 So Rex is coming up to lead this session. I'm just plugging in to be his scribe for the slides that we've prepared Okay, great so as is tradition at these meetings we have tried to capture more or less in real time a summary of each of the sessions and what we want to do is present the summary that we've captured back to you and And have you tell us a is it approximately right and B are there major things that you heard that we should have captured that we did not so that sound reasonable to everybody So the first session was really the basics of implementation science and I thought it was really for me It was a really wonderful session. I mean it really framed what it is that implementation science is all about so I learned a lot so We we heard about Defined implementation science dissemination research and implementation research and each of those phases and how they play out We learned about the timing of implementation research when the evidence is still evolving and I thought that was a really important point that We shouldn't just wait until we think we're ready and then implement We should figure out how to implement as we go along and I think I'm seeing some head shaking So I for me that was a really important lesson. It's it's because we're probably never going to be ready So we probably need to start right now and The lessons from the promo votes central line infection checklist. I also that I had heard that before but The idea that it needed to be modified at each site So it wasn't about the checklist, but it was really about the process of getting to the checklist and how that led to Better understanding at each of the sites We heard that there are dozens of models and frameworks hundred and six or something And can we work to develop a specialized in implementation framework for genomic medicine? And I think the answer to that is yes, we probably can We've got some examples of that, you know ignite is doing some of the work Caesar and a merger doing some work. We need to make sure those two are get aligned. I think appropriately But then to think about, you know, is it Reaim is it see for and maybe that's even not the right question to ask. Maybe it's about what are the right? Processes at each place and then we need to think about novel ways to fund this research outside of the slow NIH processes maybe clinical trials network kind of an approach I Thought that was sort of an interesting idea to think about could we For example amongst Emerging Caesar ignite all of those kinds of things stand up a rapid response team that could help Respond as we start to think about implementation challenges So a couple more points on the basics of implementation Could we expand implementation science into a larger sphere than just existing medical care systems? So that included the comment about direct to consumers We think that's going to become an important a more and more I Won't say important, but a more influential Factor going forward and then can we leverage payers decisions such as reimbursing for MSI testing to accomplish other goals such as getting to Lynch screen identification Lynch syndrome identification So I think that was it for the basics of implementation science. So let me stop there and pause and ask Did we get anything wrong and is there anything that we didn't include that we should have? Okay, I'm hearing silence. So I'm assuming that we're good to go Session two was to talk about resources for genomic medicine implementation We need to identify common threads of genomic medicine implementation to build implementation guidelines We saw a nice example. I think of of that in The work and the ignite spark program ignite spark program. Did I get that right? Yeah We need to think about some de-implementation guidance such as for long QT variants no longer deemed to be pathogenic. We heard about that Moving codeine from the pediatric formulary. We heard about that Payers paying for testing is a tiny fraction of the cost We need to have a broader view of what the real cost of implementation is And then question raised is do we really need a c-pick for non pgx genes, especially for non Mendelian conditions? Yeah So this also didn't and I would encourage everybody We sort of did Short shrift obviously to the CDC resources that are available to that since moving was unable to attend But everybody should take a look at those slides. I think there's quite a bit of Information in those slides if you haven't already so again resources Maybe I should just add because it's not captured here is There was a very nice discussion of what clin gen is and what clin var is that I think was is very helpful and There still continues to be I think significant confusion in the community about What the role of each of those things is so I think we still continue to have to do Significant education around around that again additions subtractions modifications One thing we didn't hear a lot about were other resources that that were needed You know there were lots of things about other partnerships and other efforts and and that we didn't hear a lot about other Resources and and maybe you don't think of them now But if you think of them later let us know remember that there will be meeting summary prepared my two colleagues have been Visually taking notes and we'll write a beautiful Summary of that so we'll capture more more detail, but but just kind of wanted to get the main points here Yeah, and I would just comment that a lot of the resource needs came out of this afternoon this morning session Around the informatics. I think we heard loud and clear that there are a bunch of Resources that are needed for that. So I