 Well, thanks, Mark, and good afternoon. I wanted to thank Terry and Mary for inviting me to come back. I was here at GM 5 when this program was just kind of a twinkle in my eye. And now I'm back at GM 10, and hopefully can share with you some of our successes as well as challenges. So Mission Health is a designated rural health care system in Western North Carolina, or a non-profit system. Here you can see the mountains. Our footprint is in 18 rural counties in Western North Carolina. Here you can see the whole state. This is basically our population mission hospital, which is the main flagship hospital, is basically in the center in Buncombe County. The system is composed of a tertiary care hospital to a large hospital, 750-bed hospital. It is the only tertiary care center in that region. There is five smaller hospitals around, adding about another 400 beds. Our patient demographic is we serve a largely underserved population. There is an increased number of retirees coming in. The demographic is mostly Caucasian, Irish and Scottish, and some Russian-Ukraine. It's a very stable population. One of the interesting components of this is people that are born in this area generally stay there, and people who retire here stay there as well. We have a challenging pair mix of 75% Medicare, Medicaid, or self-pay, and a very highly rated health system and hospital. One of the reasons for me taking on this responsibility of trying to develop a personalized medicine program at this community health system is that they already had really excellent resources upon which we could build. And I think the innovative leadership of our CEO, Ron Paulus, who was a previous director of innovation at Geisinger, speaks volumes. We also had a very established health information technology system with clinical decision support alerts being built in our Surner system. A genetics program of 20 years, the only genetics program in the region. And just a few years before I came, we had developed a new cancer center. Good primary care network. And I'll just add that in 2015, we started an ACO Mission Health Partners. So the personalized medicine program at Mission in 2013 and to the present, the vision has stayed the same. It is to utilize genetic and genomic information to minimize adverse drug response and maximize drug effectiveness in both cancer and non-cancer in cancer, where we already have many standards of care for testing for tumor markers and also new avenues of exploration, especially in response to immunotherapy. In the non-cancer areas, we wanted to focus on those drug gene associations with the highest level of evidence through CPIC as well as through FDA black box drugs, where I consider it's emerging as a best practice. Not everyone may share that consideration. When I first came there, I did an assessment of, OK, let's look into the non-cancer arena. I felt like cancer was a home run and that I wanted to make some kind of an impact in the non-cancer world and looked at the levels of use of the different drugs that were an FDA black box drug. And as you can see at Mission, and we have a large cardiology group that does over 1,400 PCI procedures on ACS patients. And I thought that this would be an ideal group to start with to have clinical implementation. And much of what I'll talk to you today about is clinical implementation. We have some research implementation going on. And as many others can attest to in this room, we did meet challenges in trying to promote pharmacogenetic testing in our cardiology practice. This is not new. In 2013 and 2014, when I came to the cardiologists and said, hey, we're here. We can do CYP2C-19 testing. Here's an auditorable. I can help you with clinical decision support. They basically said, we don't really have a problem in Plavix failures. And so thank you, but this isn't something that we feel is going to help us right now. And in addition to all of the other issues that may be present in cardiology, we also looked at a research project, an observational study, looking at ACS-PCI patients who had been readmitted to the hospital due to a MACE condition and looked up in those patients, those that might be due to a Plavix failure. And again, just an observational study, but about 4% of those patients in looking in their medical records may have not responded well to their Plavix. Unfortunately, at the time, and still now, I didn't have funding to go back and do a case control study, although that's what the cardiologist, I think, really wanted to see. We had many efforts going on, including Jeff Ginsburg coming down and giving grand rounds. As we are approaching another phase of information dissemination and through IGNITE especially, we have new cardiology leadership. And I have readdressed the question of the position on CYP2C-19 testing. We had Rick Stoffer from UNC come down and give a grand rounds talk and also talk about some of the IGNITE data. And that was really very impressive. And I think it started a swell of interest and information. And unfortunately, one of our interventionalists experienced a patient readmitted within eight days. He did the CYP2C-19 testing at that time, found to be a poor metabolizer. So we are in the process of reconsidering this. So in the interim, I was working with other groups and especially the group in pediatrics. I had gone to our VP, our head of pediatrics, and said, what are we doing about codeine and children? And she said, well, I'm really concerned about narcotics use in children overall. Now, most of the kids that come to our hospital are Medicaid patients. So we weren't in a position to be able to test every single kid that came to the door. And following many other children's hospitals around the country, we decided to take codeine off the pediatric formulary back in 2015. And as Todd has mentioned, when I went around to all the pediatricians in the area, everyone said, oh, we don't use it. We don't use it. And yet our database showed, yes, we do use it. And we use it a lot in the EDs when a kid comes in with a broken arm. So having data was very helpful to show that we were using it. And now we're doing some