Using Clinical Informatics to Inform Policy & Equity

Great all right well thank you for the introduction um great to great to be here and doing this talk um obviously clinical informatics um is near and dear to my heart um as the um see let's move through that jump to my next slide there so as as mentioned um I'm currently the associate chief medical information officer here at kilan um but also um moving into um continuing to to look at um well-being and so I consider myself an epic well-being champion and that will play into some of our discussion today so the um the obligatory um disclosure slide um and so uh with this presentation I have no Financial disclosures um but do want to take a moment and talk about some personal disclosures um in the title um we are going to be talking about Equity um and I am not an expert on Equity um I'm not an expert on um diversity in inclusion but I'm getting better every day um but as fate would have it um I am uh I am talking about Equity um and there is a good probability that I may say something that someone finds offensive and so as we start off um that is not my intent uh and I kind of um compare myself to Sheldon um from The Big Bang Theory um and just like Sheldon um sometimes I say things I don't realize that I've offended that person or group and so just to put it out there um in the beginning um if um if I do say something that um that you find offensive U please reach out to me um and let me know and I will try to do better next time so many of you have seen the objectives that we have today um and and those are outlined here um how we can um Talk about Healthcare Equity around clinical informatics um and the impact of clinical informatics that um that we can use um when we're looking at Equity strategies for um integrating um informatics into policy development and then how do how can we incorporate um the informatics into into teaching so really sum to sum this up it's how can we use data thoughtfully but I do want to to lay it out at the beginning here um that this is not a talk about how we're going to use social determins of Health um to drive National Health Care policy um a lot of what we're going to be doing today and looking at today um really is down at the at the local level so having said that our journey for today um broken up into into several different pieces looking at CL clinical informatics and how that can inform policy looking at clinical informatics and Health Equity we'll go through several case studies we'll talk about some barriers um when when looking at clinical informatics and policy inequity um and then no presentation about technology these days would um would be complete without um talking about AI um and so at the at the very end and this is the the piece stay with me until the end because we're going to talk about some really cool AI pieces that are on the on the horizon so to start with talking a little bit about what is informatics or what is clinical informatics and when I think about informatics really it's more about how we collect store retrieve and report on data and informatics in general um is about how we take um take and transform that data so that um we can we can use it kind of in an everyday um situation clinical informatics then is doing that with our healthc care data um and in my role my boss um Dr Steve Morgan oversees our health analytics and clinical informatics departments uh so they kind of both roll up under that one arm that has the the electronic health record that allows us to take input store data and then make that data available back to users of the system um and I work along with um jennif from Martin one of our senior directors in um in nursing informatics and she works more on the on the nursing side and I work more on the on the provider side but um hand inand we bring clinical informatics to kilan so the role of clinical informatics and Healthcare and um and I have a number of images that I'll be using today that that uh pop up and I will tell you that most of these were generated using um generated using AI um using in in chat GPT and so this is chat gpt's um rendering of um how we use use clinical informatics in healthcare and I have three um these three pictures remind me to talk about um the one on the right um is represents our corporate level so how do we use clinical informatics at the corporate level um here at Cilan clinic and one of the one of the ways we do that or several of the ways that we do that is we look at um how many how many providers do we need how many nurses do we need how many um exam rooms um office practices how do we use Supply supply chain information um to make sure that our or are stocked and that we have an appropriate number of vaccines so all of that is is taking in data that that we have to make those types of decisions at that corporate level the next level down um and this is um referenced by the image there on the on the left which is supposed to represent clinical informatics at the department level or the um the individual office level and I love the way that um GPT has um put the the little awnings on there to make the the doctor's office look a little bit like a a restaurant um but at the at the department level we use informatics around um metrics of of how well we are taking care of our patients so what percentage of the our patients with high blood pressure are controlled um how well are we doing on our di diabetes control with a1c's are we screening our patients for depression with the use of the phq2 phq9 um we use informatics around um around work so when we look at our nursing staff that are handling in basket messages who's completing the work um for providers um who is doing U more work outside of work or or while work outside of work pajama time who's working on their notes at night and where we might need to intervene and and help those providers to um on their well-being