 Thank you Carol. If you could have a seat at the table. We will have a discussion moderated by Shanita Hughes Howard of the Medical University of South Carolina with discussants Levi Garway at Harvard and Kelly Orman at Stanford University. So thank you. I'd like to first thank all the presenters for great presentations and perhaps we can start with some opening comments by our two discussants. So I thought those were all great talks highlighting a number of both challenges and also some very interesting suggestions for how to address them. So I think certainly the need and the challenges was very well articulated, the prevalence of disparities in health but particularly the potential for that to be a paradoxically worsened by genetic medicine. There are a few examples given. Several of them were in newborn screening, cancer of course being out ahead but there are a variety of areas though where there are these disparities and one can easily imagine ways in which things could get further confounded and worsened if we're not highly attentive. At the same time I thought this was quite an encouraging overall set of information because one can begin to envision a framework where that doesn't happen or at least where we use tools that have now matured in implementation science in comparative effectiveness research and in public health for example where as we go we can in principle implement approaches to learn and not just to learn where the problems are but to actually study how the advances that we're making might close those gaps in a hypothesis driven way. And what I heard anyway and I'm sure we'll have a conversation on this is that the themes over and over again that we hear around barriers some of them are one can imagine are quite addressable. So one of the themes that we heard is of course the issue of challenges around recruitment, challenges around understanding on the part of clinicians particularly clinicians that are taking care of patients in diverse and underserved areas even what genomic medicine is and how do you use the information. There are challenges around how do you leverage for example the electronic health records, the environments that clinicians work in all the time. How do you work within that to kind of feed information that could be used to make decisions and then of course how do you study whether those outcomes are or whether those decision making is influencing outcomes in a meaningful way. But the good news is that for each of those there are already activities in many cases in Ignite but certainly around the country where one can start to imagine if not best practices at least hypotheses of package interventions that might make a difference. And one of the points that I think Carol made towards the end of her talk of identifying areas of low hanging fruit and starting there and maybe because success can sort of breed success. So in all of us in our in fields that we're in can probably identify what some of those areas are. Certainly in Kansas which is where I'm in. It's very clear that and we heard some of this that the use of genetic testing whether it's for germline screening or for mutation or profiling for known standard of care interventions lags in diverse and underserved populations. So they're less likely to get tested and if tested they're less likely to actually be acted upon. They're less likely to have access to targeted therapies that we know work. So variance of uncertainty significance I mean it's great it's an important area and certainly there should be efforts I agree with the point that Rick made about efforts to link databases of variants in diverse populations so we don't misclassify and end up calling a variant something when it's really in a particular population it's really not. That's obviously very important but equally one can identify interventions that we know work and where there are clear disparities and and so there one could one can begin to do hypothesis driven research to say for example you know we hypothesize that you know partnerships between leading academic centers and community centers that can allow infrastructure implementations to disseminate this information or we'll build educational modules or build streams that can go into EHR and we're going to test whether or not that information you know having that information actually leads to differential utilization and therefore differential outcomes I mean one can begin to conceive of and and articulate what those kinds of studies would look like and that can be done even in the standard of care we know what the variants are we know what we would do with them and but we know that there are gaps and and and I think ignite certainly is well positioned because that's sort of where you guys are you're you're already at the point of focusing on variants where there's already a reasonable amount of utility data the question is how do you how do you broaden that and how do you determine whether when broadening it broadening it you've actually improved health outcome so that's maybe my opening self I was actually I was heartened because in listening to this you know the you know the scientists you know usually when you're reading a proposal you're you're when you're reading a proposal you're kind of you're you're looking for what's the experiment I would do what's the study I would do to test the to test whether or not this is a work and in hearing these discussions you could already your brain could already raise a head well these are very simple straightforward I mean not simple logistically but conceptually there are multiple opportunities where we know there are gaps where we could easily implement some of these approaches and the chances are there