 Good afternoon, everyone. I'd like to welcome all of you to session five of our program. And the topic for session five is establishing and sustaining genomic learning healthcare systems. Let me introduce our two moderators, Dr. Aaron Ramos and Dr. Crystal Sosie. Aaron. Thanks, Pat. Yeah, it really is a privilege to co-moderate this session with Dr. Sosie. I will introduce our three speakers, and then Dr. Sosie will moderate the discussion. As we near a close, I'll attempt a short summary of major themes and ideally proposed solutions per Terry's request. So our first speaker is Dr. Howard McLeod from Intermountain Healthcare. Howard's talk is focused on generating evidence of effectiveness and value. Thanks, Howard. Pleasure to be here, and hopefully slides will start appearing soon. We can see your slide. There we go. All set. They're in the proper mode. They are. Great. So I'm going to talk a little bit about the learning health system approach at Intermountain. A couple of general comments. One of them is the institution has to want to learn and want to iterate and want to improve. That's one of the big keys. And the second, there's a big difference between employed physicians and affiliate physicians. And we've seen even within Intermountain, the learning happens differently depending on the engagement with the institution. So I just dropped those two in there before I forget about them as we're going forward. Also, the work that I'm describing reflects the entire eight months that I've been at Intermountain, but also stands on the shoulders of what Mark Williams built, Steve Blile, Lincoln Nidal, Derek Costlem, and what Nephi Walton was currently doing here at the institution. Intermountain is a large institution, right now 33 hospitals and almost 400 clinics that are distributed across what is functionally nine states. You can see the distribution there. It's in the Mountain State. So it's the places that you're lucky to live in or lucky to build a visit once or twice a year or so. And so it's a large integrated health system, a lot of value based care, but also a lot of fee for service and other models that happen within that. And before anyone asks, with the merger with the SCL healthcare system, we have both a large center and a large epic presence. So, and no, they don't communicate very well with each with each other. The institution talks about helping people live the healthiest lives possible. And so we integrate the through precision health aspect to that, as very much the theme of what we're trying to accomplish and with with genomics, as well as similar type of tools. And then as a bit of a last backgroundy type piece, Intermountain precision genomics or Intermountain precision health is a part of this large health system, about 200 caregivers distributed across the Intermountain footprint. There's clinical care that's going on there's a genomics laboratory doing work there implementation work. And some patient driven discovery research. And a lot of the precision medicine workers in the context of value based care as as the way forward and a lot of the learning is how do we optimize value as an endpoint, as we're going forward. One, one last little pieces we've broken up a lot of the clinical activity into a small number of sectors. Well, some of them are not so small like primary care and pediatrics. Others are are much more focused. But these are the areas in which we implement precision medicine, realizing that many of them go across multiple different specialties or especially the areas. But that's the administrative structure when we do the top down type of work of going forward and know that's not a tuba or a toilet as the model for ICU. It's there. I'll briefly go over a couple of the initiatives, focusing on the value and the and the financial piece. So I'm not going to go deep into each of them but just to highlight what they are, and some of the learning we've made that has helped us shape that. In the tumor testing, the pharmacogenomics a little bit on on multiple cancer early detection, and then finish up with a large population health approach, and the way that's affecting the way we learn about delivering genomic care. The cancer model is what you'd expect. You have in this case a man with metastatic lung cancer. He's run out of options. There's no clear next option based on national guidelines or FDA approvals. He spit he wants some sort of treatment, genomic analysis identifies something that might be clinically actionable, but there's no guidelines for that. It goes to a molecular tumor board, which identifies eligibility for a particular trial. That that information from the molecular tumor board is used to influence the discussion with the insurance company to pay for that unusual situation in this case, melanoma therapy for a lung cancer. And the, you know, things go forward. And so this sort of model, you know, a lot of the value is brought by the therapeutic evaluation by the molecular tumor board. And it's not adequate just to have genomic data. You need to have the medical interpretation of that data that that goes hand in hand with the laboratory piece in order to go forward and we've shown data others have shown data where not only is there about a doubling and survival of these patients that get a precision approach versus a standard approach, but there's a cost savings and $734 may not seem that big of a cost savings. But this and this is total cost of care, because remember we have the entire office patients medical care. So their total cost of care is $734 favorably, when they have precision medicine, well that's per patient per week. So you then can scale that out to about an $80 million savings that occurs per year in in our health system. And so that's sort of data where we have genomics for a clinical reason, get the economics within our system, and then turn that back into growing the, the, the element of, of application is is one of the ways that we're doing this going forward. Neonotology is another area. There's a lot of great work that's been done. I'm happy to meet up dimension at the bottom there that I personally feel Stephen King's more deserve a statue erected somewhere of him for all he's done in this in this field, but our own efforts with that primary public city have identified a number of different examples where genomics identified therapy and one one really pointed example is a young man had seizures 31 hospitalization said at the time he was three years old. We spent how one and a half million dollars on his care genomic, a rapid hold genome identified a vitamin B six processing gene mutation started him on high doses of over the counter vitamin B six. And he has not had a seizure since spend several years now so that the idea that we can do this in the end, and if we're owning the cost of care. We'd rather spend $13,000 than $1.5 million, even if it's a small number of cases that are coming through and so again, that type of data, learning from the clinical situation, having the economics, driving more care, because we see the total picture is a is an important part of it. The pharmacogenomics piece is also important, certainly starting with anxiety depression. It's been implemented not just in behavioral health but also cross primary care and pediatrics. And that's been been something that has been well received as paid for by the insurance companies that we work with. But it's also now where we can go and take data from, for example, this is the the infamous Kentucky retired teachers paper that came out earlier this year. There was a reduction per year and direct medical charges of about $2,625 per person in this particular trial. We went and looked at how many people would be eligible for this type of analysis, based on that data. We assumed a 10% adoption rate, which is very conservative. We cut the savings in half, and it would still save us right out almost $50 million per year in implementation. So this kind of work where we look at what is out there. Can we replicate it? Can we use it? Can we apply the savings is a really important part of being a learning health system on the genomic side. The clinical care is what drives us to do it in the first place, but it's the economics that causes the Darth Vader's of our system to want to play with us and make this happen. Because often the reimbursement directly may not be adequate, but the savings is being held by others within the health system. And when that's taken into account, we can make much better decisions for the entire health system. Closing in on these examples, one that we've just started up is with the multiple cancer early detection testing. This is something where we have to be part of some of the clinical trials that were done initially that, and instead of doing it through the research office, we did them through primary care, because we wanted to learn how this would play out in the field. And now on high accuracy, high specificity, it has now changed where this is now a covered benefit. So it's not being paid for by insurance, it's being paid for by our HR department where every employee that's 50 years or older in age will get one test every