just didn't comment because I'm gonna be getting to that. We'll get to that Yeah, okay, so session three we had Several really great examples. Oh, sorry. Hi. Sorry. Just one quick thing to add I missed this session. So it might not have been talked about but it rates the point. I just brought up a little while ago Which is a resource to access? Patient-level data phenotypes and genotypes That would just doesn't exist today and I think it a federated Network of sharing would be useful and I think the other really important point We talk a lot about standards that are critical whether it be fire and and others The best way to standardize is to to create a useful sharing environment The only way you can share data is your data looks like other people's and that's happened in ClinVar I've watched all of the clinical labs Align into a common set of approaches to track information on variants Because they're a force to to format it that way to submit it to ClinVar so I think that Thinking very strategically about data sharing not just because it's an incredibly important thing, but it also Forces very quickly standards instead of like taking years to develop standards with everybody having a different opinion and Then you put them out there and then no one adopts them If you just create a useful resource Bang it quickly gets implemented the standards because you can't interact with this resource without being standardized So just something to think about from a resource. I think very important And I think given the discussion in the last hour about the importance of the phenome that probably really emphasizes that because Probably the worst place we have for capturing standard information is about the phenome All right, so session three we heard that phenotype risk cores using EHR phenotype data To identify probable Mendelians that we thought that was a really impressive use of Electronic health records data. So how do we think about develop and disseminating and maybe this builds a little bit on the question of What kind of tools do we need that would actually Standardize that and we we heard a little bit in Lisa's presentation about You know how she went about doing that and her colleagues at Vanderbilt, so We need a comparison of weekly expenditures in advanced cancer care. We heard about this From Lincoln that it's an extremely useful monatric and that every category Nearly every category was reduced except drug costs So they set about trying to create a drug procurement team to try to address that and I thought that was a really novel approach Can we develop and share? Attractive data portals, so there's a lot of discussion about what that an interface that's usable and Maybe not as clunky as some of the ones that like our electronic health records that people have to interact with might be a useful tool to create Portals for both clinicians and for patients To develop a genome ed patient registry similar to cardiovascular data registry And again, I think this might be a step in the direction Heidi that of what you you were just talking about and then Try to really do more with patient reported outcomes We heard this morning from folks that The hospital systems listen to their patients. They If your system is like my system, they're constantly driven by You know, likeliness to recommend and all of these kinds of metrics that really drive think Hospital CEOs so patient reported outcomes are really very important to them. And how can we put genomic medicine as part of that? So Other things that we didn't and especially those that presented them, I'm happy to hear more My only comment about the the portals because we talk about this all the time How do we you know get the patient engaged and also the clinicians and I agree the MR is clunky But as soon as you tell patients to go outside the EMR, they're not going to use it and so That doesn't mean it can't look like it's in New York There's certainly a lot of things that we do an epic in our system that really is an epic But it kind of looks like it's an epic and then on the patient side as a health system had on You know, we don't want necessarily our patients going to five different Portals to get information about that it needs to have a unified theme because it's not just genomics Okay, I'm gonna go look at this But they need to know what they you know, what their you know, cardiovascular what their Nutritional support is or you know in terms of dietitian rate It has to be integrated somehow and still have the experience that the health system is Owning it even though the data may be coming from different cells I just wanted to have that comment. Okay, thanks But if I could build on that just for one moment I'd like to emphasize the the need in order to accomplish what was just outlined We need the ability to help incentivize the standardization of APIs Across these systems that needs to be developed by centralized authorities like HL7, etc It needs to be adopted by vendors And it needs to be open to the point where the innovators that are that are working on developing new Applications and new uses for these data are able to build on top of those and extend the the functionality in new and interesting ways and this gets to the conversation too about having some sort of statement or something about you know PHA and cover cuz again there was a conversation like one lawyer will say one thing We're we're having the same thing and you know we looked at what all of us is declaring as PHI now And we're revisiting well our review from a few years ago because it's different from the you know The position our internal counsel took a while ago, and so it's this back-and-forth. I