Journey looking at time and inass get note Lan things like that and then there in the middle um it's all about the patient right so how do we use clinical informatics um when it comes to an individual patient so things like clinical alerts um sepsis scores drug allergy interaction so when when a provider gets ready to to write a prescription for a patient and the patient has an allergy using um using that data in the chart to alert the patient or the the provider that hey you might not want to do that um and so we use clinical informatics at all these different levels and so as we go through today we're going to continue to talk about policy around um C or policy as it relates to kilan clinic and what we do um and I definitely want to stay um steer clear of that that notion of national or state politics because that adds many more levels of complexity um it's complex enough here at kilan especially this time of year um sometimes we talk about um state or national policy healthc care policy um it can become a very emotionally charged subject and so we're going to um we're going to avoid that so we talked a little little about policy I do want to talk about equity and I use this picture here um of the the pizza because this is kind of how this I get in my mind equity and and equality um and so the the example is if we have a pizza and it's divided into eight Pizza Pie eight pieces and I think I have four people um with hands in there I kind of looked at the the fingernail polish in a couple places and um figured out we have four individual people it's like how do we divide up the pizza well if we're going to divide the pizza equally everybody gets two slices um but maybe a more Equitable way to divide the pizza would be you know give more pizza to the person that the people that are more hungry uh maybe less people or less Pizza to people that are dieting U and so there are different ways that we can divide up that pizza based upon some other parameters that might be more equitable um but may not be equal and so when I think about health care and Health Care equality healthare quality not being the same thing as healthc care Equity um and this is U just a a personal comment here um about equity and equality um when we talk about Equitable division um you know the question pops up in my mind who who decides um and and an aha moment for me as I was putting this um this presentation together was that sometimes equal is easier to agree upon than Equitable um and I I don't know if that's if that's unique but um or if I should say you know I was I I'm this many years old when I when I realized that Equitable and and equal are not um are not the same um but there um that was the AHA and just a call out that neither Equitable or equal um will be fair so I um a lot of times we get into what is fair and what is not fair and I try to stay away from fairness as well um so in the end talking about policy and and Equity what we're really looking at here is how can we use our clinical data so that everyone can attain their full potential for health and well-being and that's what I when I think about Healthcare equity and clinical informatics that's um that's what I start to think of the goal being that said um given that this is a discussion about equity and Healthcare Equity um I felt obliged to at least include a slide on social determinance of health and um you know social determinant of Health include those things like um income um housing security education unemployment um J abuse neglect um neighborhood conditions social support um you the ability to obtain food so a lot of different pieces um go into that and when we talk about those from an electronic health record perspective um several several different levels and so the first piece is capturing that data in the electronic health record and for a long time now um that has been varied but we're getting better and and so we're developing those standards so that we can gather data discreetly into the elic health record there have been some tools for collecting that data um a couple of the tools one is um is called prepare and that's prepare with an a which stands for protocol for responding to and assessing patients assets risk and experiences um there have been um several others one called Health begins or well RX and those are slowly um starting to be incorporated um additionally um we can use natural language processing in machine learning and so the ability to um to go through a a a document so progress notes or other patient documents um where we can pull out information from uh from those notes and then use that uh converting it into um discrete language so uh some area some organizations are working on that that's not on Killian's uh road map um so much at this point in time but um but maybe in the future and then there are other organizations that are working um to mine things like social media for race ethnicity suicide ideation um opioid use um again we're not we're not doing that here at corillian but some other organizations are are Leever leveraging that more from a research perspective so once we've gathered the data um another area that uh that is challenging is that sharing of data and so while we have it in our system how do we um how do we share that with our um with organizations that are close to us or even further away so how does corillian and HCA um how do we communicate around social determins of Health I can tell you it's it's it's not well um most of that is in the form of um uh kind of free text if you will so not not discreetly um we do better with within epic to Epic um so working with um with UVA Duke um an NOA um we're getting better on sharing those discrete elements between the two um between those different entities given that it's all within the same um set up epic epic in terms of gathering sharing and so as we move into using data once we've collected that