups that's a that's a PCOR net task force okay yes yes uh and it's um headquartered at duke um and as i said it's been fairly quiet but i think it's going to pick up rather abruptly in the next few months and those samples and would not be accessible to outside researchers at this point the answer is is no i think PCOR net is because we did not pay for the collection of those samples so we don't have any leverage over them you know we're we're hoping that kind of the funding of PCOR net can enable researchers at an institution that has a repository to make better use of their repositories data for PCOR net overall we really do have an aim that PCOR net will be open to outside researchers now the i think the even then though the biorepository part of it since we didn't pay for it is is going to have to be a negotiated arrangement we certainly do hope though that that PCOR net as a as a research resource is used not only by people in the 50 or 60 institutions that are part of PCOR net but um but in um you know by the broader research communities i mean that's part of what we're putting investment in for so i mean as i as i listed this and think about it some more and of course we you know we have a PCOR net center so you're about vaguely about about that at our place that it seems to me that what you are what your major focus is is to answer single important pressing questions for practice so is it better to do this or that this or that colonoscopy versus occult blood screening or falls this way or falls that way and it also seems to me as i look at terry that some of what we're doing in the genomic medicine space is getting pretty close to you and so the opportunity is to engage some large uh collection like you have like you've harnessed to and it's up to us i guess to identify those important questions so we can put that on the table for our genomic medicine working group and come up with the three or four really pressing large-scale questions in genomic medicine we could then get you to answer for it i didn't say it that way or together we'll answer them together right right right any other comments joe thank you so much a pleasure thanks terry thanks eric everybody