false
Catalog
HRS-HIMSS Joint Session : How Digital Health Has T ...
HRS-HIMISS Joint Session
HRS-HIMISS Joint Session
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
Greetings. It is my pleasure to welcome you to this HRS-HIMSS digital health session titled How Digital Health Has Transformed the Delivery of Healthcare to Patients with Heart Rhythm Disorders. I am Sana Al-Khatib. I am an electrophysiologist at Duke University. It is my pleasure to introduce Dr. Alan Young, who is the chief medical officer at Giant and is a HIMSS National Physician Committee member. I'll turn it over to you, Alan. Thank you, Sana, for the kind introduction. Hello, everyone. My name is Alan Young, the chief medical officer with Giant. Just by way of introduction, I started my clinical training in orthopedics and continue to practice medicine here in the greater Los Angeles area, primarily in urgent care and telemedicine. Today, I want to talk to you about the role of AI or artificial intelligence in virtual cardiac care. First, I want to start off by just giving an introduction or some background about artificial intelligence and the history of it and as it pertains to medicine. For many, Alan Turing is considered the father of artificial intelligence, and he invented the Turing machine back in 1936. He also coined the term the Turing test, which is the ability for a human observer to look at a conversation between another human and AI and whether or not they can distinguish which individual is actually the artificial intelligence. Artificial intelligence over the years experienced a lot of slowdown in interest and development, and those were referred to as the AI winters. For many, the seminal event that demonstrated AI could be used in a number of different settings with when IBM's Deep Blue and Watson was able to defeat the Jeopardy champion at the game of Jeopardy. Several years later, Google released a product called DeepMind and AlphaGo, which competed against a world-class Go player. This was documented that the artificial intelligence from Google was able to beat this individual. One year later, the second version of that product, AlphaGo Master, actually beat the number one Go player in the world, who was from China. This artificial intelligence system was awarded a nine Dan ranking within the Go ranking. This demonstrated the ability of AI to actually outsmart the top human player in a very, very complex game. How does this apply to medicine? Well, the initial years of AI being used in medicine started off at Stanford, where a rule-based expert system was created that combined a number of rules based on symptoms and lab data to give information to physicians. However, this wasn't well adopted by clinicians because it took over 30 minutes for the system to give any meaningful feedback to the physicians. So that led to kind of a decrease in adoption of AI in medicine. Looking ahead, AI is now a little bit more complex and well-defined, and I'm just going to go briefly through the ABCDs of AI. A is for algorithms or the steps in a process to define a task. B is the big data component, which could refer to the amount of data, the variety of the data, the value or truthfulness of the data, and the variability of the data. So all of this information can be supplemented by AI to produce a number of calculations going through the algorithm. Cognitive computing refers to the use of machine learning, pattern recognition, and natural language processing to really mimic the human's brain capacity for taking information either visually or through language. And deep learning is a subset of machine learning that's based on deep neural networks, which allow for complex processing of speech, computer vision, object recognition, and detection, and is part of the technology used for autonomous vehicle driving that you might have seen or heard about. As AI has gone through its evolution, you'll see that automation is a form of AI. And as you move into what's called robotic process automation, where you have machines or computers doing some of this calculation. We talked a little bit about expert systems, where you have input from a number of experts and you create a number of rules. As you move further down the complexity of AI, you start to get into machine learning, which includes supervised and unsupervised learning, where a human programmer needs to control the algorithm or put the data in. And as we get further and further down the line, the goal is to actually have general artificial intelligence, which is the kind of the epitome of AI, where a system can actually gather data and learn and improve on itself without any direction from humans. So in healthcare, there's a number of opportunities leveraging AI. And I want to break it down into three specific areas, pre-visit, peri-visit, or post-virtual care. How this translates into the pre-visit world is you can have an AI system or a virtual assistant that can actually perform some of the currently manual processes or processes dependent on clinicians, where we can take a history from a patient using AI. And what this allows for is the recognition of language through natural language processing, and the ability to use logic to ask the right questions based on a patient's chief complaint or initial presenting symptom, and the intelligence to actually recognize very concerning or life-threatening situations and direct the patient out to seek appropriate care. I'll show some examples of how this is done in the COVID-19 era. But for different specialties, for example, if there are certain