data um what are the different ways that we can use social determinance of heal and and I have have them outlined on this slide here um kind of in the middle um first piece being um once we once we have the data can we find places where um where there are disparities so what are those disparities um doing research into those those areas the second piece once we've figured out that there are disparities where they exist uh taking a deeper dive and trying to understand why um what is what is causing those disparities to exist and then finally um the last piece is doing something about that um how do we reduce that and I will point out um that Rush um uh in Chicago uh a group practice there in in um in Chicago they have done an amazing job utilizing um their um social determin of Health Data um within their EHR they have a um they programmed in the kind of in the in the back end that when a disparity is identified that it generates a consult um in the back to a community organization um that can can help the patient so for example if a patient comes in and we um we've identified that they um they live in a food desert um and so that there there's food insecurity um in the in the background um um nobody has to um nobody has to click a button or get a popup or anything like that it just generates a consult um to um a community organization um perhaps a food pantry um to notify them that now we have a a person um that has some food issues and that Community organization then can can reach out to the to the patient and help out just kind of happens behind the scenes in the back um which is um which is really incredible I think so um last Point here incorporating the social determinance of Health into teaching and I would say at this point um for most of the Learners that I work with and this is would be in the clinical setting just setting an intention to ask the questions and then follow up so whether that's on a on a weekly basis basis picking a different um social determinant on a week toe basis or on a month-to-month basis and so maybe um you know um October um could be domestic abuse or um or food insecurity kind of pick pick your topic but as you're talking with your learners just ask um you know how how is this thing affecting your your patient and in in most cases probably not but every now and then you'll find that one um that one area that pops up that allows you to have the conversation and and then once you've identified the patient with the disparity being able to link them back to Community Resources or resources within the clinic to help out okay so let's transition a a little bit here and talk about uh the importance of involving Equity when shaping policy uh and you my first thought was all policies should be Equitable like duh um but what does that mean um and so I found a a really interesting article this was in the um the Journal of the American medical informatics Association um a a study that was done um the original study um was looking at um uh whether or not a um an electronic summary versus a paper summary for people with implantable defibrillators um uh would um I think would be um more understandable and um and so the the uh the research took place and then they came back on the backside and did some additional analysis and that's um that's what I'm what I want to focus on today and so the enrollment um I have that here on the slide uh they looked at patients with um with icds um these implantable defibrillators um and reached out and obviously some people said no I don't want to be part of the the study um those that agreed to be part of the study they then asked some technology questions do you have access to a computer and do you have internet access because they wanted to send um these summaries electronically and so if they did not have access to a computer or um or Internet they were random or they were randomized they were placed over in what we'll call Group C which was um became the control group a control group and then um for those that um that met the technology piece they were randomized into group a or group b and one group got the electronic summary and one group got a paper summary U but the interesting piece with with this um with this study was that they went back and compared Group C with the intervention or non non-intervention group so groups A and B and what they found was that there were significant differences between these two groups um and so for example um Group C um on average uh 23% of Group C did not graduate from high school whereas only 5% from the combined AB group um in general um income was lower in group C 40% versus 9% um Group C had higher rates of um of chronic disease um so in U in group C 27% had lung disease versus the the other groups at 9% diabetes was higher 44 versus 21% um and so the the point here is that um based upon the methods there is this um enrollment bias that occurs and so here at kilan we're well Korean Clinic we're we're younger to research and so as we start to look at the ways that um our our policies and procedures around study enrollment and research uh we need to keep this kind of thing in mind that um that that we can have this enrollment bias and so if we truly want good data um good Equitable data we need to bake these things into our processes um on the on the front end another um another example of um of some some studies that I took a look at again this comes from the journal American Medical Association but um American medical informatics Association uh I was reminded um back in September they had this announcement um where Rono city is no longer in the top 10 for nonfatal opioid overdoses in the US um and how um how we've done we've we've done an incredible job here in r Oak um to um around our our opioid pieces this particular um article that I was looking at um enforces the need for um for good quality data um and and there's around overdose