Okay, it is 2.45 and we're gonna get going. First of all, I want to thank the diehard audience members and panel members who've stuck this out and are here to join us for it. I think it's really going to be an exciting session because you've heard a lot of exciting talk about AI in these last two or three days, but we're gonna kind of bring it back to earth and we're very honored to have people who are in the clinical world, people who are in the business world, and people who have a foot in both worlds sort of to talk about their experience and where we've been and where we may be going. So let me introduce the panel. My name is Rod Passman. I am an electrophysiologist at Northwest Memorial Hospital in Chicago. To my left is David Koenigsberg, who's a EP in South Florida, who is a founding member of the Florida Heart Rhythm Specialists and a clinical assistant professor at Nova Southeastern University and has IP around early detection of MI. To his left is not DJ Laccaretti. It's Tom Deering, who many of us know as the past president of HRS and also the chief of the Cardiovascular Center of Excellence at Piedmont Heart Institute. To his left is Dr. Jose Nosario, who's the market director of Cardiac EP at HCA Miami. He's done a great deal of work on quality and practice improvement, and he's the president and founder of the Heart Rhythm Clinical and Research Solutions LLC and 3PH Alliance, and he's leading national efforts to track improving outcomes, data collection, and improving quality in participating centers. To his left is Dr. Edmondo Robinson. Dr. Robinson is the former senior VP and chief digital officer at Moffitt Cancer Center at UCSF, where he founded and led the Center for Digital Health. He's also, importantly, the founder and CEO of Downey's Digital, whose mission is to leverage digital innovation to accelerate the journey towards health equity. To his left is Richard Stobridge. Richard and I spent lunch yesterday. I'm excited to have his input. He is the VP at Commure, who's an AI medical scribe, and he's the CEO of RX Health, an AI-based clinical intelligence platform, and we're really excited to have him. And on his left is Dr. Paul Varosi. Paul is the director of Cardiac EP at the VA Eastern Colorado Healthcare System. He's a professor of medicine at the University of Colorado, and he's had really important leadership roles in the NCDR, HRS, ACC, and AHA, and his focus is on performance measures, health policy, and quality improvement. So we have a great group with us, and I'm gonna sort of hit the ground running. When I asked my phone what you define as a digital care pathway, it says that digital care pathways employ digital technologies to follow and support patients through their healthcare journeys. So I want to start with the clinicians. We've heard so much about AI and what the future is, but there's already so many opportunities out there to use digital health to help educate our patients, to sort of help guide them, and I guess I want to sort of start with Paul. How do you use what's available presently to educate your patients, to follow your patients, and what have been your sort of successes and failures, and how would you tell the person just starting practice what they should be using? Well, thanks, Rod. What I'll tell you is that in our system in the VA, for actually several years, we've had a program within the VA called the Clinical Assessment Reporting and Tracking Program, or CART, and CART is both a quality management program and also a technology. It's basically a structured reporting platform that allows for structured data capture to integrate data collection into the process of care with the concept of collect the data once, leverage the data multiple times. So we're capturing information about the patients, the procedures that are being performed, the indications, so we can look at guideline concordance of indications for procedures. We capture procedural outcomes, including safety events. We capture data on the details of the procedure. So it's a way of looking at real-time information on what's actually happening, like an air traffic control system with structured data. And then that information can be simultaneously leveraged to produce the procedure reports. And for our system across all cardiovascular care, that's a couple hundred thousand procedures a year within the VA system. And we can do real-time safety monitoring. So it's all the things that would go into capturing data for clinical quality registries that in other systems are separate, reduplicative, after the fact, some full-time person pulling the data out of the system for the NCDR registries, we're able to do that in real-time within the system. There's a lot that we still want to do that we don't have the resources to do yet, but we can dream big. And this concept of gathering all the data we need, capturing it one time and using it multiple times is really what I think is very exciting. I think in some of the other talks, it seems like maybe that's the lowest hanging fruit of AI. Take those tasks that used to be some FTE going through records for quality assurance or billing or whatever, and just let AI loose on that. It seems to be the low hanging fruit. Completely agree. And this is a point that I'm taking home from these few days and really excited about trying to find ways to do that in our system. So Tom, I know you've spent a lot of time and effort thinking about remote care of patients, but how do you use what I would call digital health, what they're calling digital care pathways, not only to maybe educate our patients, but follow patients? And what do you think of the major innovations that you've seen over your