think I can paralyze the system Yeah, and as we've heard multiple times, you know data sharing is key and so it can't be paralyzed by Unreasonable limitations on data sharing all right We had our really fun, and I thought not only fun, but and entertaining But it's a really effective debate that brought out a lot of really interesting issues that I think we can think about Geneticists still largely make diagnoses and refer back for management. So the primary care providers have to be involved So called catch-and-release which I found an interesting way to think about it There are rare serious disorders which are have got to be the realm of the trained genetics because they're Complicated they're not likely to be seen in a lot of primary care settings There was a lot of discussion about pharmacogenomics and not really being the realm of geneticists maybe Not being really the revolve realm of primary care physicians and maybe should be the realm of pharmacists But I thought interesting discussion around that There was a lot of discussion about shifting more of the genetics care to genetic counselors The problem is there are a limited number of them, and I I think certainly at our place We've doubled the size of our genetic counseling program and still Everybody gets a job before they graduate. So it's it's a If you want have friends that are looking for a career. It's a great career And I think the other thing that is key here is We need to figure out how to bill for that because right now the building for genetic counseling is Probably a little on the broken side or maybe not even existent We need to develop limited training for the majority of common complex diseases So maybe we have to have training programs that allow people primary care providers to get certified as Sort of experts in a particular area and the point was made It doesn't make sense for every primary care provider to do that and just wait for some rare Person to come into their office but rather maybe what we can use is the primary care providers that have had this additional certification as Extenders to the genetics the trained genetics workforce So that hospital by hospital hospitals that don't really have these resources might at least have somebody that could know When it's important to refer and when it might be able to be handled locally We talked about seeding relevant specialties with the needed information Beyond the primary care providers and then the question came up Can some of the diagnostics and we just had a nice discussion about that be done by AI often the genome reveals along with The answer along with the line of clinicians including geneticists So you know the idea of lexie tell me what's wrong with me It's is an interesting is an interesting and maybe not too crazy idea that we should be thinking about What? Were there enough conditionals there So mark mark is raising is mark subtle yep So there are a number of bullets here that I think relate to an overarching theme that that would be important to take away At least from my perspective from implementation, which is we really need innovative care delivery models We can't default to the way we usually do things and so in some ways having a bullet saying what we need more genetic counselors Yes, but we also need to explore other types of care providers like genomic medicine assistants and advanced practice nurses and that they can You know serve in different roles. We also need to You know the AI is mentioned from a diagnostic perspective, but I think there's a lot of other things that Codra is working on as part of clean gen about information delivery and and removing You know humans from the equation in many cases where it's really not needed because of the nature of the information So, you know something that just reflects a broader Potentially researchable agenda about innovative care deliveries to support Okay, thanks Any so I think that was it for any other questions comments additions on that topic I would just add with what mark said though to make the innovative care models that includes innovative and new models of Care and utilization of genetic counselors. So perhaps, you know, maybe it's the Economy of scale of a couple of genetic counselors doing laboratory utilization versus the one-on-one Pre-test hour-long pre-test genetic counseling that so innovative care models Just making sure that includes innovative care models for genetic counseling as well Well really for all for all aspects though. I mean it's for the laboratory. It's for everything, right? Jeff I think we heard Gail talk about the fact that medical students hear about genomic in like the first two weeks of medical school And then it's like virtually absent And so how do we I guess the question is for this topic is how do we make sure that the primary care? Community is really prepared to take on their role in genomic medicine I think I guess I would just so at our place. It is true that if you look at Lectures that are labeled, you know genetics or genomics. It wasn't even as extensive as what Gail showed but if you look at the Presentations in Cardiology or in cancer. There's a lot of genetics and genomics that happens in those contexts, so I'm not I'm not disagreeing with you, but I'm just saying we need to it's not just You know lectures that are labeled genetics We just need to make there's a probe make sure there's appropriate training throughout the life cycle of a medical students There needs to be a thread and and actually the practical aspects of it And when you get into practice of how do you even order a test? You know all the things that we've probably talked about in implementation