two years, 100% covered by our health system. So far, we've had about 1% positive rate of had multiple thousands that have done the testing, about 1% positive rate. All of them have been confirmed today so we have not had a false positive yet, although we did during the clinical trial so it is possible. But again, the idea that we can see value in the health system in this case for employees for ourselves. And have that as something that's implemented and covered, in this case through health benefits, other cases through the insurance coverage. It's just a way to try to make sure this sort of thing is not languishing in the journals, rather, is being applied for ourselves and for our patients. The last little piece I want to go through is about something that was started as a research study. And it was, it was something called heredity gene. It is a collaboration between Intermountain Healthcare and deco genetics, which is owned by Amgen. According to the height that the PR department put together it's the largest genome project ever. But I would say it's not even the largest I put the asterisk on because it's not even the largest genome project mentioned today in this meeting. It's still plenty large but up to 500,000 people, all with electronic medical record, looking at at the issues of prediction prevention reducing cost, etc. So, to date, we've enrolled closing in on 150,000 just over 135 as a few weeks ago. Now it's closing on 150,000 people that are enrolled. This includes a pediatric component through primary children's as well as two other sites. We have participants from even though we're only enrolling in the Intermountain footprint, we have participants who have a home address in 49 of the 50 states. So if you are from Delaware, please come visit us and enroll in the study, because we hate that we're missing one state of the US. We have population, healthy folks, specialty enrollment, all going across there. One of the key things is this is not turned up to be just just research. We're based on Nephi Walton's work, have a full iterative loop, where when we learn something from the ACMG key genes or about 170 genes total that were currently of returning. We're going to those variants in the whole genome analysis that's being performed. We then confirm them with the patient that indeed they are with a focus testing that indeed are real. We then have a clinic that works with those patients, makes referrals as necessary. To date, there's been about 3700 patients that have been served clinically by what was a research study and I would say based on our experience, we really should not be doing large population research studies without returning results and having it turn into clinical care, because the expense while real was small, and it turned into not only patients lives, in some cases lives being saved, but also events that were then reimbursable based on insurance, etc. And so, again, an iteration of learning that helped us turn research into practice, which then turns around and informs us more as we're going forward. The saying at the bottom there it's not enough to know there is risk, we must act on it. I think in this day and age there's really no reason why health systems doing research can't have that full loop. It's just personal laziness on my part that I would not want to do it has nothing to do with the practicalities of what could be done. And so I think that's really should be the standard of error that is happening going going forward. And then the last piece I want to circle back to, as I mentioned we have 33 hospitals 400 clinics. That's fine to do the various things I talked about. But there's a large diffusion of innovation that occurs on average it's about 17 years according to literature before innovation is fully infused across these systems. And geography is a major impact on the level of diffusion. Many of our 400 centers are in small places. Don't don't necessarily get visit by the drug rep or by the, the equipment rep, the device rep of the thing, people who kind of tried to drive the, the innovation from the company standpoint. And so we have a lot of activity now, making sure those 33 hospitals equipped to administer precision medicine, and to be the sounding board for the 400 affiliated medical centers that are distributed across the states, so that a small clinic in Wyoming can have access to precision medicine expertise through telemedicine and other sources, rather than just an approach where we, we do things at our major centers, where you see the big blue dots, and then everyone else is just kind of out of luck, because they, they went to the wrong place so anyway I went very fast. Happy to, to take any clarifying questions and then look forward to being part of the panel when it, when it goes live. Thank you Howard those were some terrific examples and insights I think keeping an eye on the time. Let's put clarifying questions in the chat so you don't forget them. And Howard you can start thinking about your responses, and we will have our next speaker Dr. Dr Nancy Mendelsen from Optum Frontiers Therapies and Optum Health Solutions. Nancy will focus on genomic learning health systems and payers. Thanks Nancy. Sure. Good afternoon everybody and thank you so much for having me join you today I'm truly honored to be here. It's very exciting to see how much has happened in the last several years and I really appreciated your talk Howard and getting, getting genomics into the health system across across so many different hospitals it's great. I'm going to the next slide please. What I'd like to talk to you guys today is a little bit about what is United Health Group who we are, how policy is developed the different just mentioned the different levels of evidence which you guys probably understand much more than I do. To give you some examples about how we have had external engagement with genomics and rare disease doing some enterprise wide data analysis and an ability to try and move the field forward, because I think there's a lot going on that people just aren't aware of. And the question that was posed to me by by Terry and Pat was how do we leverage a genomic learning health system to use data differently and support people with rare diseases try and engage the payers. And I don't think I have an answer but I have plenty of questions I have a few ideas and I welcome everybody else's thoughts. Go ahead. Next slide please. The United Health Group is two large businesses. For those of you that aren't aware, we have to, they're very distinct and but complimentary business platforms. The United Health Care which is the insurance entity and then Optum and Optum is really a information and technology business that tries to enable health services across multiple different insurance businesses so not just United Health Care, but all kinds of insurance entities. And we, the United Health offers a full spec off, excuse me United Health Care offers a full spectrum of health benefits programs for people for individuals employers Medicare and Medicaid. So you can do the next slide please. So what I mean by that is policy does not necessarily mean coverage policy influences coverage and coverage is determined by each insurance entity so if we can influence the center for Medicare and Medicaid services we will influence policy more broadly. In fact, United Health Care is divided up into Medicare and retirement community and state which is generally Medicaid services and then what we call employment and individual. That includes United Health Care broadly fully insured and administrative services only administrative services only are companies where United Health Care provides the administrative capabilities for their insurance, but we don't write their policy they write their own policies and most big companies that's how they get their health insurance. One thing to keep in mind is United Health Care serves about 25 million individuals in the country. United Health group serves some are closer to 125 million people in the country so overall we touch one of three Americans. If we can influence policy as clinical geneticists more broadly at the federal level, then we can indirectly be able to influence more of the private insurers, those in the administrative services only. The policy itself, as you're well aware is determined from evidence based medicine and in general, randomized clinical control trials which is very hard and genetics and particularly in rare diseases. And it's also influenced by evidence based healthcare practice guidelines like those that are developed by ACMG or the AP or COG. So all of these things together influence policy but don't determine it in a black and white fashion. Next slide please. So there are some things that can be done to partner with Optum and or United Health Care overall. In general, I think most people know about this study that was done by Grace Yang and Annie Kennedy and Grace through the Lewin Foundation and Annie through the Every Life Foundation. The Lewin group is actually an Optum company. And the data that they used is the Optum data. So I think this study where they show the impact of rare disease and the