quality improvement studies to just check about how often this policy change and this removal from the pediatric formulary has been consistent. There are still some sites that are not connected to our CERNR system, which is an inpatient and an outpatient system. So we have basically walked up through this list recently developed clinical decision support for carbamazepine in the process with digitize of doing this for a back of year. Each one of these require submission to the P&T committee and also a clinical champion for each one of those areas. So the strategy in trying to develop this program continues to change based on really where the interest is and where the support is. But when I step back, I was really focusing a lot initially on the inpatient mission health system to try to develop point of care decision support tools when a drug was being prescribed. And in that process, making sure that there was an orderable, making sure that there were drug alternatives while you were either waiting for a test to come back or deciding against using that drug. In other words, in coding, we had alternatives at different age groups. And then realizing that there's this huge primary care network throughout the region, we have a very strong AHEC called the Mountain Area AHEC. And what I was trying to do then was to try to develop some pilot feasibility studies in the outpatient setting to try to push preemptive testing in the outpatient setting and still have point of care clinical decision support for those on the mission EMR. Not everyone has a mission site. We do have an accountable care organization, but there are still many primary care practices that are not part of mission. So let me just step back a minute, and these are our current services that we provide as well as our team. I can't stress enough the raising awareness and education and training and being a resource. And this continues to be a challenge for both educating clinicians in the region as well as patients and communities. We help to develop policy and implement those policies and best practices. We work with the informatics groups to develop the actual language in the clinical decision support, and thank you to Mary and many others who have helped in this regard. We also, especially in cancer, conduct QI outcome studies, which is done mostly manually to ensure that we are complying with the policies we've put into place. And recently, in 2016, we started with a clinical research study that was funded by the North Carolina Biotech Consortium, in which we were hoping to bring pharmacogenetics into primary care. And I will show you a few slides in just a minute about that clinical research study. But these are feasibility pilots. So the pilot in primary care, bringing pharmacogenetics to primary care is about a year into the study. And we just started a supportive care study in cancer looking at pharmacogenetics for cancer supportive care. And both of these studies are basically just to get physicians used to using pharmacogenetic testing, when especially the major barriers are removed. So we have funding to cover the cost of testing for a limited number of patients in the primary care study. It's directed at patients that are 65 years of age or older on Medicare polypharmacy, where one of the drugs is one of the CPIC guideline, High Evidence-Based Drugs. And we also provide education and training. And we have a clinical interpretation summary that goes along with it. So it's basically a way for physicians who have said to me in the past, Lynn, you're not even on my radar. I can't even think about pharmacogenomics to just try it and see what the real barriers are when cost and education are taken off the table. And since the arrival of our clinical pharmacist, Dr. Jillian Bell in 2015, we can now offer clinical consultation in both cancer for genomic profiling and in non-cancer as we have a clinic that we started last year. A lot of our work is really moving silos into a matrixed integration philosophy as everyone is working in their individual areas, and yet our program cuts across all the different service lines. We have a small but efficient group of people. My VP, a hospital administrator, is very supportive of the program. I have three full-time people. Myself, the clinical pharmacist and our program coordinator, through some of our research efforts, we've been able to get funding for a half-time research nurse and a pediatric pharmacist who's been the NICU pharmacist for 10 years and wanted to do some professional development. And so he's actually learning pharmacogenetics spending a day a week in our facility. Trainees come through and do rotations in our program, and they're very valuable resources to learn from and also to conduct a lot of our smaller studies. We really couldn't do this without our consultants, Howard McLeod and Mark Dunnenberger, and without all the partners that have been really so responsive and so generous with their time that are listed here and many others not. So let me just go over a little bit what we're doing in cancer. When I came in 2013, we were still bringing in disparate oncology practices under one umbrella. And so what I tried to do is have a way to more or so bring people in for a consistent way of using tumor markers, developing the first kind of integrated tumor marker program at Mission Cancer Center. And it was basically to walk through each of the major disease sites for which NCCN had guidelines for tumor markers, make sure we were all doing it the same way, and make sure that this was common among all the different oncology groups. And also to bring together the different disciplines oncology, pathology, genetics, pharmacy, whomever was appropriate, to develop our internal guidelines. And that was something that people really hadn't done together. We've instituted quality improvement studies for tumor markers to ensure compliance with evidence-based guidelines. So for example, when I first came, we were looking at our non-small cell lung cancer patients of the adenocarcinoma variety and looking to see if we were doing EGFR-mutational analysis, ALC, and ROS on all our