career and what do you'd like to see in the future? I think right now that we do have, unfortunately, a significant gap between what patients need to know and how much we educate them about. And I think we need to work on filling in that gap. What we also do, and to build on what you just said earlier, Paul, what we also need to do is look at what the guideline-directed medical therapy tells us we need to do and make sure that we put programs in place and educate patients about how their particular care will be better enhanced if we do that well. So for example, you asked about remote monitoring and remote options. I'll just give you two things which we're doing right now. And we do it in partnership with our third-party vendors very, very frequently. So for example, we all know that if you have atrial fibrillation with a certain thromboembolic risk score that you should actually be on anticoagulant therapy. We know across the world that that's somewhere in the 60 to 70% range in general. Now, there are reasons that wouldn't come to place like patient refusal or other contraindications, but what we have done working with Pfizer and BMS is put together within Epic a program where we can identify all of those people. That's the first part. Then we have a program where we can educate our patients about why they should need to be in it remotely or by nurse calling them and interacting because it's a mixed bag right now. And then thirdly, we track those results. We track it systemically. We're a 25-hospital system. So you can see there's going to be a lot of variation there. So across the whole system, we look at it by system. That is hospital A, B, and C. We look at it by specialty, EP, cardiology, general internal medicine, et cetera. And we look at it by individual clinician so that they get that data and they can then hopefully use that data that we've been able to derive electronically and drive it forward to improve. We also get together on a regular basis to review this data within our subsegmented groups. I'll give you another example really quickly and then let you move on to the next person. But for example, we do know that inappropriate shocks and appropriate shocks are problematic for patients, both because of the psychological issues and the clinical issues. So we have worked with one of our partners to identify a third-party vendor that we use for remote monitoring of implantable devices to identify people who are having long runs of non-sustained VT, any therapy such as ATP, so that we can get them in either for pharmacologic therapy, for ablation. And we do this across the board with a broad group of individuals. We have palliative care folks. We have heart failure folks. We have folks that treat for sleep apnea so that they all get this data, like you said, Paul, in a unified manner. And we can approach all of those things so that the right therapy can be delivered. And it's oftentimes multifactorial. But to sort of go back on what you said, it's clear that what you're describing is very a multidisciplinary approach to what manifests as an arrhythmia issue, right? Correct. But we've not been very good as a system in sort of breaking out of our silo, partnering with other subspecialists. There's no payment model for the time that a sleep person may take and say, is this a sleep issue that's manifesting? So how have you made that work and is it exportable to other institutions? I think to some degree, it is exportable to other institutions, but the way that we've made that work is, I think upfront you have to, before you start looking at the technology, because the technology is a tool, although I think maybe AI will be someday a partner of ours, but it is a tool. So what you need to do is bring all of the players together, administration so you can get some fiscal and operational support, and then you have to bring all of the doctors together. So for example, Rod, on that example that I gave with identifying people with ventricular tachycardia treatments, what we would do is bring together the palliative care team, bring together the heart failure team, the sleep team, and say, this is how we're going to address these issues. Are you on board with that? So when electronically something comes up and we see something is missing, they will address it. Now there is minimalistic payment for that, and I don't think that's gonna change in the short term, but there are consultative fees for the sleep people seeing those individuals and palliative care at many institutions. Although at our institution, we have it at our mid and our large size institutions, but we don't have it at our smaller institutions. So that becomes a problem there. But what we try to do is bring up front, create the, communicate about what the gap is, get buy-in on that, and then put together a program and use the digital data that we derive to continue to improve it because we will, we do, and we have had made mistakes. Good answer. I want to switch gears for a moment to sort of, again, understand what is available to us as clinicians today and what works and what doesn't work. So Jose and David, this next question is for you. You know, you're both very busy electrophysiologists, and it's very important, you know, that we're using all the technology out there both to educate and follow patients. So I'll start with Jose. I know you've set up a great consortium where you're doing long-term follow-up of patients at multiple institutions to talk about outcomes. So how have you leveraged technology to do what would have been a very difficult task to do even a decade ago? So I'll tell you, my journey started because when I started doing afib ablations right after