warning sign symptoms that the patient is able to share, such as severe chest pain, sudden shortness of breath, or any stroke symptoms, or palpitations, or heart irregularities, the artificial intelligence can triage the patient appropriately. The next step is really around when patients are going to enter into a clinical visit with the clinician. The AI can support this visit by taking a more thorough history and actually performing some of the documentation requirements that most clinicians need to observe as they see a patient via a telemedicine or virtual care encounter. So I think of this as a virtual scribe that can help me take a history from a patient, ask some of the standard or routinely asked questions, and then be able to document that on behalf of the clinician, either through an electronic health record or some other capacity. And finally, the other area where AI can be leveraged is after the patient visits a virtual care visit with the clinician, it's always great for them to get some follow-up. And due to the volume of patients that are now being funneled through telemedicine platforms, it's impossible for any system to maintain good contact with their patients. But if you have an AI virtual assistant that can reach out to a patient to inquire how they're doing, are their symptoms improving, have they picked up their medication, are they scheduling another visit, all of these things can be done in an automated way. So I want to share an example of an organization that has actually launched a virtual assistant. And while this does not apply directly to the field of cardiology or heart rhythm specialties, this is really an example of the impact a virtual assistant can have to help a health system as it manages an influx of concerned patient calls. So in March of 2020, at the height of the pandemic, Intermountain Healthcare launched a virtual symptom checker on their website. And you can see by the numbers in the first two and a half weeks, over 157,000 people used the system for an average of over 9,000 per day. On day three after it was launched, they saw a peak of 23,000 users online using this virtual assistant. And surprisingly, over 90% of people completed the entire screening process for COVID-19. And at the same time, they witnessed a remarkable 50% drop in the call center volume for inbound telephone calls to their nurse or a hotline or helpline. And at the same time, of the users that were using the symptom checker, 80% of them were symptomatic for something possibly related to COVID, and these were directed immediately to telemedicine. So this is just one example of how the use of a virtual assistant has really opened up a new channel for care and helped direct patients to the right follow-up as appropriate. The other place where AI can play a huge part is really around the clinical decision support side. So as busy clinicians manage and consult with patients, they oftentimes don't have all the information they need. As you can imagine, combing through an EMR for a patient's medical history can be tedious and time-consuming. You could have data in disparate places outside the EMR that you need to gather, and I've listed a few of the data points that are sometimes helpful for clinicians to make decisions about a patient's diagnosis or care plan or next steps. What AI can do is it can actually surface and search through all of these different data sources in a way that allows the clinician to get insights from this data, reminders about failed therapies or information about a patient's genomic testing or propensity for sensitivities for certain medications, and also looking at their trends in history for vital signs, labs, and other types of important information can help support a clinician as they make a decision about their patient. And finally, I just want to touch briefly on the idea of remote patient monitoring. Obviously in this day and age, people are more and more concerned about getting real-time information about their health. Apple Watch, Fitbit, and others have now included not just the heart rate monitor, but actually a single lead or multi-lead EKG device that can transmit this data to the cloud or even back to their healthcare provider, doctor, or specialist. And so a lot of things that can be done around monitoring a patient and understanding their heart health when they're not in the clinic or hospital has created another area of opportunity to leverage virtual care to improve the outcomes for patients. With that, I want to turn it back to the panel and I want to thank everyone for listening today and appreciate your time. Thank you very much. Alan, thank you so much for an excellent presentation. You gave us a lot to think about and to discuss. At this point, I'd like to ask my co-panelists to introduce themselves. Mike? Thank you, Sana. I'm Dr. Michael J. Miro. I'm a cardiac electrophysiologist in Indiana. I'm the chief academic research officer at Parkview Health System, but also a clinical professor of medicine at Indiana University. My area of interest in research has been messaging patients their implantable device data over many years, and also how do we integrate that digitally into the electronic record in messaging patients. Thank you. Of course. And Martha, do you want to introduce yourself? Thank you so much. Thank you for having me. My name is Martha Ferrara. I am a DMP and currently the assistant director of EP services at White Plains Hospital, New York. And my doctorate degree is in nursing science and patient advocacy. So let's actually delve into it. We heard about