importance of across the genetics healthcare is really moving the field because they're proving what we all know, which is that rare disease is expensive and trying to get a diagnosis is a green field and we need to pay attention to this area because in my opinion it's a green field. We haven't done much to take waste out of the system and try and reduce the healthcare costs. Next slide. The same group is looking at a set of six diseases to try and compare the economic impact of early diagnosis and treatment versus not. And the results of this study are pending. So they're going to the next slide. These are the seven rare diseases that they're looking at. They're going to calculate the direct and indirect medical costs of each disease at different milestones and compare those that have had early diagnosis versus late diagnosis. Those that have had treatment versus not had treatment and there are several people in our genetic community that are participating advising as experts. And I'm very excited about this study because I think it will hopefully have as large of an impact as their previous study. Next slide please. There's also something that has been developed in our Optum Genomics group called the evidence engine. And what the evidence engine is doing is they're working with particular labs and that have developed tests and trying to prove clinical utility. So it's real world evidence. They are working towards analytic validity to check the accuracy of the laboratory process used to perform the test and hopefully eventually clinical validity. So really the ability to test correctly and identify the health condition or disorder. And they also are using the Optum data, which is quite large. Go ahead. Next slide please. So the question that I would pose to this group is, is there a third circle? How do we take a basic, you know, this great paradigm that has been developed to take the healthcare learning system within particular health systems and partner with insurance companies or partner with Optum and work more broadly to think about a framework, a framework for rare diseases, a framework where there might be less stringent criteria for diagnosis and treatment. What evidence do we need to develop a policy and state of guidelines for people with rare disorders? Is there something other than randomized clinical control trials that we can prove? Can we as a community agree on what the outcome should be that we're measuring after gene therapy or how we determine a change in therapy and the natural history of diseases? How do we determine what the best outcome is for a rare disease? And you guys have talked a lot today about the implementation science and decision tools and how important that is. So I think looking across the healthcare system, thinking about patient experience and population screening, and mostly if we can impact CMS or Medicaid rules, even though they vary from state to state, I think we'll have the ability then to impact the insurers. So I'll just pause there and see if people have general questions and I look forward to the conversation. Thanks so much Nancy. We do in fact have a little bit of time for any clarifying questions. I think the questions I'm seeing in the chat, we might want to hold until the discussion, but I do see Terry's hand is raised. Yes. Thanks. Thanks so much Nancy. I wonder when you mentioned that policy doesn't equal coverage. Could you just expand a little bit on what you mean by policy and how the research community would contribute to it? Yeah, I mean, I think there's general policy and then there are each individual insurance products that dictate whether or not coverage happens. So you may have a policy for whole genome sequencing that says whole genome sequencing is equal to whole exome sequencing in terms of ability and coverage, but payment and coverage does not include whole genome. So some companies, some individual ASOs may elect to have whole genome sequencing covered. So it may be in what they call their particular insurance document that dictates that coverage is there, but the policy in general doesn't include whole genome sequencing if that's helpful. And how we can influence that as a community is to and to change policy is to have very clear outcomes and guidelines that show when we should be using a particular test or a particular capability. So we don't have a lot of guidelines about when to use whole genome versus whole exome and that I think would go a long way towards influencing policy if that's helpful. Very much so. Thank you. Thank you. And Nancy, I had a quick clarifying question as well. You talked about the new study that's evaluating the improved outcomes and economic impact focusing on those seven rare diseases. And just say a little bit about the rationale or criteria for selecting those. Yeah, I was, I'm not part of that study directly. So, I mean, I advise them so I can't speak to it. I think the idea was to try and get a smattering of diseases that some are on newborn screening, some are not. Some we have good treatment for some we don't so that we can have a broader understanding of the impact of diagnosis and treatment. Perfect. Thank you. Okay, that was terrific. Let's move on to our final speaker and then we'll have plenty of time for discussion. So it's my pleasure to introduce Dr. Darrell Pritchard from the personalized medicine coalition. Darrell will cover progress and the integration of personalized medicine and common metrics. Thanks Darrell. I'll follow up the discussions about the learning health system and payers and about generating evidence of effectiveness and value that Howard discussed with a short presentation on the progress toward the integration of personalized medicine within the broader US healthcare system, especially regarding metrics related to a genomic learning health system. So first, I want to thank our host NHGRI for leading the genomics medicine work stream and this 14th edition with special thanks to Pat Diverka and Terry Monoglio for inviting me to participate to give this presentation to the group. Thanks guys. Thank you Aaron and Crystal and thanks to Renee Ryder and the whole group of all of you for all of your efforts to lead this meeting into advanced genomics and clinical settings really appreciate this effort. I as Aaron had introduced me and Darrell Pritchard senior vice president of science policy at the personalized medicine coalition based in Washington DC. I'm going to describe an effort to assess the current landscape of clinical integration of personalized medicine and genomics in the US healthcare system. Now during his opening keynote presentation yesterday morning Peter Hewlett introduced this effort he provided a couple of slides that I that might seem familiar to you because I have them in my slide deck as well. And he briefly discussed how it provides a measure of readiness for a genomic learning health system and a measure of progress in genomics integration needed to further build out the learning health system. And this is key to the conversation how these measures that I'm going to discuss really kind of show the readiness and progress for genomics integration in a learning health system. The initiative was designed to look at integration throughout the US healthcare system at large so it's not just early adopters and academic health centers centers with well funded research programs, but rather it includes a representative sample of healthcare delivery institutions across the United States. So, the objective of the study was to assess the adoption of personalized medicine across health systems in the US. What makes it different is it includes a broad survey across different hospital systems, different geographies therapeutic areas and adoption levels, and involves a multifactorial definition of personalized medicine or genomics adoption. Beyond just testing it includes data utilization, consistency data sharing leadership funding health equity and other things. A critical outcome of the study was the development of a novel maturity model to design to objectively measure personalized medicine adoption across the health systems published in the journal of personalized medicine last year. And it involves interrogation of each system's unique unique fingerprint on various metrics which can highlight specific challenges and potential solutions. And hopefully this will be insightful for the development of a learning health system. Learned through many presentations to not bury the lead so I'm going to provide the findings overall first and then get a little bit more into detail. What we were able to do was develop that novel maturity model designed to really objectively measure genomics adoption across health systems. On the top line, about 83% of institutions studied scored a two or higher on a five point scale. When we examine their integration efforts and that's, that's important to note because that means that personalized medicine and genomics is being recognized and adopted at some level, pretty much across the board in the US at this system so we, we were having some one level of genomics adoption throughout the system and that's a positive. However, only 