eligible patients. Well, we did a QI study after we implemented that policy of doing that and saw that 62% of the eligible patients got tested for at least EGFR mutation, if not the others, if there was sufficient tissue. And that the 32% of those patients who didn't get tested, there were really good clinical reasons why not. They refused therapy, they went to hospice early, or they left and went to a different institution. So a lot of what we do in personalized medicine is operationalizing process from the point of test ordering to sample submission to an interdisciplinary problem-solving, which also includes interpretation of results. And that's something that Dr. Bell is doing as we have advanced to have a lot of different genomic profiling laboratories and opportunities that we interact with. One of the things that Dr. Bell does is provide a clinical summary of these 20-page reports that's specific to the individual patient and actually looks up if there is a drug that matches their mutation to make sure that that patient, based on their other clinical characteristics, is able to not only take the drug, but if there's a clinical trial, is actually eligible for the clinical trial and that that clinical trial is really open somewhere in North Carolina. When needed, she also helps the physicians with some of the appeal process. The other research study that we're just starting right now is really to look at the feasibility of providing pharmacogenetic information to physicians, to patients, at the time when patients are starting on chemotherapy. This is for moderate or highly emetogenic chemotherapies in which at the time of education about their chemotherapy, the patient is asked if they'd like to participate in this study. Again, the testing is done free of charge. The physicians now get exposed to more of the germline orientation. So I've listed here some of the challenges in coordination and communication. A lot of the data is not in discrete fields, so manual data abstraction is quite cumbersome. I'm just going to go through some of these other challenges. In the CDS for drug gene pairs, we don't have a pharmacogenetic subcommittee of our PNT. Our group is very alert averse. We do have a personalized medicine clinic that we started last April. And most of the referrals, and it requires a physician referral, most of these are patient driven. And these are generally patients who are very complex cases that have heard about personalized medicine and are coming to their physicians. We are hoping with more awareness and some of these feasibility studies that we'll be able to improve that referral rate. Funding is another challenge for us as a community hospital. I really don't have startup funds and work on smaller grants as well as philanthropic donations. So I'm just going to end here. I had this slide and one other just to let you know that one year into this pharmacogenetic feasibility study in primary care, we have four practices, 10 clinicians, 32 patients who have had their pharmacogenetic testing completed. We have recommended changes to current drugs that they were on in 38% of those. Now again, these are pretty targeted patients. And we can go later over some of the barriers in this, which are not uncommon to what other people have described. So in closing, our lessons learned, and they continue to keep coming, is that having a top down as well as bottom up physician support is needed in our system. Solving existing clinical problems was very important rather than just saying, hey, we have a personalized medicine program. Don't you want to do some testing? Timing is important in terms of responding to national concerns as well as new professional guidelines coming out. So the new tramadol and codeine, we will be updating our CDS in that. I'm hoping that a pilot try it kind of approach is a way to help clinicians who are uncertain of the clinical utility of this testing in their practice and are not willing to try this in a clinical setting, but would much rather do this in a research setting. And I wish I would have done more work on applying innovation theory and implementation and diffusion science as well as understood a little bit more about behavioral economics up front, but we learn as we go. And over four years, I think I'm getting to the point where we understand a little more about what motivates our organization, our community, and clinicians in terms of changing approaches to health care. It's a work in progress. So thank you. Thank you, Lynn. Josh, I'll have you go up. We have time for one quick question, and I'm going to take it from Rob because you're going to have to come to the mic. We'll get you for the discussion period at the end. Not quick enough. OK, quickly. So I have a very nice talk and a very interesting approach. I'm from a physician or provider education standpoint. One of the challenges we have in designing training programs is understanding how well they're able to implement that information. So we can test their knowledge after we've given them a lecture. But that isn't guaranteed that there's going to be a behavioral change. So you had some things where you were able to go into the EHR and look at who had tested. And so with how hard were those to implement? Is that something that you think can be done generally to see after training what the difference is in how they actually take care of patients? Well, that part of the research study for the primary care, we're still in the process of finding out what the physicians are doing with this information. So for those that are in our EMR, we can go in and look to see what changes have actually been made. We make a clinical recommendation or suggestion into how to change care, whether or not they do that. That will be in the second year of the study to look to see what was done with the information that was given out. So I don't have that information, but that is exactly what we're trying to understand. Great, thanks very much. So our last talk before our discussion period is Josh Denney, Vanderbilt University, someone who needs no introduction. PGX implementation research programs at Vanderbilt.