deaths um not specific uh to opioids but um in in general um kind of all overdoses uh but found that uh people that um were dying from from overdoses in counties that had higher income they were less likely to have unclassified overdose um information um in in the in the death records um also racial and ethnic minorities were less likely to have unclassified drug overdoses and so this leads to um with without having kind of the full picture if we're going to Target interventions in spefic specific counties not having a full data set may lead to um to mistargeting um you know so rather than um you know focusing on opioids we may F we may Target on um meth cocaine other pieces like that and that that applies to kind of all aspects of the data um here at Killian um I mentioned Dr Morgan earlier um but Jason Brown who's our Enterprise data analytics officer it's his role to make sure that our Enterprise data warehouse um that the information that's going in there is as complete um as possible and that we can report on that in a in a meaningful way and he does an amazing job but again kind of coming back to data quality is is key so getting good data um making sure we're storing that in a way that can be reported on um is is very important Switching gears to some case studies um and this is this is my technology and colon cancer um image um and so uh here at corillian um we we obviously do colon cancer screening um and for the purposes of this talk uh when I talk about colon cancer screening um I'm I'm lumping screening and surveillance um kind of all those aspects together and and my my um GI colleagues May throw things at me for for doing that but um but it really is about who who needs a who needs a colon colonoscopy that doesn't have colon cancer um and starting to talk about what what is Equitable when we talk about um colonoscopy screening and and as our surgeons and gastron neurologists are working down the backlog how can we use data in a way to um to help prioritize keeping things Equitable so um many of you may know there are multiple ways to screen for colon cancer colon oscopy being one uh fit testing so looking for um for blood um in the in the stool is another and when we think about people having colonoscopies people with a positive fit test um you that have screamed positive because there's there's blood in their stool um they move higher up the list in terms of people who who need a colonoscopy because now you have a you have a person that has the has a much higher potential for having something going on also people that have had a colonoscopy previously that had issues and need to be um need to be rescreened or or or followed up they need to move up higher on that list compared to say a 45y old with no um no no previous issues um that wants to have that first screening colonoscopy they're going to be a little lower on the list and one of the one of the things that we're doing here at corillian to help with that prioritization is we are using um a third-party system to do some natural language processing on um on old or older operative reports um the procedure reports the previous colonoscopy reports the pathology reports to to pull out information from those reports that may not be um discreet um in the um in the need for followup for a for colonoscopy so if a patient had a colonoscopy done and um it kind of meets the meets the care Gap so the screening interval for colonoscopy should be every 10 years and so the provider says they've had that done and so they're good to go um perhaps in the operative report the um the gastrologist um indicated based upon the procedure that um the patient needs a followup in five years and if the the provider missed that or if that um document came in from a from another organization and the patient has has moved in between hospitals or something and the provider um does not realize that the patient needed a five-year followup versus a 10year followup we can we're working to go through with that natural language process and pull out that data um and present that back to um to the providers to say here's here's a patient that needs um that needs followup kind of outside of what we're currently recommended and so we can then stratify them and get them back in you know um um get them back in sooner so we can we can do the appropriate screening and so again one another place where we kind of look at that Equity who needs to go to the top um based upon um based upon their their risk scores another um another area where we use um informatics um around U around Equity um here at um Cillian for those of you that don't know what this is this is CAC and this is um kind of the the brains of bed management here at kilan put together with Paul Davenport and his team um this is an amazing amazing facility and if you've never had the opportunity to go over and take a look at CAC um it is it's it's futuristic um but basically it's I mean this is a command center that um that looks at um all all aspects of of bed management you can see on this on the screens there um there looks like in the middle they're looking at at weather data and you see maps of the um maps of the area um they really do have their their finger on the on the on the pulse of our bed utilization um you know and they they continuously um tweak um their processes so that we are using um our capacity to its fullest but they um you know they're looking at who needs a bed how many beds do we have uh how many beds are we going to have who's who's going home so um so that they can allot those beds to the people that need them most so not only um patients that are maybe boarding in the ER that need a bed but they're also looking at surgical cases that will be coming in today so um you know what surgical cases are we going to do where the patient won't go home where the patient will need to have an impatient day afterwards they're also looking