AI from Alan, and I'd like to start off by asking you, how have you implemented AI in your own practices, and what have your successes and failures been? Well, Sana, maybe I can start off. So I'm in Indiana, and I'm part of a large health system, Parkview Health System. We have an employed physician group, and with APPs and physicians, about 900 now. And the cardiology group, which is fully integrated, 29 cardiologists. And in that group, they also have 15 nurse practitioners working with them. And we've really looked more at data governance at the front end of implementing any AI, because of the AI solutions we've used in the hospital, mostly around prediction of sepsis, for example. We have found that the data governance piece was one of the biggest failings of AI. It's not a failing of AI, it's a failing of data at the front end. So the data quality at the front end of implementing any kind of machine learning tool is critical. And I think that without that, sometimes what appears to be a failing of AI is actually a failure of the health system or the clinicians from entering structured data at the front end. So we're big proponents of doing that. So we've struggled with getting all of our clinicians on board to enter systematically very clear structured data elements at the front end. And as you know from your experience at Duke, which is very much embedded in entering structured data, and particularly in the cath lab, in the EP lab, that the success of doing that is really getting all of the clinicians, nurses, physicians, anyone who touches data entry to be on the same page. So most of the tools that we've actually implemented are pretty simple around decision support. Nothing advanced, like Ellen was referring to, because we're still dealing with the data governance issues. That's great. Thank you. Martha? For me, and I want to thank HRS for having me be part of this panel, and thank you all for inviting me. I have a new service. White Plains Hospital did not have an electrophysiology service. So we, myself and my attending physician, Dr. Daniel Wong, basically have started this EP service from the ground up. And so what I tried to implement is just start the service with the most advanced technology that's out there for devices. I have some say in the setting up of the EP lab, but that was already set up. We use the BioSense system because, again, it's a small lab that is starting, and now we're here two years, and actually we're going to need a second lab. And we are in the process of mapping that out. My role is basically just making sure that when we set up our remote monitoring clinic, we set up our services here, that we are using those technologically advanced CIEDs. We have Bluetooth technology. We have a cardia in place. We have had also much of the work with setting up a successful remote monitoring program, and none of that can be done without using some of the tools that are out there. I agree with Dr. Miro, it's the human element sometimes that needs to be teaching a little bit. I don't think that we can do electrophysiology without using some of the tools we have. I'm waiting for the electronic medical record to also be part of that integrative process that needs to be done with using an electronic pacemaker record. That doesn't exist. So I think we have a lot of work to do, and AI is playing a big role for electrophysiologists, with no doubt. Great. Thank you so much. You both actually shared a lot of information with us. I do want to go back to some of the points that you brought up, but before I do so, Alan, may I actually ask you, because the vision that you shared with us, which was fascinating in terms of what happens pre-visit, during the visit, and after the visit, do you know of any programs that have been able to implement AI successfully throughout the care of a given patient, like pre, during, and after? Because it seems like many of us are doing bits and pieces of what you shared, but your vision was amazing in terms of having AI incorporated into every step. Do you know of any programs that are using that? Yeah, that's a great question, and I'd love to share some of what we've seen in the industry, and also with some of the partners that we're able to work with. And this is not specific for electrophysiology, but just general healthcare and medicine, in speaking about how AI is being deployed. I think one of the examples that comes to mind is the use of a triage or symptom checker to help really navigate patients to the right care. So if you can imagine if patients are starting to experience cardiac symptoms, for example, and they're not sure if it's an emergency, or if they should go to the urgent care, or if they should call their cardiologist, a lot of patients don't have the benefit of that access to healthcare information or advice right away. And sometimes they show up. They come to the ER, they come to a clinic. And what we found is that a digital tool like this can be deployed on a hospital's or health system website. And what they do is they interact with it and find out where they should go next. And if you can take a detailed history and apply evidence-based protocols or triage guidelines in an automated fashion, you remove the barrier of kind of constraint where a health system normally would have to have a team of nurses 24-7 in order to address every patient's needs. You now have an automated way to do so. And some organizations like OSF Healthcare in Illinois, Intermountain Healthcare in Utah, they've started to leverage this capability to give their patients access on