fellowship, I truly hated doing afib ablations because I felt like we're doing a procedure that we're seeing all these redos. I don't understand what the success rate is. I don't understand the complication rates, and what am I telling my patients about the chances of success? So I had a moral dilemma with the understanding outcomes and how I present a procedure to a patient. So that's how it all started. And the first thing I feel like what we struggle is there is this very distinct separation between clinical practice and research. And then quality improvement becomes this thing that no one cares about, but that's really where we should sit in terms of what patients care about, which is understanding how to measure outcomes in a way that you can talk to a patient. And patients don't care about the effect score. They don't care about, you know, that you don't have 30 seconds of afib. So how do you collect data in a very pragmatic model? And so what I did is find a pragmatic way and find a way to make it sustainable. How do you leverage technology for this? That's, it's getting ever better. So we have, we develop a software that allows the idea of building a database that it's used for multiple purposes, maybe for patient transparency, benchmarking, or physician publications, but also for policy, for label expansion so that you allow industry to partner and help you fund that data collection effort. But now the new frontier is to use, you know, digital biomarkers and use, you know, patient reported outcomes, not in the way that we're doing right now, but in a way that patients care about. Something as simple as in our practice, we didn't use effect. We used a, how do you feel after your ablation? You know, I feel markedly better, slightly better, I feel worse, so on and so forth. Patients understand that. They don't understand the effect questionnaire, and this simple question correlates a lot. So I think the digital tools today allow us to be able to speak the language, collect the data and speak the languages that the patients want to talk to us and speak to us, and offer a tool for us to build bridges between institutions, particularly, and I think where we have failed for a long time is community doctors have been completely marginalized in that effort of collecting data and pushing forward, essentially, care pathways. And so now we have a chance to do it with collaboration and digital pathways. I want to follow up, because I think part of that data that you're collecting, you know, now that everyone's walking around, you know, look on the panel here with their watches and their cardio devices in the future, their rings and their who knows what, I mean, you know, could you envision a time where we're always collecting data on our patients, right? Where we don't have to bring them back for, you know, medical grade technologies, and that will be an acceptable endpoint for what we are doing. So can you foresee that, or are you there yet? We're not there yet, but I hope we get there very quickly, right? And not every patient's going to care about it, but a lot of patients will. So we need to create the pathways that patients choose to partake. So some will want to be extremely data-driven and have access to every episode of AFib they have on an ILR, on a digital, on a wearable device they want to know, they want to be a part of the decision. Some don't. So I think we need to build all different pathways to really be inclusive, based on how the patients see their own journey. So David, I'm going to sort of bend the question a bit for you. You know, you mentioned that you started your program right out of fellowship as a solo practitioner, and I'm sure you've seen things change. So I'm sure you want to, you know, obviously educate your patients in an efficient manner, in a thorough manner. So which kind of tools do you use that you have found to be effective? And do you feel that all of this digital health that are available to us, A, makes your job easier? B, makes your patients more anxious? Is it allowing you to find things earlier? So is it a pro or a con? Thanks. So yeah, like you said, I started my practice in 07 out of fellowship and I've seen quite a change in that period of time. I'm located in Fort Lauderdale and our patient population in Broward County is 10% of the patient population is over 80 years old. And I have kind of a bimodal distribution of my patients. So I have patients who come in with their Apple watch with a diagnosis of atrial fibrillation, asking for an ablation, because they studied with Dr. Google and they know everything there is to know about atrial fibrillation. They ask for the energy source, they ask for, you know, what kind of mapping I'm going to use, if I'm going to use intracardiac echo, if I'm going to use fluoroscopy, they're very in tune, involved, and they use all the information available to them to make their decision to see a doctor. And then when they see a doctor, they want to order off the menu. And then you have the second group of patients who are, you know, octogenarians, who are referred for watchmans and some of them for ablations. Yes, believe it or not in South Florida, we do lots of ablations and people over 80. And these patients don't understand when you hand them a digital brochure made by Barostim or a brochure that says, you know, go to this website to look at your video on how a watchman is done, or, you know, refer them to different types of digital media, online media, how to get there, what to do, what they're looking for. So these patients need an old fashioned explanation with a chalkboard or a piece of paper and a pen, or some, you know, something on the wall or a heart