would be successes and success can breed success in terms of galvanizing broader interest and engagement so I'll stop there okay so I'm gonna talk a little bit from a perspective of having been back in the clinic for the past year and some of the challenges that I think I've experienced that are reflected in the things I've heard from this group and I want to start by saying first of all that I think ignite has as Carol said done a really nice job at trying to engage these diverse communities from the start and I'll invite you to target California next because we can cover the Latino and Asian populations that you're missing from your diversity map one of the things that I've seen a little bit is how important the access and the coverage are to truly being able to integrate genomic medicine and in our pediatric setting at the hospital we offer exomes to all of our pediatric patients where we think it might make a difference and that population is probably about 50% Latino and more than 50% on Medi-Cal and Medi-Cal is just not covering most of these exomes so our hospital will have to choose whether or not they're going to cover those and then make decisions at that point and then out of those that go for insurance authorizations only about 50% of those are being covered so I think that coverage is really clearly going to be one of our huge barriers and if we're going to see any sort of access for diverse populations that's going to be one thing that needs to be tackled and I was just really thrilled to hear Tony present on the payer engagement because I think that that's got to be a national engagement in order for it to go anywhere so that's the first thing I'll say. I think the second thing that's become really obvious is how hard it's going to be to actually get all of these different clinician groups to understand enough about genomics to triage the right people and to get them where they need to go and to make sure that whether through EHR or otherwise people can actually find the genomic results that they're looking for. Something as simple as the fact that most hospitals are scanned in PDFs that show up on the media tab that you can't see on care everywhere in Epic can really influence how much this actually matters and that seems like a really simple stupid thing but I can't tell you how many patients that's meant like I don't know what's going on with you so the fact that you had an exome it's sort of worthless medically so that's one thing that I wanted to say the other thing I want to talk about is something that Carol talked about the diversity of our genomics workforce so we've been wrestling with this forever as a genetic counselor our numbers really haven't changed very much we still hover around 90% Caucasian and 90% women and only about 10% with any ethnic diversity and in the most recent professional status survey only about 12% report fluently speaking some language other than English and that is a huge problem it's probably worse if you look at people in clinical genomics on the MD and PhD front and I think you're exactly right that we need to engage a diverse workforce and a linguistically diverse workforce we try really hard in California to train at least a couple people every year to go out and be fluent in Spanish or Mandarin or other languages and I feel like it's complicated enough to explain these issues to patients in English to try to do it in a language that isn't theirs makes it even harder and that's going to impact our uptake we want to make this actually happen in diverse populations we really need to work on our linguistic things and we need to have that health literacy taken into account and the numeracy and I think we also actually need to not reinvent the wheel there because there's been a lot of work done in that area I was just thinking that Isaac Lipkus did a grant probably almost 10 years ago where he actually worked on risk communication in diverse populations around genetics so we need to start with what's there and then kind of see what we need to do with the actual resources that can be handed out and then the final thing that I want to say is I feel like there's a lot more space for psychosocial work around how diverse communities approach things like healthy genomes and the predictive ability that we may have here whether or not people are going to think about family cascade testing in similar or different ways and how well accepted that will be because as Muen was saying you get people out there passing the information along then you get papers that say maybe we shouldn't be doing this testing in the first place so we need to see how that's going to work on a diverse level as well so I'll stop there Thank you I'd now like to open it up to the audience for your questions, comments, reactions Yes Mark Williams Geisinger so this is directed to Muen and he'll not be surprised what he's going to hear I think the public health perspective is critically important but I think we also need to distinguish between public health and the public's health because in many ways I think programs like Lynch syndrome screening and HBOC surveillance and familial hypercholesterolemia while clearly public health problems are not necessarily well served in the traditional public health infrastructure in the sense that the resources that are needed like mammograms and colonoscopies and that are not within the control of the public health sector and the other problem is that the public health sector the pie is not growing and certainly in relative terms it's shrinking and so if you really try and force this into traditional public health delivery you really have to ask the question is it going to be something in food safety water safety those