is Is somewhat lacking from and I think should be on the agenda Greatly lacking again in our place the concept of genetics is mostly genomics. There's no pot is mostly Mendelian genetics There's no almost no population genetics at all in our program. So So it's a real weakness All right a session for this morning EHR is an implementation So I think we heard a really nice presentation about how improving CDS is is limited by the lack of both institutional evidence institutional acceptance of supporting evidence And then that the CDS architecture itself varies so it sometimes creates a barrier between Individual for at individual sites There's a tension between operational IT and those terrible researchers going in and mucking up with the Operational IT system. So I I think we've all experienced that at our various sites We need to think about a defined framework of stakeholders transactions and clinical systems and how those interact to better I use that better use of the electronic health records and clinical decision support and The great opportunity of using clinical decision support for patient screening. I think that's a really important Activity going forward that would help us better do clinical trials and get better evidence for a variety of things We heard a lot of discussion about the needs to standardize technology variant specifications We heard about some Systems that are really important resources including fire We heard about the HL7 clinical genomics workgroup, which is likely to bring forward some important information to help us going forward We heard about the importance of the clinch in NCBI allele registry and how we need to Integrate that and and and I think maybe now is a good time to say and make sure we're integrating it with the phenotype as well Dan keeps talking about adding a phenotype column, and I think it's a really important contribution We we heard again from several people the need to embrace a common data model Can we how do we go about doing that? And then we also heard though that the everybody wants to use a common data model But the problem is everybody wants to use their common data model And so we need to come to some agreement about what the common data model might might be we heard about Odyssey and I think I think it was Chris who said but maybe it was maybe it was Larry said that Fire really may be the ultimate CDM specification Using some kind of a pluripotent data model. So maybe there's even hope that there's a target that we could be driving towards Going forward to that. I think that was it for EHR. So Sure, people will want to add or subtract to that Gil One of the things that I thought was kind of a nice development is the domain analysis model document for clinical sequencing and then more recently the clinical genomics Because it has a basically a standard list of the use cases in genomics. So it's like for the first time we can talk about Okay, here's a standard use case and it's workflow and so forth and and I've seen that being used by not just institutions by national programs by Other consortiums Yeah, and if I can add to that so I agree with with that resources It's a really good resource. I would say that the one thing I didn't see up there was this notion of establishing a community-driven policymaking Decision-making kind of committee to do things like validate that HL7 use case document. I was there, you know, I was there for a lot of sessions of how those things get created and There's there's more experts in this room than combined of what was used to make that So that gives you a sense. It's it's not it's good. I'm not trying to knock it It's just I think you guys want to have a group out there that blesses it and says this is good And we agree with it that will get more people to engage with it and adopt it quicker So anything you can do along that lines to get that going would be really fantastic Okay, Chris isn't here. So I'll channel him I think he would say that we've got our nomenclature confused a bit I think what he was really saying was that while there are common data models that are out there What we want to do is to move away from a common data model and to use data elements Which is somewhat captured in the fire bullet fire. Maybe the ultimate common data model specification I don't think he would say it exactly that way I think what he would say is is that it's a pluripotent data model that actually Allows for much more flexibility than our traditional approach of using a common data model So I think that's the takeaway that I had from his presentation in particular and I happen to agree with him I'm not sure if that's Agreement among the the majority of our other informatic experts here The second thing that I didn't see represented there again reflects comments that I made during the strategic planning Talk which was gills presentation about the pace patient-facing aspects of that I think that some of it is a little bit implied, but I heard that as being much more Explicit in terms of how we can actually use fire as a patient engagement tool around genomic medicine Okay Yep If I could just add one point to what mark just said about the common data model and the data elements One of the challenges that we have in spades right now is that when we have a problem That's related to data sharing that needs to get solved. It's typically because it's on some sort of timeline We have a project. We have a grant We have some sort of need to exchange data and the way that that happens is two parties or a small handful of parties get together And they come up with the common data model That's going to work for that research network or