the front end to guide us to where they need to go. The second part is really around supporting the clinician encounter. And this is where we've seen a lot of use of things such as scribes. There's actually some companies that deploy a voice recognition scribe in the room. Our company likes to deploy a kind of a text-based virtual assistant to take a history for a patient. And by doing so, we can actually incorporate this data and surface it to the physician to allow them more freedom to do a physical exam, ask more detailed or complex questions without having to repeat kind of the basics. And I think that really will help organizations to alleviate some of the burdens on their clinicians from a documentation or clinical record-keeping perspective and allow them to focus more on their patient. And so some organizations are rolling this out as part of an experience for, for example, if you're in the waiting room of an urgent care clinic or in the emergency department or even your cardiologist's office, this can be deployed while they're waiting to see the clinician. And the third example, as you deploy it, I think one of the organizations that is doing this extremely well is the Cleveland Clinic. They actually use an automated process to follow up with all of their discharge patients from a number of hospitals and ensure that the patients are doing well. They do this in combination with their nurse call line. But as many of us realize, people do not like to pick up their cell phones anymore when they get an unidentified caller on their phone for fear of a telemarketer or a scam. But it means a lot more if you get a text message from your health system saying, hey, we have a few more questions, can you interact with us via your phone? And you ask the same questions that a nurse would, document the information, send it back to your doctor's office or the health system, and then they can initiate an appropriate workflow or escalation if there's anything concerning. So those are just some examples we've seen of people experimenting with AI to really help support care. That's great to hear, Alan. Thank you. Without a doubt, the pandemic has had a major impact on the way we deliver care to our patients. And I'd love to hear from Martha and Mike about how the pandemic has affected your workflow, perhaps maybe starting with telehealth. How have you implemented telehealth in the pandemic? Martha, can I start off? Absolutely. In our health system, just five years ago, we started a robust effort at transitioning to what we would term virtual health, which is a combination not only of telehealth, but the technology surrounding that and the health IOT space, integrating devices into that system. The clinicians' uptake was very slow. Cardiology actually was the most advanced, doing teleconsults to E.R.s because we're in an area where we have servicing a lot of rural hospitals. Rather than transporting patients 60 miles to our facility, to do a teleconsult. And that has been very successful. So in January of this year, the total number of cases that were virtual visits for ambulatory visits was only about 650 per month. And that's a lot of patients. So we've been able to do that. We've been able to do that. And the total number of ambulatory visits was only about 650 per month throughout the health system. This is primary care, cardiology, everything. By April, it was 25,000 per month. So an absolute explosion of telehealth services. So an absolute explosion of telehealth services. In cardiac electrophysiology specific, our remote monitoring clinic, we monitor about 4,500 patients with CIDs. And many of those patients, of course, as you know, like to come in because they want to interact with a nurse or someone. But that now has transitioned to greater than 97% are done remotely. And we've actually supported the ones that need to have some human interaction with telehealth monitoring. So the only time a patient would come in is if they needed their device reprogrammed or something, there's some issue with a lead, issue like that. So we've seen a huge shift and I don't think it's gonna go backwards unless the CMS rules on reimbursement for telehealth services goes backwards. And I think that's a very ill-advised given the CMS cost curve. So anyway, that's been our hand-to-hand combat experience here. And by the way, I really wanna make a comment for Alan. He touched on some really important things. So one of our colleagues, Eric Topol, I'm sure read his book, Deep Medicine. So a lot of the AI solutions will actually free up clinicians to spend more time asking very pointed questions to their patients and spend more time with their patients. And so the clinicians shouldn't be concerned about AI. It actually will allow them to actually practice medicine and be more rewarded, have a rewarding relationship with their patients. So I think AI will actually be a revolution in healthcare with regards to the joy of medicine. I like to echo that sentiment, Michael. I'm an avid listener of Dr. Topol's podcast and we have had basically, I would say the same experience with implementing telemedicine at White Plains. Initially, this was a little project that was started September, 2019. And prior to this conference amongst us, I had gone downstairs to ask the team to just kind of give me a little background how we were doing with the telemedicine. Basically, there were maybe 20 to 30 visits a month starting back in 2019 in September. And maybe it was a total of 200, 300 visits total from 2019. As we stand now statistics from this week, we are up to 