at patients that are in outlying hospitals outlying Cillian Clinic facilities outlying um other facilities where patients need to be moved to a higher level of care and they're coordinating um all of all of that using um using data and algorithms to figure out um how to um how to assign those beds um as they come available um just truly truly fabulous so another area where um their their policies and procedures are really um aimed at being as Equitable as possible um this is my rapid fire slide um it has several other areas where we um we're promoting Equity through killing Clinic policies and processes um there in the in the middle um that bright colored slide that is labeled Healthcare that's um that's ai's representation of um of access to care in a um in a health system and one of the things that we are doing right now with um with Dr Mike Jeremiah um he's heading up our access to care um initiative and so looking at how we can um how we can provide access to to patients and do that um for the patients that need to be seen most so um in dermatology where we have a waiting list or Rheumatology where we have a waiting list are there ways that we can identify those highest risk patients and um and move them up the list so that they can um see our rheumatologists and and dermatologists and one way we do that um is through the use of EC consults and so um our Primary Care Providers nowadays if they have a question about a a particular um uh patient with a uh with certain diagnosis or maybe an uncertain diagnosis they can reach out to a specialist through an EC consult uh provide some information uh to the um to The Specialist and get those questions answered and so that can do a couple things one it can start treatment earlier so we can start um we can start treating the patient before they get to the specialist the other thing that that that can do is that can help move that patient up up the list so the specialist says here are some things you need to do and we probably really need to get that patient into to be seen um and so they they move up that that list a little bit um a little bit higher because they have um their their needs are are bigger um same type of thing in in dermatology you you have a patient that maybe has something that looks like a melanoma we're going to move a little bit faster to get that that person in than say somebody who has acne um but through the use of e consults we can start treatment for um for one or or both of those in the interum another area that we are uh will clinical informatics comes into play and will come into play even more so in the in the coming years um there on that middle slide that says throughput um that's o throughput um and so that's GPT um kind of visualization for increasing o throughput and I know that um that Dr Tony Spa has made this a um an is a larger issue as we look at 2025 um he's working very closely with um Steven lever in the o to look at how we can um how we can improve throughput and they're using they're using a lot of data to do that so things like or um start times are we starting on time um do we have the the timing for different procedures correct so that we really are making the most efficient use of of of our o our time um are providers using their block time appropriately are there places where where we can where we can do better also looking at the the processes around the O so can we get patients to the O faster um from a from a primary care office um through through the specialist into the O and what are we doing then um post o um in terms of overall length of stay are the things that we can do to continue to decrease that length of stay so that we can input the throughput or we can improve the throughput um I have sepsis on here um sepsis and deterioration um other areas that are um near near and dear to my heart working with u with Dr Kramer and the Caps group um some amazing work in sepsis um helping to implement the Epic early warning or early sepsis detection which is a machine learning algorithm that helps um identify patients that maybe have sepsis that um that we may be missing um and so uh each patient gets a score if you will that um the higher the score the more likely that the patient is developing sepsis and so once a threshold has been achieved uh we can we can notify a nurse or provider um on the nursing side the nurse can do a further evaluation to help us figure out is this really sepsis or if it's not um we've even have gone so far as to implement some um some policies that allow the nurse to implement an order set so some specific orders um none of which are are medication orders but um some lab orders and some some other things that allows us to help um figure out is this sepsis or is this not sepsis um and and then obviously um more work in in terms of treating sepsis so um using that data to help identify um identify the patient so we can intervene earlier likewise we have a tool called the Rothman index which is a sign of deterioration and so taking uh a number of data points from the chart combining those using machine learning we can say here's a patient that um that is is deteriorating we might not be able to identify why through the index but we can identify that that some things have changed for this patient and now we need to notify the provider we need to notify the nurse so that they can step in re-evaluate the patient um from a rounding perspective maybe move that patient to the top of the rounding list um so that so we can figure out what's going on um lastly I have um the the joy of medicine um because when I think about um Healthcare Equity it's about helping everybody um kind of live to their fullest and so provider wellbeing is is is very important and so are there things that we can do within um within the electronic health record or within clinical informatics to help provider well-being and that would be things like