model to tell them what exactly we have to do. And that takes time. And no matter how fast technology advances, it still will require someone to sit them at a computer and show them the video, or sit with them and explain things to them. And I don't think, you know, that Apple's gonna figure that out for us anytime soon. So I think that as a clinician who's very busy in a private practice setting, yes, there are tons of resources available to me and I utilize them, but I still feel that patients are somewhat left out of the equation, especially in my situation, the elderly patient population. Well, you know, you raised some really good points and I sort of now want to sort of pivot to Edmondo because I think you brought up an issue that deserves special attention. And I sort of, you know, we have these wonderful tools, right? Cardiology, the reason I think HRX is here is because you can measure the ECG and you could measure a PPG from a relatively inexpensive device. So I think that's why we are at the forefront. You know, Edmondo and I were talking beforehand and I told him that I could show him a map of the city I live in, the city of Chicago, and I could show you who owns a Apple phone and who owns an Android phone based on where they live. And then I could put an income map directly on top of that and it correlates 100%. So, you know, we have these great technologies, right? But A, are we just furthering the quality and care gap by having these, you know, commercially available that are not paid for by insurance, number one, for the most part. And then I'll follow that up with this other piece of thought that, you know, whereas you talk about AI and understanding using EHR and all this digital health technology, if we know already that underrepresented minorities are less likely to get anticoagulated or less likely to put on rhythm control strategies or less likely to get ablated, then are we just perpetuating those failures of the system? So we only have 25 minutes. If you could figure this out, I would appreciate it. I got you, I got you. So no, great, great points. And here, so look, there's, spot on. The challenge with the digital innovation and really the rapid pace of the adoption and implementation of digital, one of the challenges is that you're exacerbating disparities, right? So we know that we have health outcome disparities and we know that many of them track along income, education, these other things, access, these points. And we also know that when you add more technology, you're gonna exacerbate those disparities. So, you know, there's a couple of different concepts that we should, at a minimum, understand, if not follow. I think we should follow them, but at a minimum, we should understand, right? One is really kind of this idea of digital equity, where, like, which is really straightforward. I think, I really think it's table stakes, which is try not to implement a digital innovation that's gonna exacerbate disparities. Like, just like, let's not do that, right? That's, to me, that should be table stakes, right? Which means, though, that you should understand what your disparities are, and you should be measuring in a stratified way what happens when you produce an implementation and implement that. You actually have to measure, right? What's happening across your different demographics, your rural patients, right? Your different socioeconomic status, you actually have to measure those things and track them and make sure that you're addressing those disparities, right? So that's the first thing. To me, that's table stakes, okay? The second part, though, is where I'm really excited, where I think we should be, which is, can we design digital innovations that are purpose-built to actually eliminate the disparities that we already know exist, right? So it's one thing to not make it worse, but can we actually make it better, right? To your point, right, we know that there's disparities in the folks that are getting anticoagulation and ablations and so forth. So what can we do from an innovation perspective to actually reach out, grab those patients, and actually bring them in and start to trim that gap in health outcomes, right? That's where I think the real promise is. But it does require that you're very intentional about how you design those tools and how you design those implementations, right? Which means in my, you know, so what we work on from, you know, in DownEast Digital is, you know, how do you bring the actual end user to the table early on, right? As opposed to saying, hey, let's all like go into some windowless room somewhere as a bunch of docs and like come up with all the answers and then, you know, thrust them onto the community, right? Why don't we actually have the folks we're trying to serve there early on and say, wait, you know what? That's not going to work, right? This, why don't we go this way? And then we have to be maybe humble enough to listen and maybe pivot around what the community is telling us. If we did that a little bit more, we might actually start to bend the curve around these disparities that we, again, we've been documenting, IHI and others have been documenting for decades. This is not new, right? So, but what can we do? How can we use digital innovation to actually start to move the needle the other way? Let me ask you, I mean, you say it's sort of a non-starter that we shouldn't do anything that may exacerbate the divide. But the other viewpoint is that you need to start someplace and technology gets cheaper. And if you figure it out with an expensive technology, ultimately, you'll get a cheaper technology that can be more generalizable. So are you arguing that we shouldn't