that whatever it is and they come up with a Solution and it gets the job done and that's great, but it's fit for purpose. It's relatively short time time Lifespan, and it doesn't really go beyond that what we need collectively is a longer term or sustainable Thought process about how we develop those Lego building blocks those common data elements that are going to provide the common Semantics for how we can exchange these data elements for a much longer period of time one of the challenges in genomics is the rapidity with which the domain evolves and We cannot be in a situation where we're continually having to Redesign our data models every time we come up with a new use for this data or we are never going to get anywhere So how would you go about doing that? It's a challenge It's not easy if it if it were we would have been done 10 years ago One of the the so I mentioned the challenge about the fit for purpose And the fact that the drivers for a lot of these efforts are typically Relatively short in terms of lifespan compared to what we're really needing in this domain Which is a much longer more stable Platform of data elements that we can build off of One of the challenges and I think I'll look to my other colleagues in the SDO is that that can perhaps speak to this One of the challenges that we have is that when we have Participants coming to the table to help us develop these things as part of a more formal process Everyone comes with their own use cases and their own Objectives in mind of what they want to accomplish and it takes a tremendous amount of effort to take those specialized use cases Generalize them so that they apply far beyond those those driving projects and Then make sure that we have enough Common understanding about what we're trying to represent that it will stand the test of time That is extremely labor intensive it involves conversations that literally last for hours If not days Larry's probably laughing over there at me, but it it's absolutely true And so I I think being able to take the time to breathe and to have those conversations will allow us to Develop standards that will be much longer lived Then some of the ad hoc standards that we have today, and I think at the risk of maybe saying what's obvious My guess is that there's probably an implementation science approach that could be applied to how to actually develop that Yes, right, and I want to Bob made it sound like So big that you guys might not want to even touch it So I wanted to try to go and say that I think the approach that I would take is that with this whole domain There are parts of that model that we've spent enough time on now that are getting pretty solid And the part is getting it over the goal line so people could start using it And that's what I think we should be doing just like you doing a software project when you're getting requirements You sort of you get the requirements and then when the requirements get solid enough you you say okay We're ready to risk development resources on actually building this piece because it's it's stable enough and we can then Develop on it. That's sort of what we have to do here And there are parts that are close to the goal line We just need to push them over the end and we need a little bit investment to a little bit You know we need investment to get it over that goal line and then build on the successes Okay Just any anything else on the HR? I was gonna add one more. It's just that tools for supporting tools for learning from others. So whether that's artifacts or Webinars to share lessons learned and and also The Areas for trying out Decision support before it's actually implemented talk for a while Rex Well, I think that was about trying to make more tools available to people to think about How to share and then the second piece was clinical decision support somehow a test platform for before it goes live a Sandbox as it were. Yeah So then Session five this morning. I just found Fascinating and I hope all of you found it as fascinating as I did it it completely flipped my way of thinking about this so The title of the session was like, you know, how much evidence do we need and The message that I heard from the folks here is we need to stop worrying about evidence. We've got evidence We just need to figure out how to use it. So We're gonna stop thinking about Deficits and evidence. We're gonna continue to develop more evidence, but we're not gonna be held hostage by oh, you don't have enough evidence yet The other thing I thought just completely Revolutionized my way of thinking about this is you know We've been thinking about oh, we need to get the payers on board And I thought the case was so compellingly made today that that's not the right question the right question is how do we get the people that are some the employers on board so that They're you know, they're representing their employees and we heard they're motivated to actually do the right thing For their employees and oh by the way, if you have the big picture of things, it's probably gonna save some money So I I found that just again I think it really we should really be thinking about building on that and so for example, we need to think about Research that shows the employers benefit from adoption of genomic medicine and And the other thing that came out of that is we probably should it whether it's one of these meetings or a smaller meeting need to convene a group of employers and really Understand this even better than we understood it today and maybe including believers and then maybe some non-believers who might be brought along as believers in in the course of that discussion and then the other