types of things and so to me the critical question is how do we move this use public health to identify what the problems are but then to use a delivery system model to really try and implement that and how can we work together that to me seems to be a really interesting research agenda and I think that's one of the comments on that Thank you Mark as usual somebody said that the definition of madness is to try the same thing over and over expect a different result so I've been trying it for 20 years in the public health world but there is some movement I think you're absolutely right there is the public health agencies and states, federal, it's local but there is the public's health I think where we are right now is that the need for a partnership between all of these groups together because the delivery of genomics is going to be in the health care system what we can do from a public health agencies and states sort of build on a number of activities like what Michigan and a few other states are doing one is educating the public educating the providers working with the health care systems helping indicators for population health improvements at the state level like how do we know that we're doing well surveillance systems taking care of disparities taking care of the policy issues that allow sort of access to data, access to test results lab quality some states are better than others like New York and California ensuring the lab quality so there is a role for public health the small P, the public health agencies but at the end of the day health care, public health have to partner together in this implementation space especially if we have to take care of health disparities because at the end the beginning of implementation will occur with early adopters with academic institutions there will be big holes even in implementing tier one applications the data are clear as was said earlier black women are less likely to get counseling less likely to get BRCA testing less likely to act on the result of the testing less likely to have access to services so there is a role for the both public health and the public's health and all of this and I would encourage the ignite network and of course NHGRI to kind of develop a schema for implementation that allows for that interaction to occur now in extreme situations like newborn screening where you have a clear public health role because the health care system was not doing its job PKU kids were missed at birth and you needed a rapid ascertainment schema that's how it evolved my friend Jim Evans keeps saying we screen newborns don't we so could there be an adult generic screening program that could be population wide that includes maybe a subset of the 56 genes for which everybody gets access to the results of that maybe not in an opportunistic setting but more of a testing scenario that's something for the future something that work can explore thank you Howard I really appreciate the comments Leva made about getting people to do what we know we should be doing now I think there's a great opportunity there both in terms of the testing and also the access to some of these therapies the question I have for the panel is love their thoughts in particular Rick so around do we need to go back to the model or to reconsider the model around specific cohorts of patients in our trials in our NCI cooperative group trials we get very excited if we would get 25% actually we get 25% non-white patients on our trial of which there's a whole bunch of different people in that category because the NCI looks at the national averages and if we beat the average we get a gold star and the idea though that we now have a data set where we don't have definitive numbers to actually draw a conclusion about these populations for outcomes for toxicity frequencies certainly from molecular aspects and so Rick's been involved in some studies that were plus and minus in terms of cohorts and I'm sure I'm ripping a scab off for you but the idea that we have definitive data to actually advise a group is what's missing right now and so I'd love Petal's thoughts actually I think there's room for multiple sort of strategies I think we should go back to at least having some kind of cut off some threshold some quota right but at the same time I really like what the example with the April 1 example where there's here's something that you go to the community specific community with this particular test that actually could be impactful right so irrespective of numbers just say this is something that is potentially actionable for this population so I would try all types of strategies to be honest with you I wouldn't necessarily say one's better than the other or what but we have to do something I know that you know the sad part is that you know the train has left right the train has left the station and if we don't do something now to sort of increase those numbers increase that knowledge base for the diverse populations it's you know you're not going to be able to catch it can I just say one thing I I'm obviously passionate about engaging diverse people in research one of the things that Ignite gave us the opportunity for I'm sorry if I'm not supposed to say this Ebony is that we are not going to finish our recruitment on time and it's not because my projections are wrong it's because the group I was working with when we wrote the grant was like yes we can enroll you know 200 people every month starting month 3 and I said well actually we can't because we're going to have to take six or eight months to figure out who our population is do some formative research understand why people with diverse backgrounds would want to be tested how they want the results returned both the