22% of institutions scored a four or five on the personalized medicine integration scale. And I think as I described the scale a little bit more. You'll see that these, this is where we need to get institutions to be in order to be considered part of a learning health system. The distribution of the overall level of personalized medicine integration was broken down by different types of healthcare institutions, different practice types and demographics, as well as other criteria. So just quickly looking at the setup so we know who were what kind of systems we're looking at when we're talking about adoption of personalized medicine and genomics. Virtually all of the respondents were actively involved in personalized medicine initiatives. And they were representing a various set of roles within the healthcare system most, most of the respondents were lab directors, or C suite executives and different health delivery system administrative groups. And you can see here on the right. The breakdown by type of health system whether it was a health system and independent hospital or an integrated delivery network such as Intermountain Healthcare and Dyson. They are represented pretty much as are represented across the country as well. And the graphics of the different systems that were involved. You can see the breakdown, we had roughly about a third of our respondents were academic health centers a third were community teaching hospitals and a little more than a third were community non teaching hospitals and that's important as we look at the differences that may occur in adoption between these types of of systems. So we ranged from from large systems to single hospital systems, and we had a nice breakdown regionally across the United States. Organizations were evaluated based on the level of personalized medicine integration across five key clinical areas including oncology, rare and undiagnosed disease which Nancy just talked about pharmacogenomics prenatal or neonatal and even healthy patient screening was a place where we're still developing and still needs a considerable work towards clinical adoption. But we evaluated PM adoption across these areas and developed through a weighted system, a score which then was used to assign a level of integration from one to five. The next slide was shared by Peter Hewlett yesterday as well. And it shows the multifactorial set of criteria that we established to assess personalized medicine adoption, eight different independent dimensions, including different types of information that are collected and utilize how they're processed and utilized, whether it's data sharing and how the, how programs are structured for this conversation I'm going to focus on a couple of these independent dimensions. The first being the collection of genomic data, because these are the ones that are most relevant to a genomic learning health system. And secondly, our fourth independent dimension testing guidance and data accessibility, and lastly data sharing. Before I get into that let me show you the overall integration. As I mentioned, most of the systems you see a little bell curve here, most of the systems scored a two or three and that's great news genomics are being integrated throughout the US health community hospitals as well as academic health centers at some level, but you can also see that there are far fewer that far fewer health care delivery institutions that are scoring at the level of four or five and again I think this is the level that we're going to need to get our health systems to in order to be considered and really a part of a learning health system that shares data externally, and we'll talk a little bit more about that. Just a breakdown, because I think this will be of interest to the group of of how these scores played out. You can see this through our academic verse community verse community teaching where the scores broke down where the levels broke down health system independent hospitals integrated delivery networks in the middle, and then the suburban rural and urban breakdown, based on on region. Most of this is what you would expect what a couple of interesting that things that came out. As you can see, all the way to the left that that academic health centers really had very very few that were scoring at the low levels, especially level one really academic health centers are at least personalized medicine and genomics at a higher level than some other community health care systems, but you see the level of four or fives are pretty consistent whether it's a community system or an academic health care system across our sample. You also see just mentioned that the rural verse suburban and urban breakdown shows that most of the rural systems scored at a level two or three. This is also a feature of the fact that we're only able to get six rural systems in our out of our 153 that were sampled. So that means four of those scored a three and two of them scored a two so a little bit sparse data for looking at the rural community hospitals. Now, to talk about the genomic integration. Real quick, I just wanted to show that health systems were assessed both on the breadth and consistency of their genetic testing, whether or not they were implementing multi gene testing or exome sequencing or genome sequencing, and you could see that that at least some physicians were ordering testing or most physicians were in testing multi gene testing across the board and the breakdown for this across different clinical areas shows that while oncology is receiving the most genomic testing of the five disciplines we looked at it's across the board genetic testing is genetic testing is being implemented. It's been more broad targeting and whole exome and whole genome sequencing in oncology and rare undiagnosed disease than in areas like pharmacogenomics and prenatal testing. But this shows a nice breakdown of where we are as far as genomic testing is being implemented. We got into testing guidance and data accessibility and we measured this through whether or not systems were manually ordering testing, whether they were pathways or clinical protocols in place through the electronic health record that then prompted a doctor input or develop results for biomarker testing and genomic testing within place, or whether there were results automatically integrated into the electronic health record, and you can see that across the board, that about a quarter of the, and this is the top thing about a quarter of the systems that we looked at had pathways or protocols in place that had results automatically integrated into electronic health records. This was slightly less in pharmacogenomics, but for most disciplines we're seeing only about 20 to 25% of institutions that are are at that level of integration. However, we are seeing a promising look at having electronic health records, prompting positions to do testing in certain areas. Lastly, I wanted to show the data on data sharing that we came across and we measured this by whether data was being shared internally within a department internally within an organization or externally. And I think it's important to note that the external sharing is really where we're going to need to be to develop a learning health system as Peter Hewitt defined a learning health system yesterday. It's a system in which internal data and experience are systematically integrated with external evidence, and this knowledge is put to use for to practice to get patients higher quality, safer care. This external sharing is going to be necessary and it's one of our key challenges which we can discuss shown here in all of the disciplines where we're seeing data sharing occur. We're doing a decent job, especially data that shared within institutions and within departments, but we could do a much better job at external data sharing. External data sharing is only occurring in between 10 to 25% of the institutions that we were able to look at, which I think is a representative look across the US healthcare system with slightly more about 25% in oncology. So the conclusions that, and there's a lot more to this study, but the conclusions for this discussion that I can get to is, is just that based on the survey of representative samples of healthcare providers reveal the system wide but incomplete push to implement personalized medicine in clinical practice. This, I believe underlines both the momentum that the field has as well as the limitations associated with the utilization of new technologies. US health systems are making great progress but we must build on this momentum in order to raise all healthcare delivery institutions to the highest levels that will be needed for a learning health system, effective learning health system. Now, also, a key conclusion is that by interrogating each of these systems unique adoption fingerprint, we can highlight specific challenges and discuss potential solutions that may be insightful for the assessment of progress and for the development of a genomics learning health system. And this includes but isn't limited to the collection of genomic data, the test result database automation and data sharing. Oh, thank you