lesson, I think they came out very clearly Or need that came out very clearly of the Discussion this morning was or this afternoon. I guess it was the need to develop a basic genomics formulary and There's already an example that was generated several years ago now In the UK probably needs some updating But maybe we can even think about start that as a starting point to think about what a basic genomics Formulary might look like so I'll stop there and Does this capture what's missing? Pat yeah, so I would just have a little bit more My impression was a little bit more nuanced with the evidence Evidence exists stop focusing on it and I see Terry that you already put a little parenthetical in there. I think if we're going to Disintermediate the payer and go directly to the employer. There's still need evidence I heard lots of examples from Jay of where peer-reviewed publication was persuasive So I think it's the timing of the evidence They may have a lower threshold for wanting to implement than payers who tend to focus on clinical utility And and so I would just say that it's still important But it's how it's packaged for example whether we include indirect costs Which would be very important but typically most of our economic evaluations focus on direct health care costs And the other piece that I think is missing is I also heard that Consumers are another important audience and they are going to drive a lot of the decisions around genomic medicine So it's not just skipping over the payers and focusing on purchases But there's going to be a lot of pull-through from consumers who ultimately are all patients at some point in their lives Right and that we need to also focus on them And then I heard a lot of agreement around sort of a major barrier Not just being lack of reimbursement, which we kind of have always heard but that that that whole Need to educate that everyone involved in the implementation process Otherwise all of this wonderful evidence is going to fall flat Just real quick to say that you know the evidence thing I think The first statement makes sense at least what we heard But it makes sense to me in the context where pockets are deep When the context where pockets are not deep, which is a lot of medicine I think that Having the evidence the scientific evidence is still a good Concrete milepost that we can hold out to people Payers or whomever to get them on board. Yeah, I just want to make it clear I don't think we were trying to say we don't need to continue to gather evidence I think it's just we need to make sure we're not held hostage by Some sense that we don't have evidence Mark the one other perspective that I heard That of course wasn't necessary wasn't represented here But we do still have issues on the public payer side with Medicaid and Medicare where the traditional rules of engagement still hold and for which There's still a huge Percentage of the American public that is funded under those programs So that's something that needs to be considered although I think that's probably the toughest of the nuts to crack Unfortunately because it encompasses one of the issues that Bob raised which is they don't have deep pockets And there's an inherent conservatism in terms of the things that they do now The one point that I would make in you know in contradiction perhaps to Bob is that The reality much as I made the comment that clinicians use evidence as a canard to say I just don't want to change things In many cases payers use Evidence as a canard for things that they don't want to deal with or pay for The example I always love to use when I talk to insurers is to say you know you hit us about evidence evidence evidence when K-RAS came out for I Can never remember which Abbott is a satuximab or whatever You guys all covered that overnight because you realize that with that one test you could not give this expensive medication in 90% So you used a very different criteria of your evidence level for something that's going to save you a ton of money So let's not you know, let's Understand that there's a certain hypocrisy here So you know we do need to recognize that we have to have venues where we can really honestly Talk about are we really talking about evidence? Are we really talking about the risk the financial risk that people that payers have about being out in front of the Curve and paying for things that are in fact highly unlikely to return value But the reality is is that a lot of what they pay for in health care now that doesn't have an ohm associated with it Is also not delivering value Completely agree And I think it's probably Important to say that it's not always the case that Medicaid and Medicare are the last to adopt. They're often actually the first and so There's maybe hope there if we can get private payers to adopt because their Employee-based systems are telling them to that'll also help on the public side Let's keep going Alana So I wanted to just remind her that it's not so my I guess my issue with the Stop focusing on the evidence is the What kind of evidence just because it's happening doesn't mean no evidence is being generated So can we what kind of evidence do we need? Can we how can we utilize the evidence that is being? Generated so this bullet need to need research showing employers benefit from adoption what we were talking about in that session They're generating evidence, so let's use that evidence and how do we how do we take the research and the information from that Evidence from the people that are clinically implementing Finding those places that are generating evidence the direct consumer industry. What evidence can we get from it? So obviously I was being