clinicians and the patients so we're kind of right on time from where we thought we were going to be but a little behind in this and so you know this idea of it might take a little more time and sometimes it might take a little more money to do some of these things and I worry because I know that when most of us get grant funding what you want to do is you want to recruit on time on budget and and some of the stuff like that I saw with the PMI I was worried was going to be unrealistic if you want to recruit tons of patients really quickly what are you going to do you're going to go to the friendliest clinicians the easiest to recruit patients and the rest of the diverse population is going to fall by the wayside so I think that really needs to be part of how we're doing things or when I don't think we're ever going to see enough diversity thank you yes ma'am so I had a question and comment primarily for a moving so my I'll do the comment first I guess so in my presentation on clinical outcomes I think you're going to hear that we're really thinking along the same context that you're suggesting which is as we implement we need to garner data we need to figure out how can we develop data that support long-term use of the implementation as we also show the implementation so so so we're definitely as a network thinking along those lines my question was on your statement in Michigan about African-American women under utilizing the genetic testing I think for BRCA is that is that sort of the normal access cost kind of barrier or is that provided free so there's some other sort of barriers there not so sure I have to consult WCAT who is the architect of those findings in Michigan and I think it's a little bit of both there may be some sensitivities around testing but I don't want to give you the wrong answer so stay tuned I think Eric a comment first to mark is I think it's not appropriate to define public health is what county public health departments do that's a very small part of public health this meeting is public health I think we need to look at modern public health much much broader and not to find it by traditional agencies but my question to the panel is about diversity I think it's a little dangerous and I think we pat ourselves a little too much on the back to define diversity only by skin color sometime by language one of the things I'm afraid of as a citizen is the growing economic diversity in this country and I'm a little bit worried by focusing on a narrow definition of diversity we're going to leave out the lower incomes as technology decision health grows and what can we do as a group to make sure that these initiatives are not the initiatives for the wealthy I think that's an excellent point and I would hope that the entire panel sort of agrees with your statement but you know when you think about the gaps though there are gaps that are biological and so being poor doesn't mean you have a particular set of genes so what you would pick up in that sort of population would be more of the social cultural behavioral influences on health which is utterly important but there's also gaps biologically that really I think are astounding in a sense if you really think about the amount of money that we're putting into this implementation I agree with that but your point is very well taken and many of the suggestions and the approaches that have been discussed actually if they were applied would encompass exactly what you're saying the issue of obviously there's the issue of knowledge of what the information means and what have you and that actually if one says okay this is what these variants are this is what they mean and obviously if you think about the development of the population that's part of the education if that's being matched with additional implementations that overcome barriers to access and low populations I think you're right one can in principle put together a set of policies that in essence will address a lot of what you're saying so the economically underserved if that's the target you're going to get a lot of the groups and so obviously there are specific considerations about variants in different backgrounds that ought to be part of that but the big picture you're right we ought to be able to have an umbrella that covers all kinds of diversity and I haven't heard anything that would preclude that from this conversation one of the things I was thinking about earlier is how much of our premise here is on showing up for preventative health care and low income population and you're getting most of your health care kind of in acute settings who's going to be the person that raises these issues and translates it and then if you end up in a population where you get a variant of uncertain significance because we don't have enough data about your ancestry who's going to follow up down the line and bring these things up too so I think those are big picture things that are really enormous health care systems Katrina there's a lot of practice in the Pacific Northwest and it makes me uncomfortable to get through a whole diversity session and not hear the word Asian because those are my patients and they are very underrepresented in genetic research and you get no brownie points with NIH for collecting Asian people there's not a box for that that you're checking to meet a certain target and goal and I'm just wondering it's a pretty big population in the U.S. in general it's a very large population and we have the same problems in that population of not understanding what these variants do and not understanding what the best care is for these patients how do we get their voices at the table too because it's not you know it's not in the definition of underrepresented