even ask the question if the... I'm arguing that I think the reverse is actually more true than what you said, right? So you're saying that, hey, we start with the expensive stuff, then it'll trickle down. I actually think the reverse is true. So I actually think that if you design for the lower economic status first, it actually much faster rapidly will progress to the other side, right? So I actually think we're approaching it the wrong way. I actually think you should start... So for example, and I actually saw a panel, I think it was yesterday on this, where if you designed... For example, they went to a couple of Western African countries, and they designed solutions around that. And they realized, holy cow, this works perfectly in Baltimore, right? And you actually didn't even have to... There's not even a whole lot of change you have to do, you just deploy, right? So I actually think you start there and go the other way. So the follow-up on my other question, though, if we don't have... If African Americans are getting less ablation, right, than other race-less ethnicities, what is the AI going to tell us about follow-ups in a group that aren't well represented in the data to begin with? I mean, enrollment in clinical trials of underrepresented minorities, particularly in cardiology, particularly in EP, is dismal. So how are we going to reverse this? The AI ain't going to help you there, right? So this is a Geico issue, right? Garbage in, garbage out, right? So if you train your AI models on all the stuff that you've always done, which actually produce the disparities, guess what? You're just going to replicate the disparities. There's no magic bullet there, right? So I think, again, I think we got to go back to our fundamentals and actually engage with the folks that we're trying to serve. So this is what happens. I spend a lot of... I spend my time in oncology, right? And so we have a similar issue with getting folks into clinical trials, right? Whether it's in rural areas or other folks, the answer is not, hey, let's figure out how to do an AI clinical trial. That's not the answer, right? The answer is just, you know, rolling up your sleeves, engaging the community, understanding what the needs are, and figuring out how to meet them where they are. It's not... I wish it was an easy, like, tech answer, and unfortunately, I don't think it is on this one. Well, I'm going to make sure that I engage the person to your left because, Richard, you and I spoke yesterday, and when I first heard about what your company does, I just... My pulse rate increased because I think that when you talk about AI and how it impacts our day-to-day lives, and maybe less so in electrophysiologists, but certainly primary care doctors, people who spend their time seeing patients, you know, to me, this is just an amazing application. So would you sort of talk about what you do and where you... How this may impact the quality of life of the doctors themselves and perhaps the patients too? Yeah. Absolutely. I'm very happy to be here. I don't... I'm not a physician, and I don't play one on TV. I'm actually a trained telecommunications guy, and I've been doing healthcare for about 40 years now. And I think we're really at an inflection point. I look back on my career, and I think I've seen two other inflection points in healthcare that might be close to comparable to what we're seeing now with AI. What I was doing was I was the CEO of a company that was spun out of the Innovation Center at Mount Sinai in New York. And what we do is a little expanded sort of definition of care pathways because really most of what we do is we engage people outside of the clinical setting. So I can reach out, and did during COVID, actually, I reached out to a million people in two days and educated them on how they could sort of self-triage, don't come to the emergency room was sort of the message, unless you really need to, right? And that company grew through COVID and after COVID, and we recently sold the company to another company called Commure, and now I'm a VP at Commure. Commure adds the AI portion of that to what we were doing. We would reach out and still do reach out to either one patient at a time who may have an elective surgery via SMS messaging or IVR, and to your point about, you know, the octogenarians and being able to get to them, we do language, we do interactive voice response if they don't have a smartphone, and we've worked with inner city populations, very rural populations in all different ages, and we find that we can overcome this, you know, the digital disparities that we see as we see them. And we've had an NIH grant for the last five years and did some of the very foundational work on digital equity and continue to do that today. The whole idea is to be able to get out there, educate, you know, give them, you know, bit by bit by bit pieces of information instead of giving them a load of information or force them to go to Dr. Google. We give them, you know, curated information from our customers, our health system customers' own websites. We provide them with additional information. We use chat bots to collect information from them. So it's not about how is that patient doing the last time we talked to them, the last time they were in front of us, but how is that person doing today? How are they doing this morning? Collecting information from remote patient monitoring devices, ePros, you know, all of those kinds of things, and being able to do an outreach that never was possible. And now with the AI capability, we can individualize that outreach and really personalize it even if we're reaching many, many, you know, tens of thousands of people. We're