with that I'll leave you and take questions and get into the panel discussion with Howard and Nancy. Thank you so much Darryl that was terrific before I turn it over to crystal to moderate the general discussion were there any quick clarifying questions for Darryl. I don't see any in the chat okay turn it over to you crystal and if we could have all the videos for our speakers turned on. Alright, thank you so much if you are able to could you please drop your questions in the Q amp a that way we don't lose the questions in the chat. There's some amazing discussion going on, but I just want to make sure we don't miss anything but I also acknowledge that if you're a panelist or moderate or you may not be able to use the q amp a function in the same way. I want to start off with actually a quick. I think this is a clarifying question or just a quick question from new and park directly to Howard. And just it just reads Howard does Intermountain work with Mountain State regional genetics network. So, on an individual level there are some of our folks amongst our counselors or positions that do engage with them, but institutionally we, we, we haven't nothing against them just haven't haven't, haven't just haven't had a need. Alright, well thank you so much. I'm kind of going in order of when where I saw this in the chat so I do apologize if it seems like I am prioritizing one speaker over another but I also do want to highlight Adam burgers question. Has there been exploration of the intersection of genetic services and CMS alternative payment models as an incentive for adoption. APMs are meant to incentivize payment for provision of high quality cost efficient care. Darryl I know that you entered a response in the chat. Would you mind reiterating what you stated and then I'll also open it up to both Howard and Nancy. And thank you Adam for the question. It is something that the personalized medicine coalition in the whole community all of us have been thinking about how can we leverage this value based response and the new evidence that we're developing that shows the greater value of personalized care and genomics to CMS and to others that are key decision makers so that we can have better access to these technologies. And I think the key challenge is just in that and the evidence and Howard talked about this at what Intermountain has established but a lot of these data points that show that show an economic benefit required a long term outcome measures. So you're looking at a lifetime, because the costs for a treatment, a targeted treatments may be more expensive than small molecule treatments that have been traditional care. And because the cost of diagnostics is added to that upfront the cost may be higher but we are showing that this is leading to reduced costs downstream. But we need to have that long term data, we need to take a look at endpoints that are five years 10 years of lifetime down the road, and that's what's been difficult to come at most of this evidence that we have is based on models and these are good models good cost effectiveness and clinical utility models that show improved value that will be useful for CMS and others. But that that data is still sparse because we were accumulating it and we're doing these models. This includes an oncology rare and undiagnosed disease and in pharmacogenomics, but we're seeing that evidence developed now, and I think more and more we're starting to move the needle and push those decision makers into recognition of that value. Thank you so much for those comments Darryl. I asked about acknowledge that the timing of this question coincided with actually Howard's talk so Howard do you have any further comments or would you like to elaborate on that question. I think Howard's hit it pretty well I think, you know, CMS has been, it's been a bit mixed engaging with CMS. On the one hand, they have to be involved with innovation, because they're going to ultimately pay for it. But they also don't want innovation because they're going to ultimately pay for it so it's, it's a group that we've tried to engage with a little bit more clearly. And there's some champions there that goes better than others but overall we just, we've ended up just deciding you know most of us we're going to do ourselves. And then we'll work with CMS as we're able but we can't rely on them, because there's not a consistent path to go forward. Thank you. And Nancy I also want you to have an opportunity to respond as well if you'd like. Sure, thank you. I mean wasn't there there was something that was called the Medicare coverage of innovative technology 2021 proposal and I think I think it was repealed. I don't know how we get them to consider it again but that might be helpful. That was one thought and the other thought I have is that you know, when they do the health economic outcome studies often they're looking at a 12 to 17 month outcome, not over years, because that's how often people switch insurances. I was, it's, you know, your car be twisted in between and if we can try and do some studies that look at that short term financial gain by getting a diagnosis sooner saving testing, decreasing hospitalizations and emergency room visits having people not continue to look. I think that would go a long way as well. Thank you so much. This question comes from Aaron Howard, does Intermountain share aggregate variant pathogenesis dissertations with ClinVar. It does and the US we're getting more of our data back. We're working through the process of what, what can we share so we do a whole research whole genome in a CLIA environment but it's not CLIA, and then follow up with CLIA sequencing so I think Aaron and I are hopefully going to be talking sometime in the future. That'll be one of the topics, you know, is it worth putting our research grade data in there, realizing that not all of it will validate which at the moment we're not putting that in so kind of I guess is the answer to the question. Okay, thank you so much. Um, Jeff actually had a couple questions. The first question was directed more toward Nancy and I see that she answered that in the chat already so I'm going to go directly to just second question. As we move toward population screening. How do we address the issues that underinsured patients, or those who have no insurance will need medical services. So I'm going to start with the mammography colonoscopy after finding a pathogenic variant. And that one is to the group. Any thoughts on how do we ensure, and Jeff, please, if you'd like to chime in and perhaps clarify or add a little bit more details to the question. The next question of this is, as we are moving toward population screening like this to me seems more of an equity question and ensuring that patients who are already underinsured or underserved actually are able to access these services. And then they determine that they have a pathogenic variant of the year. No, that was a spot on and I was thinking what we've had discussions in North Carolina about doing large scale precision health screening for tier one disorders and this issue always comes up about going into a community that you know that may not have the right support from from payers and I was curious as to how it is you've launched your large scale program. Have you run into this issue and how have you dealt with it. Yes, so thanks Jeff so on the population health side, because we have a lot of value based care, including Medicare Advantage saw Robert had questioned there on that. A lot of the precision health approaches. We pretty quickly in a, in a value based system, you get to own the value, you're not just transferring the value on someone else. And so what we found is, is in the context of especially of rural populations, which do overlap with, for example, some of them have a lot of, in our area a lot of Native American groups, some Hispanic, a lot of Polynesian and of course you grow Utah. There, there we get some extra value in that we're able to find things earlier. We're able to benefit that because we're, we're holding the risk on that patients care. And so by reducing the risk, we're able to intervene in a better way. There's the issues with rural folks and with the those without that are economic disadvantage has been they have not been treated well in the past. And so there's a lot of work needed to make sure that they see that there's value for them. And as a, as a health system at least one that's not completely addicted to fee for service. It is not hard to make the case where population screening population health strategy can can be economically favorable. I'm saying that, that more health system should take, should take risk on these populations and then they'll be able to provide more care for people that are in rural situations are underserved. So we're, we are, we have a lot of our population that it that we hold the risk. And within that group. So I haven't done the actuarial work to say, we should go out and look for one group or another. But within the groups where we hold the risk. It pays off to that now there's a lot of overlap, because of the way our health system is designed with rural, rural