looking at um the pajama time or the work outside of work looking at um note length um time that people are spending in in in in their in basket doing that type of work compared to their peers so we know that that providers are going to have to write notes and providers are going to have to be in the electronic um electronic health record doing things but are they doing so more than their peers and if they're doing things more than their peers you know helping to understand why and are things that we can do like additional training for the provider um maybe their maybe their workload is higher than than their colleagues and we need to cut back on their workload um but taking that first step into identifying um you know where do those disparities lie among our providers and then being able to um to adjust uh as we go forward wellbeing I think um we're going to we're going to be seeing more and more of that um in the in the coming years which is a great thing all right um as promised um and I think this is my my last slide or second to the last slide um talking a little bit about Ai and and the future um and the and the future is here so everything that we're going to talk about on this slide um exists today so this is not this is not vaporware these are these are things that are um that are present um and so the first looking at risk scores companies like event and alidade um they are taking electronic health record data claims data and and they're the the terms that I've been I've seen um they they run the data through deep learning models um and they use words like factorization and igen Vector um basically they develop this risk score um or um uh how like how much it'll cost to take care of of this particular patient with all of their their various um conditions so so a cost model um and then they get into prediction so if we do these targeted interventions it will cost this much to take care of the patient so what's that what's that different difference in cost that we can um we can realize if we have these these interventions and in some organizations that's helping to to guide those clinici and decisions so around around cost um the other area um and and this gets to the AI Navigators and I um I think this is is amazing so companies like uh like live person Amelia and deliver Health they're developing tools to translate complex medical jargon into understandable language um and this leads to Patient Empower empowerment and can improve compliance um with medical regimens and and and medical plans um one company has something that's called um they call them answer agents and it's AI um that draws from uh like knowledge documents and and other websites that the um that the company has to answer patient questions so almost a little little chat bodish um but helps um those patients with with that low health health literacy um additionally uh what are called action agents um a little different than the answer agent but the action agent can help automate processes um that en um encourage and enable patients to do um self self-service scheduling can help with with billing that that management of appointments um and for people that that may have a difficult time navigating the system these these AI Navigators um kind of sit on top and and help with those things a digital front door in some cases um and then finally um this notion of a pocket doctor or doct Google and there are a number of of of uh companies that have been popping up um sley adaah Health aund are a few um and these companies um are getting into more diagnosis and personalized care so like the um the Ada app has a um has a system a symptom checker um where it interacts with the patient and supposed ly and I have not seen the data on this but supposedly the Ada app has been better than human doctors and accurately diagnosing rhe some rheumatological diseases skin rashes and the source of abdominal pain um in um and this I believe this was done in the in an emergency room setting um so um these apps are getting better better and better but one of the things that these tools um help with are um around access so when a you know we have patients that are living in places where they don't have easy access to health care yet they do have a cell phone um they can put their information in in that cell phone and get answers that can can help navigate the system or in some cases and this is the um um this is the this is the way that some of these groups make money is that they can then con convert into a tele medicine visit or T Health visit and so they can then um interact with a physician if if they need to okay lots of lots of fun things on the um on the horizon so just just want to bring back do a quick wrap up um in terms of what we've talked about today um clinical informatics should be part of the discussion as reforming policy and so if you remember back that to the pre-enrollment v um pre-enrollment bias um that we saw with the um implantable defibrillators as we're forming policy how can we bring those types of discussions um into that policy um we've got to have good data we've got to have accurate data how do we collect um um social determins of health and other data that will um uh affect patients um and then the last Point um in terms of integrating this into your into your teaching just simply be intentional um pick a um an area um weekly monthly and focus that in asking the Learners um and just kind of make that part of um of the teaching as as you go forward and so now is the time for questions um that's again I I asked Chad GPT to develop a a picture around technology and clinical informatics and questions and so that was the um that was the the piece that that popped up lovely rendition thank you so much Dr speaker this is just lovely yeah who has questions what kind of questions do we have for Dr speaker while we have him with us what that is coming in the background