working with a population in the state of Nevada, 220,000-person managed Medicaid population, eight community social workers. It's an absolute impossible task to be able to reach these people and really understand what's going on with them. We can do, you know, SDOH screenings. We can do health risk assessments. And all of that information comes back and is compiled and is available within the electronic health record. So I want you to talk about sort of ambient listening as well, but I also want to expand on something you said. There was an article in the New York Times a few weeks ago, which kind of blew my mind, right? Because we're used to going to these meetings and speaking to colleagues that we train with, and we all are in urban centers with lots of competition, but there are huge swaths of this country without any cardiologists, period, certainly no electrophysiologists, right? So how do you sort of take this model to, you know, educate even primary care physicians or patients who should be being seen by cardiologists and get them in the right hands? Is there any role of what you're doing to help with that issue? Well, you know, I think there's an educational component that we provide as well to the primary care physicians to, you know, these are the kinds of things to look for. And we've worked with, you know, very diverse specialties in our customer base, Jefferson Health, Yale, New Haven, Mount Sinai, of course, and the UC system, and really are trying to understand, like, what are the key components to look for, you know, within that patient population as you're seeing them? And then also create a roadmap where we can now outreach to those patients automatically using the device that typically they use every day, which is their mobile phone, and collect the information that you guys as experts need to be able to say, hey, we need to get to see these people right away. We're doing one of those right now with an aortic aneurysm kind of model where we're using AI to look at the size of the aortic aneurysm, and then outreaching to those patients, you know, after they've been triaged by you guys to get them into surgery faster, get them scheduled faster, get them educated faster, and sort of move them along that path. And that's been very successful. So, Rod, let me just make a quick comment. You asked a question about the provider, and he was discussing about the provider, the physician. One of the things that was really curious to me about the journey on outcomes assessment was that just doing AFib ablations and not knowing how much you're impacting your patient's lives ain't fun. At some point, it just becomes some, it's just a job, it's just a task, right? So, the best way to address physician burnout is to address physician's intrinsic motivation, and relatedness, autonomy, and competence are the three pillars of addressing intrinsic motivation. So, I would argue that this, you know, patient education and outcomes assessment actually brought back to me such a joy at the fact that I'm an electrophysiologist and I'm impacting patients, and I didn't have anything objective before to measure how much I was impacting my patient. It was very subjective. I do an ablation, I see them in follow-up, I'm doing okay. Now to know, because of this, because of, you know, digital tools, outcomes assessment, that this is the true impact, I think it's also an interesting way to address physician burnout. That's a great point. I mean, I guess if we didn't want to follow our patients, we would have become emergency department physicians. So, we're doing this because, you know, we do care about those outcomes, and for many of what we do, it's quality of life, right? Which we don't get objective assessments of when we simply do a monitor on them, for example. So, I think more and more, that is going to be important. But it's sort of, my follow-up question to you, Richard, is how do we balance these privacy issues, right? What are the amazing advantages that we could learn from digital health versus, A, sort of to David's point, the people I think most tuned in and most likely to give us that information may be least likely to benefit from it, all right? And those people who are most concerned about their privacy, least likely to sort of trust the system, may be already disenfranchised to begin with. So, I'd like to hear both Richard and Armando's perception of that. Because in the United States, I was hearing a roundtable yesterday, and like in Germany, they couldn't do any COVID tracking, because none of the German people would allow even their location to be known. So, how do we balance those issues? Because we all believe that if we could, you know, I won't say get over, but we all believe there's a great benefit in having that information available, not only to take better care of patients, but imagine the research that could be done and the resources saved to do trials if we had this information accessible. So, there's this imbalance, though, of who's willing to share information. So, I'd love to open it up to whoever wants to talk about that. Well, I guess I'll start and I'll give it to Armando, because I think he's got greater expertise in this than I. But, you know, I'm just looking out across the country that we serve, rural, urban, you know, food deserts, I mean, you name it, all the disparities are out there. And by bringing the kinds of models and techniques that we're bringing, I feel is really reducing that equity gap. It's not perfect. Not everybody has a cell phone, but many, many people do. And through our work with some of this digital equity, you know, the research that we've been doing, you know, we know how to get them