a little too glib Impacted me Heidi So the bullet on we should convene a group of employers, so I think it's more than convene I think you know unlike the health care systems that are part of the academic community. They're used to publishing and Have statisticians and looking at data I don't the motivations for the employers is a little different and I'm not sure that they have access to The or motivations to get this data published and put it out there and so you know And I was talking to Eric at lunch about this. Is there a way that NHGRI could could fund and Resource of and a group of people that could actually work with the employers and help them Ensure they're you know rolling things out in a way that the data can be captured You know have the expertise to do the statistics and the analyses on it and get it published And then they have the as Joe, you know papers to point to to actually Justify what they're doing and get it rolled out further But you know just convening them in a two-day meeting and then sending them off I'm not sure is enough to actually really make use of what is likely to be an incredible resource I think I think that's a really important point that I would also say that we One of the things that we heard relative to health care system adoption of Genomic medicine was they often rely heavily on consultants and so I think it was maybe Jeff Who recommended maybe we should think about? Using the resources available broadly in NHGRI as a consulting team that could go in and provide additional Support to the people that are at least willing to consider Implementation if I if I can respond to Heidi's comment I would hazard a guess that with the exception maybe of Klesson And are the NIH NIH folks we all work for employers that are self-insured Academic medical centers are for the most part self-insured. We have a mission of academic. We have an academic mission I think that there's alignment to look at those when I say hey, we're in the publishing business anyway. Why not? Take a look at this Yeah, I think about I mean this motivated me to go back talk to my CEO about that they're you know 8,000 employee health system that we should be talking about this with At Heidi's point, too I think there's an opportunity here as you know narrow networks take you know hold of areas in terms of you know Different health systems certainly in the Chicago land. That's something with a lot of you know Large organizations that have their headquarters and the creation of those but that could be a perfect opportunity for that And I can say that There are a couple of projects. I know that are going on at our place That are doing just that working with like for example Boeing and some of the larger Employers and to actually think about how can we do more preventive care to actually help and so this would fit in perfectly Yeah, sorry one more thing the other thing that Jay referenced is the idea that you know Self-employed insurers are already aggregating together to tackle issues like purchasing power and Sharing quality improvement metrics and those sorts of things there'd be the potential for Taking you know an idea that Jeff had proposed about you know creating this Genome add type of thing within a consortium of self-employed insurers where there is richer data and that they would actually serve as the Consortium that would pull that data together to define outcomes and do that so wouldn't always fall on you know the resources that are You know limited from NHGRI and I have a delimited time frame, so it's another interesting model to potentially explore So I think Gil are you next I think are you Okay, Robert. Yeah, I just think we should be very careful about some of this sentiment because I Really think that NIH and NHGRI and and the academic traditions of evidence generation are a kind of bulwark against over-enthusiasm Marketing and so forth and I don't think I mean certainly a geisinger or a self-insured academic center is one thing, but a Self-insured company for the most part doesn't have the resources traditions thinking processes skill sets to to generate the kind of what we call evidence so Absolutely, let's let's reward and cheer on early adopters Let's encourage You know experience with evidence generation at all levels, but let's let's reserve for NHGRI the the focus on the highest levels of Evidence the most rigorous methodologies. I hear in this maybe I'm maybe I'm just hearing it and it's not true But I'm hearing in this a kind of I don't know rush to Reward experience rather than rigor and I just want to Yeah, since I was sort of the author of that I that was not the intention the intention was I was just struck by the fact that None of them said We don't have it with the evidence was a problem I mean not that we can't always have more and it can't be better And so that I I'm guilty. I hear everybody saying that I overstated it I mean they funded people who looked at things that the world was saying the world was not ready for In each year I has to be ready to fund things that say wait a minute You know because because it's marketed because it's adopted and because people like it it may not be effective I mean, it's it's got to be able to reserve that sense of equipoise in both directions Good point. What else we're now down. We're now to two slides. Yeah, so Oh I'd forgotten we weren't done with that. That's okay. No, it was one slide. Okay made it to all you were talking Because of all the adjustments that people good suggested. All right, so let's move on We had a I thought a really good discussion about the NHGRI strategic plan and there will be This is just this is just the beginning