can you hear me yeah hello hi my name is Milena stov I am from raford University and this semester I teach graduate students one of the topic is Health informatics and the use of artificial intelligence and chat GPT I asked the students to create a policy for like support of chat GPD and it was very interesting how the students are more like inclined to not to see the benefits of it but to like prohibit the chat the use of chat GPT prohibit the artificial intelligence so how do you like any suggestions any tip to provide that there's a positive uh like implication of artificial intelligence especially in clinicals um no it's a great it's a it's a great question and um and and one of the things I I think about and and I've talked with um with several Learners about um is I don't think that chat GPT is going to replace doctors and nurses and and other people but chat GPT and co-pilot Etc large language models um people that use large language models will replace those that don't so I I don't think um I don't think Banning um or um or preventing use um is is a good direction to to head in I think the question becomes how do we thoughtfully incorporate these large language models in into what we do um because if if we don't if if Radford and corillian and um in other areas if we're not doing that we certainly know that other organizations will and so as we've seen um with uh you know the application like a to health right people are doing that and if and if and if we're not if we're not keyed into how to use these these tools because that's what they are tools if we're not keyed in on how to use tools correctly um other people will figure that out and and that will put us at a disadvantage um so you've got to understand the the limitations right please don't put health information into chat GPT unless it's secure corillian or a secure Radford um large language model um but certainly asking questions of GPT um will get you answers and then you need to verify them right because you can't trust everything thank you I I do see us moving like it or not we're all moving in that direction right so I like towards the end when you were talking about you know Dr Google just a few years ago I you know you would walk into a doctor's office and they'll be like well tell me what Dr Google told you um and they were sort of afraid of what patients were coming in with what type of information they're coming in with nowadays you know that your patients are coming in with that information and it helps that Physicians are getting on board other providers are getting on board and saying okay let's work together with what you've seen and what we've seen and then wouldn't it be lovely if at some point um we could get get to where uh Physicians and other providers could recommend um some types of AI that were reliable right um as opposed to just not knowing what you're going to get um I would Envision that would help out um from from those patients who instead of um searching or looking into it themselves immediately jump on my chart or something along those lines which can be overwhelming for Physicians who um would like to end their day short of 22 hours um you know those types of things so I I I love the fact that we're sort of getting to the point where we're realizing this isn't going to go away so how do we work together and and really make sense of the resources the tools that we have um other folks what kind of questions we have just a few more minutes any questions or comments Dr speaker great presentation one of the most powerful ways that we can really use um the EMR for research purposes is my chart but as you know we only have about 45 to 50% you know people in it and when you look at who's in it it generally tends not to be people with sdoh so is there ways or mechanisms that we can use to basically get more people on my chart because this would be a great Avenue to for recruitment oh God um you you asked a great question and um and I I don't know of easy ways to to make that happen um I think um I think that there's some opportunities that we have with our patient experience group um to um to continue to refine um some of the some of the pieces around my chart um but you know at the at the end of the day um it comes down to you know access to the app access um online um and and making it easier so I I think it's it's a journey for us here in Southwest Virginia um I think other places um that have higher technology up uptake um may may be easier but I think it's going to be um it's going to be a challenge for us um as we go forward um a question in the chat can you speak to the role that data sharing agreements plan in operationalizing or not uh clinical informatics yeah um yeah data data sharing um is is fun and I I will uh first of all say that I am not an expert when it comes to data sharing agreements that's where um uh Jason Brown and his and his group um are are very helpful but you know corillian has a lot of data in their EHR and and the question becomes I think in my mind is how do we how do we appropriately partner with others um around that organization or around that information and currently um kind of a challenge between Virginia Tech and and kilan in terms of access to that data um I like I don't I don't have the answer but making sure that we have good agreements for for data sharing is is key thank you so very much is there anybody with one additional question before we wrap today's session up quick question anybody comment got a lot of comments in the chat just thanking you for a great presentation okay well uh fortunately Dr speaker is um right here and local with us so um available for questions following our session um he's in Outlook so feel free to reach out to him if you have additional questions or you like to have conversation about this Beyond today's session other than that I hope you all have a wonderful day um stay safe.