cell phones if we want to get them. And then we've run into other problems. I mean, we've run into problems with patients that, you know, live under bridges, they can't get their phone charged. So we can get them a phone, we can pay for the phone, we can teach them how to use it. But if they can't keep it charged so it's functional, then it just is even that much more difficult. But those are the kinds of things that we're running into. But I think that, you know, as time goes on, we're being able to, you know, more personalized, better understand those individuals' problems. And so, yes, I think that's spot on. And I think there's another component I'd want to add in here. So part of it, so one, generally, trust in the healthcare system has been eroding, right? Certainly, and that's been accelerated since COVID, right? So that's part of the challenge is that we have a bit of a trust issue, broadly speaking. But I think there's even a more equally fundamental issue, which is that, you know, we are implementing innovations that actually make people's just regular lives or jobs harder, right, and not easier, right? So we do this to docs all the time, by the way, right? So every time there's a new innovation comes out, that's another click for you, another thing you got to do, another, another, another. Why can't we innovate and make their job easier and not harder, right? So that's why I get a little bit excited about the generative AI stuff and, you know, some of those options, which, by the way, it's just a workaround because our entire approach to documentation and the rev cycle is completely screwed up. So then we have to do a workaround called generative AI to fix all that stuff for us, right? It's a complete mess. But, you know, we keep making the job harder. And it includes for our patients, right? So if we're going to, if we're going to present an implementation, we actually don't, where we work, we don't actually call it patient reported outcomes, we call it patient generated data, because we don't necessarily want the patient to report, right? Can we, can we do something that allows the reporting to be passive and not actually have to ask them to do another thing or fill up and fill out another survey or show up in another place? Right? So, you know, as an example, there's a, there's this, there's this technology called the bio button where, you know, we actually, it's just a, just a little patch that we put on in the, in the hospital, right? So in the hospital, it actually helps the nurses because the nurses don't have to actually check vitals every four hours, right? You actually, it's actually continuous. Now you've got to put AI on it and find the, the, the noise versus the signal and all that stuff. Right. But if we send them home with it, now we're getting continuous data on their vitals and the patient's not doing anything. Right. And we're getting more information. So increasingly, I think, got to get past the trust issue, right? So big brother, the whole thing. So that's still a thing, right? Don't, don't get me wrong, but how do we make it easier for us to collect those data and how do we make it easier for us to then work with those data in ways that don't make everyone's job harder? Those are great points. Tom, I want to do you and then David next. I think you make it a very good point there, you know, Edmondo. One of the things I think that we need to do is make things actionable, both at the clinician level, because we have significant burnout among clinicians who are overwhelmed with data and they have to read these reports and maybe 10% of it is something you need to do. So the data going to the physician or the clinician should be, you only have to deal with the things that are actionable. Now, I would argue, like you said, Edmondo, on the other side, the communication to the patient isn't, you know, someone's watching you 24-7, but you are being watched under the radar screen 24-7. And if something is actionable on your side, we will hear about it and we will communicate back to you. I don't think we communicate as well as we should. I think we tell patients, take this monitor and, you know, watch yourself. We don't tell them how it operationally should work. And if we put that in place, we would be much better off. And I think along the lines of what you said, Richard, with education, I mean, a patient comes in to see you, Jose, okay, rate control, thromboembolic risk reduction, rhythm reestablishment, you know, risk factor modification, my gosh, that is a huge amount of data in a very short period of time. If we can use digital electronic educational components to say, look, we are going to deal with this today, but I want you to also learn about that. And it has to be at the educational level that the patient wants. You have that Ph.D. scientist who wants to look and find out exactly what particular technology you are using and other people who are functioning at a third-grade level. And depending on their background, their family goals, their individual goals, we need to have education on the components, but also on those other things. We have mechanisms right now to check and see whether you are not just taking your medication, but you are being adherent over time. So if you are not, it shouldn't be just take your medicine. It should be educational things. Why? Are you having problems financially? Talk to your doctor. Did you not understand about the risk of bleeding versus thromboembolic risk reduction? We need to have educational things that upfront communicate the generalities and then drill down to the specifics relevant to each patient in their geography, at their educational level, and make