folks So we heard some interesting ideas about can we develop partnerships with regulatory agencies and payers to clear needs of priorities for evidence generation, so Again, we're not done with evidence generation Are we shifting? Are we shifting goals of research from high value publications to convincing payers? And I think we've heard even in the discussion just now that we need to continue to do high value publications as well We need to define what NHGRI can own and not what it can just partner with that was a pretty strong theme We need economic studies for preemptive testing And we've all Dabbled in thinking about the economics and know that we're not trained to do that So we really need to get some people that are involved We need to improve standards standardization of genome related phenome questions I think we've talked already quite a bit about that we need to increase emphasis on the last mile problem of the clinician interacting with the patient and We heard not to forget the babies, so what is not here that should be or what needs to be adjusted the second the last bullet Again, I think it reflects a Traditional way of thinking I think it's really the patient interacting with the clinician again. I think you know one of the point Was made is that we need to be more patient focused. I didn't see that you know come through and in terms of some of the comments that that I made There was one other thing But I'll think on it and I think Bruce you are one of the people that made the comment about the last the need to deal with the last mile Last mile doesn't have to be that a clinician interacts with the patient. It could be direct interaction with individuals And then the other point that was the medicine-based evidence. Oh, yeah I also heard some comments that were made about the longitudinal use of the data and being able to have infrastructure to be able to Use the genomic information Over a lifetime or have ways to be able to access the information from multiple different points as well as beyond just the Interaction between the clinicians and the patient Ways of utilizing technology to be able to make the information more accessible to patients in general and involving them more in the process of care and as well as using that same type of technology to be able to look at ways to Impact many of the points that are labeled here as far as the care process and Implementation thinking about the actual process itself and are there ways that we can use technology in place of the way that we currently implement genomic medicine other comments or or thoughts Yeah, I'm not quite seeing other to the I don't know if this is what's meant by the the last mile those longer term outcomes of the impact of Genomic medicine. Yeah, I think the last mile problem. No, the last mile problem was something else. It's about the interaction Yeah so say what you just said again your longer term outcomes of Implementing of implementing genomic medicine the longer term impact on patients on providers and how they How to improve their use over time? Through the engagement with them. I'll fix it in a second. Yeah Okay, anything else? so I think This has been very helpful and glad to get all of the feedback and revisions We're Traditionally from these meetings. We try to generate a manuscript We will think about what that might look like one idea that Terry had suggested Was to go back and look at the GM one meeting Manuscript and to see what the world looks like a few years later And I think that might be an interesting construct to think about the Outcome of what we've talked about at this meeting. So we'll we'll think on it We'll seek some input. We'll get the this Summary out to you as well as some additional summary notes that have been taken and I just want to say I you know I We when we started organizing this meeting it was it was sort of like hmm. What are we going to do for GM 11? But I hope you have found this as informative and sort of eye-opening as I have it's I think it's been really Terrific and it really gets to the foundations of what it is to do genomic medicine implementation. So Thank to thanks to all of you for your a sticking around to the really bitter end And then also to all your input during this and we look forward to figuring out what the next steps are So, thank you all and I don't know and also special. Thanks to Teji Rocker Burris Julia Walker and Camilla Celia and Gabe Yeah, I read too much about the Royals And also Kiera and Alvaro who will be putting together the the videos from in our AV colleagues from the hotel We'll be putting together the videos and they'll end up on the genomic medicine Working group webpage of NHGRI in the meeting section. So probably within the next two weeks or so They'll be there in terms of a paper typically what we've done is to have people who were presenters and moderators Be sort of the co-authors and in that if people feel that's not right Let me know and we'll you know, we'll see if we can expand that a bit and I did want to make sure Teji Did you send around the slides for moving query? Did those get? Okay, great. So those will come out to you. You'll see what what resources there are for for him and Eric any closing thoughts from you? Thanks to you. Thanks to you Rex Terry, of course, it's thanks to Terry The root of everything. Yes. Thank you Rex for helping to co-chair this worked out of the entire Genomic Medicine Working Group who are instrumental in organizing it and everybody for participating great now We can apply. Thank you And those of you that are staying for the meeting we'll see you around and those of you that are traveling safe travels home