by NIH yes I'm sorry I can't see your name I got glasses on I can't see and for in terms of reaching to I would say many diversities have you guys thought about first of all looking at places and I can only go off and I have family members in Arkansas right and the low income but a lot of their health issues is stuff that they don't know like maybe the doctors they go to aren't explaining things to them they're not explaining you know preventive measures or anything like that properly to where they understand that the other issue might be you know look at your public schools I mean some of those kids could probably communicate or help communicate this issue to the populations that you're trying to reach and then for every poor person there's also that family member that is you know middle class wealthy I mean it's all getting that information from different angles to the right people you can't just look at urban I mean go across rural America those are my two cents thank you yes sir thank you this has been extremely helpful and very interesting discussion I have a few questions so I'll address it but then the individuals on the panel to Dr. Currie I wonder if there's materials that you have or CDC and other groups have to both train populations consumers as well as providers for this collaborative that you are thinking about there are existing tools on our website and the cancer division website the NoBRCA tools for providers and the public R1 we have an implementation toolkit department for a few tier 1 applications they are not very well used because of the lack of funding there are tools that are being developed as we speak as part of this action collaborative so within the next year or two there could be many more and we'd love to partner with you on some of that the other question I had was for you one of the big great limiting steps for us as we're skilling this genetic counselors we cannot especially the kind of genetic counselors we now have are primarily trained to counsel parents about monogenetic or rare diseases and not really for the kinds of things that we're talking about here and while we're trying to scale up our program the curriculum has to be changed and that's a big challenge that we need help from a group like this because of the types of counseling that you'll be doing I wonder if you have any current plans or this group has any plans to do that. So this is something that the genetic counseling profession is well aware of and they've actively engaged in a workforce assessment which I have heard some preliminary data on and not surprisingly there's a huge gap between the I think there are probably three positions for every genetic counseling graduate at this point as a training program director I can tell you that all of us are actively engaged in trying to train our students to deal with the complexities of the adult genetics conditions and how we do it and our students are not just taking positions in pediatric genetics anymore that's actually the minority so it does take time I think we're going to have to all work hard to find ways to bridge that gap and continue the expansion and kind of not get steamrolled and make sure we're doing it appropriately but we are at the problem. I think that part of it is defining when and why we need genetic counselors for everything so in our model for April 1 which is you know it's a risk for kidney failure like high blood pressures risk for kidney failure and other things risk for kidney failure or pharmacogenomics where it's like your creatinine is too high don't take this you have this genetic variant don't take this the way that our model is is we have our genetic counselors be the super counselors they're the ones teaching other people how not to screw it up you know in the old days with HIV it used to be that I as a primary care doc was not allowed to test my patient for HIV without a counselor first counseling and the counselor had to return results and then we realized well we can't do that anymore there's too much going on so I think it's going to take a definition of the counselors elevating to a higher level and saying what can we have other people do and what do we have to do we have the kind of the equivalent of a community to health worker returning results and of the 1800 people we've tested so far and we've offered all of them a free meeting with a genetic counselor none have availed themselves of it and I actually think that's really appropriate I think that of all of the conditions we've talked about when we've interviewed non-genetics healthcare providers as part of our ClinGen study pharmacogenomics for example is one of the things where people really agree we don't need a genetic counselor but the hard part is that even when you get genetic counselors and geneticists sold on the idea that we can sort of teach people how to do this the other providers are very frequently pushing back and don't feel ready and I think that's going to be a big challenge too yeah maybe a better definition of genetic educators rather than counselors finally in terms of diversity one of the big problems is the rural community which is while it's complex it's very different kind of challenge it's not so much language as opposed to access but the other problems particularly for states like Indiana which are very high farming states approximately 10% maybe even 15% of the Indiana population is totally migrant population we have farm workers who come probably from July through October and they're all from Mexico and they just stay for four months and then they go back so we have this incredibly sort of transient population that is Spanish speaking that's not a simple challenge to address so there are some