populations, and with some of the historically are currently disadvantaged groups. So they're neglected groups. So, so there is a high overlap there. But I think, you know, there are people much, much smarter than me that can do analysis to say what the business case is for seeking out one population versus another. I think it's great. And I'll just say that a lot of these patient screening programs are really pilot programs at this point I see Mark Williams with his hand raised and I'm sure he'll talk about the micro project at geisinger. Again, to scale this up, which has been a big part of this conversation, we're going to need the data to show that there's an economic benefit. And I think Howard just laid it out perfectly if you're talking about at risk populations where you're screening, then I think that value proposition will be clear but when we're talking about full population screening, we need to show that there there is a cost effectiveness to it. And I'm not sure we're there yet. Thank you I do want to give mark the opportunity to respond and thank you for having your hand up. And thanks, Darryl for the lead in there. Yeah, I wanted to just talk a little bit about this from the my code geisinger perspective. Of course, we're set up very much the way Intermountain is, in terms of our integration. What that means is that for people that participated in my code which is open to any individual, any patient that receives care geisinger. About 40% of those are covered by our health plan and we pre negotiated with our health plan that any recommended medical care that would come from the my code recommendations when we identify an individual the variant would be covered so we, at least for the 10% where we had the responsibility. We said that we would make sure that our health plan covered it now of course there are other issues like co pays and deductibles that also can sometimes lead to people's inability to get the needed care. That's something that we can't address as much. We also for all the initial return of results in that that was all covered by the research project so there's no auto pocket costs to any participant in my code for that initial return of results, and for the transition into care. In terms of the really challenging issue of that is, you know not limited to genomics but that there's a lot of people that can't get the care they need, because they're uninsured or under insured. We have at least made available all of the programs that are available within our institution to help people that have difficulties to obtain needed care because of financial challenges. We don't have access to those services as it as any other patient, but there are IRS rules that don't allow differential application of those types of programs. So we have to follow the those eligibility rule, but we do make sure that we help navigate them to those services, so that they can. We can do as much as we can to eliminate that type of disparity. Thank you so much. And I do want to address Josh Peters since question in the Q&A. Just reading this question for Darryl. Surprising to see the relatively robust implementation across different types of hospitals seems to be a big improvement. Any concern about response bias, or our rural hospitals adopting quickly. Can you give us an idea of what a level five health system looks like. Thanks for the question Jeff. So, as I mentioned, there was 153 hospital systems that were of all different types that were part of the survey that led to that framework development. Really, I think with that kind of end we're minimizing bias, however, I also mentioned that with the rural systems in particular, they were self identified as rural system and we only had of the 153 systems six that self identified as a rural community hospital system. So there is a definite chance for bias from that particular cohort. And as I mentioned, most of them measured a three. And I think that would be considered terrific if the bulk of all of our rural hospital systems across America were measuring at a level three on a scale of one to five and genomic implementation, not sure that that's true. I think the ones that we had in our system were that, but it was very difficult to reach them. An example of a level five system would really be a system that has programs and personalized medicine programs in place or genomics medicine programs in place with all of the genetic counseling with all of the data integration that we've been talking about throughout the conference for the last two days in place and actionable, they're sharing data internally and externally to help drive clinical practices and pathways forward that in that increase our knowledge of genomics in clinical care and utilize it for better patient care safety and efficacy. Some examples would be like the Mayo Clinic they've been doing this for a long, long time. We need to move our baseline closer and closer to those levels of four and five. Thank you. Thank you so much. And just also wanting to address Rob Raleigh's question in the chat at Howard. What has your been been your success with Medicare Advantage plans versus direct Medicare coverage. Yeah, so we kind of hit on that a little bit with some of the other discussion. Medicare Advantage is on its surface seems like a place where personalized medicine would not be conducted because of the excess cost that brings in testing. We've been able to work through both on the oncology side as well as on the geriatric and chronic disease management sites where where there could be benefit and what we found is in the oncology side by able to select medicines and by identifying which can be oral versus IV medications. There's different implications on which budgets those come out of and so suddenly when they realized that there was some positive budget implications, they became big fans of tumor testing. Same thing on the pharmacogenomic side by by demonstrating where the benefits are by now there's a little bit more data out there. We have some of our internal data. This has allowed the discussion to be less theoretical and more practical. And I think that's kind of held the field back a bit is we have this theoretical idea of what it could mean as as guys near and as other groups come out with real data. So if you have to to guess what's going to happen. We're now seeing that the decisions are are being made in much more practical manner and more often than not leads to adoptions of the technology adoption adoption of the technology. Thank you and Mark also can speak to Medicare Advantage. Yeah, we have the advantage of our health insurance plan actually having a Medicare Advantage offering. And so our discussions with our health plan also included the Medicare Advantage plans, which was really important since the median age of the folks that are in our my code. The other participating my code isn't in the 60s so it's it's definitely a population where over half are of Medicaid eligibility, and we have a significant number of those that are enrolled in Medicare Advantage. So, the other advantage of Medicare Advantage is it allows for coverage of preventive services which are generally excluded from traditional Medicare. You have to do within the constraints of how much available premium dollar you have for things would go beyond Medicare coverage and of course, if Medicare says you explicitly can't cover something. Then you can. But it does give more flexibility and so we have been able to extend coverage for medical services that derive from a variant identified in my code. That have our Medicare Advantage plan, but of course that doesn't have any impact on the other Medicare Advantage offerings that are that might provide coverage for people in our area. But that also allows us to collect data to say, is this really a good use of the preventive funds that are available to Medicare Advantage. Okay, well, thank you so much. I'm going to get to a Lana she's, she they says, I would like the panel to discuss not just the cost of effectiveness, but the effectiveness of engaging individuals in meaningful care after receiving a result. Do any of our speakers have a comment. So, I'll, I'll jump in first I guess that. Very important for really practically, if we don't achieve what that question was going after, then kind of who cares about the rest of it. It's just something for a symposium not for a for really moving the needle. And so we've had to do some things differently. And we have the classical genetic counseling way forward tell and otherwise, but we've set up some some separate all telemedicine based where they're one of them is a longitudinal care model where we're looking at a basically a medical home for those who have hygienic risk. And it's hard to find those out there, you know, typically, the clinics are very episodic or they're a triage based. Nephi Walton also set up a more of a triage based clinic so we get whole genome data, one can identify what needs to be chased, and then refer to specialists who are ready for that diagnosis. Not every neurologist knows what to do with someone who has a high genetic risk of early onset Parkinson's, but is 37 years of age. You know, so