sense. David? Yeah. So you said something, Armando, that really struck a chord. So when I started my practice, I was checking devices myself. And in time, I trained a whole staff to check devices. And now, my practice, we check probably 10,000 devices between remote and in the office. And I get AI reports that are garbage. And it's created more work. Whereas if I were to just do this myself, the way I did it, and scale up with human resources, I'd have less work. So AI is only as good as AI gets. And in time, it will get better. It needs the feedback. But it really has created a lot more work in my life. Paul, I want to go back to what you said in the opening statement. I think that the VA system is unbelievable. They were so early with the EHR. And there are so many advantages to their system. And yet, as I hear all the work that you're doing, it frustrates me that we can't generalize this to all the major academic centers. Look at what we could learn if we pulled our data and created these models, not just for patients taken care of in Mayo Clinic, but patients taken care of in Arkansas, and Texas, and the diversity that we could achieve. Is that never going to happen? Are the privacy issues, and the legal issues, and the potential monetary issues just insurmountable? Well, this is a philosophical question of, are you a conservative in the sense of, I want my independence and freedom to do things the way that I want it? Or are you a liberal in the sense that you want big government to regulate things? Want me to answer that? There isn't a right answer. I love the fact that we live in a country where we have a diversity of opinions and it's the ballot box that determines these things every few years. It's beautiful. Personally, I like working in a system where we have a system. Beyond the VA, we have a health non-system. And what's even worse, what are called electronic medical records are not electronic medical records. They're electronic charge capture systems that are designed to capture everything for the revenue cycle. And because they're selling to the CFOs and CEOs of hospital systems, they're not looking out for us, the frontline clinicians. So they don't work for workflows. What I'm hoping is that the efforts of the people on the panel here and the attendees here will find solutions that will actually work for frontline clinicians and ultimately drive us in the direction of being able to do that. How are we going to deal with those issues when the data is that half the patients in my EHR listed as having hypertension don't have hypertension? The medical records are wrong. Patients have a history of aflutter because someone misread their EKG 10 years ago that was afib. Are we ever going to be able to get through the human errors that are going to sort of corrode any future predictive models? These are good questions. And I can say that in our system, there are ways to correct those in real time in the electronic health record to ensure that you've got the right diagnoses and remove things that are incorrectly captured. And we need processes to address that on a broader scale. So we have about two minutes remaining. I'm going to just go down, starting with David. What excites you most about sort of digital health in terms of how you manage your patients, how you assess outcomes? I mean, what would you like to see to make this technology that's already out there more usable to you, right? Not to create more work, but less. So just sort of go down the path and figure out what's most exciting to you. What's going to make your life better and your patients' lives better? I mean, personally, I think that wearables will really be able to help us as clinicians if the patient is educated properly. It's not just, you know, your watch does all this stuff and you don't know how it does. I mean, I spend time teaching some of my patients how to use their Apple Watch. So I think that wearables are part of our future. And I think we need to embrace that. But we also have an issue with too much information. So we have to kind of balance both those things. And I look forward to that happening. Carmen, 10 seconds. This is a session on AI. I look at accuracy. I don't have less false positives, less garbage in and garbage out. And I also want it to be innovative and interactive. Jose? I like what Edmondo used the term, patient-generated data. I think that using digital pathways to bring patient-generated data to the clinician to then speak the patient's language is where we need to take this. I'm a big fan of generative AI. I actually think that generative AI democratizes AI in a lot of ways and actually starts to put those tools in the hands of folks who never had access to those kind of powerful tools. And I think it can really lead to some interesting outcomes. Richard? I think that's an excellent point. I would say access. My example about the 220,000 managed Medicaid population and being able to communicate directly with them on a daily basis. Paul? Technology to unify data management, everything from patient-generated data to the information we need for clinical care, coding and billing, clinical research, quality improvement, supply chain. If that can all be integrated and seamlessly incorporated into one process, it can reduce physician burnout, improve our documentation, make it help us deliver better care to patients. Well, the time is up. This is a fantastic conversation. I want to thank our outstanding panel and thank all of you for joining us. And hopefully, I'll see everyone next year at HRX. Thank you.

AI in healthcare

clinical settings

digital health tools

patient care

wearables

health disparities

real-time data

clinician burnout