complexities even to minorities that we have so we'll take our last question and that from thank you clinicians so this is a question for the group to think about practice group health systems they usually follow national guidelines and at least in pharmacogenomics I want us to think about how this group can actually be integrated into chest guidelines working groups that provide a balanced view of the evidence in these national guidelines because this is what most hospitals follow the second thing I want us to think about is health disparities picking up on the thread of not just racial diversity but socioeconomic diversity and perhaps the thing we can think about is if you're studying Parkinson's it's okay to have just 1% African Americans but if you're studying failure and that patient group is 40% African American then you should really be you know the threshold should the bar should be set you should be recruiting 40% and the last thing is I presented at a health disparities conference recently and I got the whole a sense that all the genomic findings in press have the way it's presented in the lay literature has a lot of negative connotation so we as a group need to do better in terms of communicating for example the EPOL1 story is this is impactful to you and to take what Carol said is be at the table so you're not on the menu so those are the comments. Those are great comments one brief comment that your second point seemed to might have been a bit of a completion of so if you're doing clinical research you're doing a clinical trial on a disease obviously if it's a disease that predominantly affects diverse populations that obviously it's important to understand how those are being represented and hopefully be a part of the barriers that may entail to doing that research but there's also the issue which I think a lot of this is about where there's already a certain level of clinical utility data out there and then it's about how do you implement that particularly genomic based evidence in broad populations and those have a different set of challenges potentially than the clinical research so there's some overlap that's important in these health disparities population if you don't engage them in those clinical research. Thank you so I'd like to thank all of you for a very robust discussion and I'd just like to bring this session to a close by summarizing what I heard as the key points. First is that diversity is important not just from the perspective of race and ethnicity but also from the perspective of gender and so as we move forward with thinking about diversity in ways to not have genomics and precision medicine enhance disparities it's really critical that we think about diversity from a much broader perspective. The second thing related to that is it's critical for us to develop and implement strategies for having greater inclusion of diverse populations in our cohorts and in our research programs. I think I would just want to draw your attention to the work that's being done at Carroll Center and I don't remember the catchy acronym for it. Thank you because I think it really does really exemplify my opinion some of the things that and approaches that need to be replicated to ensure that we include diverse providers include diverse patients in our research. Related to that I think it's really important this was something that I think came out in the panel discussion and the presentation about ways that it's important for us to leverage existing resources to understand better variants in diverse populations and whether or not those are actually variants of uncertain significance so we need to focus our effort and our energy there. And then the final point well maybe the next to the final point is the model and the methods that are used for delivering genomic services. Many years ago when I first started doing work in BRCA 1 and BRCA 2 it was sort of the model of meeting with a genetic counselor for an hour and a half to two hours. And now as we're learning more about genetic factors and variants that may not be appropriate for most clinical settings where we want to see more things integrated into primary care so thinking about new models for delivering genomic services and how to leverage our existing resources both from genetic counselors to community health workers to other types of providers is really important. And just a closing comment and perhaps more of a reflection many of the issues related to diversity are things that and points that have been raised over a number of years. And I just wonder and one example is you know in Muin's talk about African American women being less likely to utilize genetic testing for BRCA 1 and BRCA 2 mutations and it strikes me that that has been sort of a consistent theme over a period of time and I just wonder about the need and the value of instituting a really nice term for the translational genomics working group at the level of NIH that really thought about developing tools and resources and disseminating and developing and disseminating best practices for how to address diversity issue and disparities issues across the portfolio of genomics research that's being done at NHGRI. So I'll leave that as a final thought and turn it back over to Ebony. Thank you, Shanita. Just we're going to have the break. We're going to meet back promptly at 11.15 to start the next session so if the moderators, speakers and discussants can be at the front of the room at 11.15 and also we are going to have a group photo if everyone could make their way to the front of the cafeteria. That is where the group photo will be and hopefully if we get there quickly we'll have plenty of time for a room or anything else by 11.15. Thank you.