that that sort of thing, getting to the right specialist is there. And so, at least in our hands we've had to set up new clinics that really were more fit for purpose, rather than using the traditional models. And then lastly, supporting the primary care folks so when they do want to take it on, making sure that we have their back in real time and tell us really changed a lot of that. And then we get a specialist into a room for a few minutes, as needed, if it's, you know, if it's arranged and so I think it's just a case of responding to that need, as opposed to, you know, how do we fit people into the way we've been doing it for a while. It just doesn't work. We're going to have to, everybody that's talked, this on the session, has been talking about new things that they've had to do over the last year even. And that's going to be the story really for the next five years. Great, thank you. Do either Daryl or Nancy have any other elaborations on that point. Thanks, Crystal. And thanks for the question. Just to, to follow up some of what Howard just mentioned. A lot of these studies that we're looking at to show value are, we're talking about them in a way of focusing on the cost effectiveness but they are designed as clinical and cost effectiveness studies, and that clinical effectiveness that clinical utility piece is the primary pieces, as Howard mentioned, without that, even the cost effectiveness doesn't matter. The reason why we're talking about cost effectiveness so much is because we're, we're talking about access and we're trying to show evidence that patients should have access to this care. We're driving those messages towards those stakeholders that control access to these technologies, payers and providers. That's what the cost effectiveness really goes to. But the most important stakeholder in this is the patient and the clinical effectiveness and that clinical utility piece is what really matters to them. It's also important obviously to providers and payers and industry and all stakeholders, and we're really trying to show that. Now, the projects and programs to show clinical and cost effectiveness that I've been a part of, and that I've seen coming through the personalized medicine and genomics community have shown great clinical effectiveness. Again, this has been, again, it's been shown very clearly in rare and undiagnosed disease, somewhat in pharmacogenomics, and then again it's been clear in oncology. One of the things that we've also seen is we're not realizing the full clinical utility or the full clinical effectiveness of these technologies because of policy problems, including access problems, and because of practice based problems. It's not being used. There's not enough education. Some of the things that we've been talking today about what we can improve through a learning health system that will help optimize this strategy will help improve that clinical utility. If we look at all of these issues and address them, then yes, we will have a clear picture of great clinical effectiveness that, you know, will be undeniable for patients going forward. It's really helpful. Thank you. I mean, I have to admit earlier I was typing an answer to something else in this part of what you said, Howard. Sorry. The one thing that I would add is we, we often measure a net promoter score NPS, which is sort of like a customer satisfaction score. So an NPS score in Disney World is, you know, over 80. And in one of the programs that I was part of previously, our special needs initiative, we started out with an NPS in the negative range, like negative 100, and got up to 70. So we considered that a huge win and really seeking to understand what it was that our members and staff, such patients, which took me a long time to call people members instead of patients actually need and want. So I think including that and our studies might go a long way. Thank you. Thank you all. Howard, the next question is directed toward you, but I think all three of our speakers can possibly comment as well on this considering the relatedness to the question we were just discussing. I want to comment on the possibility of sharing data related to cost savings using personalized medicine initiatives in your health system with payers to help shape future policy. Yeah, so there's some of that data, we are trying to get out into the public domain, either via white papers or via publications but there's a lag that occurs there. There's sort of forum that we found so far where we can share that data in a safe enough way to allow people to see it but not give away all the trade secrets and and the elements that some of our leadership doesn't want to give away. And so that that's been the, you know, the difficulty is trying to like how do we make sure that the field benefits from from that. But not, but not compromise the institution's ability to be competitive going forward. And so that, you know, some of the savings we're talking about is the secret sauce for having a profitable margin in negotiations, you know, so it's, you know, that's kind of the hidden thing with precision medicine is a lot of that is secret sauce. And that's why we can do Medicare Advantage and not lose money, because we're identifying risk and mitigating you know so I think that that's, we do need to find a forum where we can better share these things because you know academic publications are fine we all do a lot of that. But you know, the lag time is too great and, and, you know, frankly, there are very few journals want the kinds of things we're talking about today. It's fascinating to me that you say that the leadership doesn't want to share because when I listened to you and to mark these two places that have been so successful, right, in showing a shared cost savings and being able to move the field forward are places where your health insurance is part of your health system. And that is really powerful. I think the partnership between a health system and a health insurer to share in the savings is part of the reality that allows this to move forward. It helps the patients and it helps the members it's it's the, the push and the pull of the difficulty of the US health system because it's a business here in this country. And, and that's tough because none of us went into medicine to be part of business we went into medicine as physicians and genetic counselors and PhDs to help people and be clinically active so how do we lean into that, find the savings find a way to share in it so that we can best move the field forward and care for more people and get them supported. Seeing some head nods of Darrell, I definitely hear your thoughts. Thanks crystal and thank you Nancy and Howard. I'm actually part of the mission of the personalized medicine coalition so it's my job to do a better job of, of spreading this word of accumulating this evidence and bring it forward to the appropriate stakeholders and Howard McLeod is on board of directors for and so we work together a lot and I've worked with the Intermountain team with Howard and Lincoln to move forward a lot of the data they've done, but we also build this data through projects, we really community wide projects and part of our strategy of doing cost effectiveness and clinical effectiveness studies is to involve a payer advisory group. We have a healthcare working group, which consists of providers, we have patient advisory groups as well. But by including a payer advisory group when we develop these projects to show this evidence to use real world practice based evidence to show the value and a personalized medicine and genomics approach. We can bring the plan to our payer advisory committee up front and make sure that it's going to be meaningful. If we do something that payers don't feel is credible or don't feel is providing the evidence they need for their decision making then, then we don't want to do it we want to make sure that we involve them. We want to be in those discussions so that we can, we can, we can move the needle with our impact of these studies. Great great comments. I want to focus on a question that's directed to you Darryl. Any plans to repeat the survey you presented and track changes over time and reasons for increasing implementation. That question from Pat so thank you Pat for the question. So there's two things we can do and we hope to one is to update the survey so that we can show progress over time. And that's a plan we're working and I want to thank Gary Gustafson team and health advances that's helping us do that that work. And I think that's something that we hope to do within the next couple years. That'll give us a good sense of how things are progressing how things looked in 2021 and how they'll look at actually that data was from 2019. So that will look in 2023 or 2022. To show our progress and to show different elements because there are these individual fingerprints that we can look at to show what progress has been made in those different elements of the personalized medicine score. So what we can do is fine tune this this as a tool to look more at a genomic learning health system and make sure that this framework for our adoption scores is is built with a genomic learning health system in mind so we can ask more questions about genetic counseling about health equity about data sharing and try to drill down a little more to inform this particular effort for the that NHGRI is doing for the development of a genomics learning health system. So I think both are things I would like to discuss with all of you. So that's my plan on doing the update. I'd also like to talk about how to make this absolutely relevant and impactful for a genomic learning health system. So hopefully have those conversations to up and coming. Well that sounded like an opening if Nancy and Howard had any thoughts before we go to our last question. I'm hopeful that Darryl does that. And doesn't ask me to do any of the work. Just advise Howard I want you to always be involved. I think it's great. Thanks Nancy. All right well great. I think we're coming on our last question and this is coming directly from Aaron, but it seems to have elicited quite a bit of interesting discussion in the chat so I have asked Aaron to re ask your question to the group and perhaps, in case there are any other questions like please post them but this one. Thanks crystal you've been doing an amazing job. Yeah so my question in the chat was directed to Nancy, because Nancy made an interesting point that the virtuous genomic learning health care system cycle could potentially have a third circle, which would talk about more direct interactions or collaborations with payers so I asked Nancy for a little bit of time on that Mark weighed in, but would love to hear, maybe Nancy expand and then mark to follow up if you have any. Yeah, I was going to say, I can start and I'm sure Mark will have more to add. He's insightful about this and he always does. So, I guess, a big mouth that's the difference, Nancy but I appreciate your, you're putting a positive spin on it. I always have people always tell me I have a big mouth mark it's okay. So, a couple things one is, you know, there's always the opportunity and, you know, and likely a cost involved to be transparent to partner with optimum and the optimum data sets which are very large. That's one possibility. The other possibility is to do as Howard and Mark have done, or as Darrell is suggesting which is to engage a more local health system partner so if you're a health system that is located in one particular part of the country, there's often a local network or or a Medicaid provider or a private provider that may be interested. I also think there are opportunities to partner with particular companies so if you live in a state where there's a very large company that has their own health insurance. That would be a place to partner. In fact, I had one person from a very large testing company call me and say, how do we get whole genome sequencing covered because they wanted it covered and I said, well, does your company cover it. I mean, people people people forget right like your own company all these private companies are they including the genetic testing and what is their approach to that. And then, you know, I'm also in our system certainly happy to connect people locally with who I know that may be willing to partner as well. The one other thing I might add is, I think from a broader research perspective NHGRI what should they be you know Terry's always saying to me, what should we be doing what kind of researcher will be doing so we can make sure we're getting the right answers. I think if some of the people that are leading in the field the people that are on this network can help us ask the right questions. That might be helpful. So one thing that bothers me is, who is an expert in what and how do you define it. I'm going to make some standards around that. You know, those kinds of questions and helping to address answers for those that are very that are basic and create some infrastructure we can all build on would be very helpful. So I will take the opportunity to just add a little bit I wanted to explain a bit more on some of the issues of alignment. As I mentioned in the chat. I think one of the reasons why some of the leaders in the personalized precision medicine area are integrated systems like Kaiser or an inner mountain or geysing or others is because we can have discussions where we have clinicians and we have the hospital and we can have the pay are all sitting around the same table. And if we take a condition like limb syndrome. And we say well we identify this individual has a variant in Lynch syndrome and if we enhance surveillance, then we may be able to prevent a cancer in this individual. And the payer says well okay, we might agree with that but that might be 10 or 15 years down the road and the Nancy's earlier point, we may not be ensuring that person at that time. And so it's going to incur a cost but we're not going to realize the benefit. But if you're in an integrated system, the clinician can say well wait a second. We're going to generate some revenue, because we're going to have more colonoscopies that we're going to be doing. And if those are done in an outpatient surgery center or in a hospital facility, the facility is going to generate some revenue there so in an integrated system you're going to say okay health plan you're going to take a little bit of a loss on this, but we're going to make it up on the procedure side. That's an indictment and, you know, a reality of the United States healthcare system and how it works. But if you're a health, if you're an insurer that's just sitting off on your own, you're only going to see the cost, you're not going to be able to recover any sort of benefits because you're not aligned with those other systems and that's where the problem really comes in. In terms of identifying how can we actually make an argument that aligns the strategies and values of these different organizations around this particular topic. I think it's possible but it's a much harder conversation to have. Thank you. Oh, I'm just going to ask if there are any other comments before I kick it to you. If there are none, feel free to hop in Aaron. Thank you. Yeah, wonderful. Crystal and I can't thank our speakers enough for your great presentations we have a terrific set of questions and discussion. So I'll just, I will summarize my the main points that I captured and then if any of our speakers or Crystal wants to add anything that would be terrific and then we can wrap up and take a break before next session so we heard from Howard some specific examples where precision medicine reduce costs and improve outcomes I think well we should continue to catalog and disseminate these advances. We talked about not just the cost savings but more importantly the value and the meaningful impact that this has on patients Nancy emphasize the challenges and opportunities to leverage the genomic learning health system to support people with rare disease. Highlighting the economic burden estimate to be at over a trillion dollars. We also heard that policy does not equal coverage coverage is determined by different insurance entities. And the answer reminded us about the hierarchy evidence to develop coverage policies. So the community we need to have clear outcomes and guidelines showing for example when to order particular tests this is critical in helping to leverage decisions. The virtuous genomic learning health care system cycle could include more direct interactions and collaborations with payers which is we just ended on that topic. And then we had some discussion on outcome measure so Darrell summarized the PMC study and that was really enlightening the adoption model covered eight main areas. We had a presentation collection of data testing guidance data sharing leadership etc. We were some discussion on sort of what standard outcome metrics do we have with them those categories and how the community can begin to use those more globally Nancy mentioned the net customer satisfaction score, but there's other metrics we should have in the community using and capturing that information in a standard way. And then lastly work to be done to able to follow people and their outcomes across insurance providers especially this is especially important through the lens of health equity. So I will pause there we have two minutes. Does anyone want to add anything to that summary. Only a note of gratitude and really appreciating part of this today and everybody helping us out. Great. Thank you. Thanks, Nancy. I'll just add my final comment that the data, the real world evidence and practice based data is critical for this entire value proposition which we, which we raised today we need the real data and those real outcome measures in order to have an impact. And I think we've done a great job Travis Austin and yesterday with Bob Dolan and Christopher shoot and Carol block bolt did a great job in talking about how we can leverage some data metrics. It's key to this, this endeavor for turning this into a learning health system. Thanks, Darryl. That's an important note to end on. So I think we are on time and we are now up for a 10 minute break, and we'll be back at 310 for our next session. Thank you. Thanks, Crystal.