20,000 visits, which translates to 15 to 2,000 visits per week. And where it impacted our clinic, basically my entire clinic was deployed for COVID-19 coverage to the hospital. So it left two team members in the clinic. And even if we did not want to be efficient, we had to learn very, very quickly how to learn how to do telemedicine. And not only that, but also how to teach our patients how to do telemedicine. Because they're used to traveling to the office, that's easy. Everybody gets in their car, everybody knows where to park, everybody knows where the office is. This was a lot of learning with, you need to click on that site that we will send you from WPH Connect, which is our service now. And I say that for the past time of the pandemic, which we're still implementing protocols for, we have seen in the office perhaps a handful of patients, everything has been done through telemedicine. And I do agree with your comment earlier that in a way it makes that session a little more intimate because you have this block of time with the patient and they are in their element, they're in a comfortable setting. So now that they're in a comfortable setting at the very beginning, it felt like I'm the one that is in their house. So that was a very different sort of feeling for us. Mostly for myself with my remote monitoring team, the protocol that I implemented so that I would do two things. I would check the remote monitoring working. If my patients forgot how to send transmissions, my team would teach them on the spot, 24 hours earlier to the telemedicine visit, they had to send a transmission so that they would relate that transmission to the visit that we're doing the next day. And I think it worked very nicely so far. We have had everyone either learn how to do a transmission, remember how to do one. And if the monitor wasn't working, we basically just did a phone visit and we order a new monitor. For me, my experience has been in our remote monitoring team, it just made the work more cohesive and more meaningful because I outsource our remote monitoring team. They work out of my office in another location. They don't get to see the patients when they come into the office. But now they had to basically build a relationship with that patient in order to just get that transmission prior to the medicine. So our patients have had, I think that's, and I have asked my patients, have had the positive interaction that this little monitor on that night table really does have people behind that monitor. We are the little elves working behind that to keep this going. So I think it was a positive, if it can be such a thing, a positive event that happened for us. And patients are happy not to be outside their homes. That's my findings for our team. Well, thank you. That's actually a very insightful perspective and thank you for sharing your experiences. I believe that remote monitoring is such a great way to utilize the technologies to keep an eye on these patients. And at least for us, I can tell you that the volume of remote monitoring, even though I would say maybe close to 90% of our patients are already using, or were already using remote monitoring before the pandemic, but still our remote monitoring volume has really skyrocketed. I suspect that that has been your experiences as well. Can you just talk a little bit about how your remote monitoring volume has been? So- Yes, go ahead, Martha. So I'm an advocate of remote monitoring. I truly believe that the value in it is knowing that data from the old way we used to do this. I knew nothing about this patient's device for three months because that was the old way of doing things, which I cannot even think about how that used to work. You would lose sight of that device. So now that we are connected, it was meaningful to me in that this is how we follow our patients and we can always act on things that come through that remote monitoring before that patient shows up. So that was always something that came natural to me as I do it more and more. My experience was that it became sort of meaningless to the patient because they lost that every three months I must get in the car and go. But this pandemic has brought it full frontal that this is not just a toy to keep here because it's just a pretty little box to have. And what they have done for our practice is they have been incredibly understanding and kind in that now we realize, and this is the feedback they give me, now we realize that what you always talk about, connectivity, connectivity, keep it plugged in, don't keep anything in front of the monitor, really has a meaning because the data that you get, I share that data with the telemedicine with my patients. Like I press the button to share the screen and I show my patients, this is what you sent yesterday and the data stamped on there. And it really has become an act that we do that they didn't realize that we actually are doing. So for me, I've always had a very high connectivity rate for my patients. I aim for, and I know this sounds funny, I aim for 98% because my connectivity right now is 92% because I truly believe that what we do is work that is meaningful for our patients and the impact that it has on them. They're elderly, they're scared, they don't know what it all means. So now it's a little easy to explain, this is what we do. And I don't get to see what you're doing over the remote monitoring, doesn't have that technology, but I do get to see the health of your device. And it's just has been easier for education purposes, just much easier. And I'm happy to report that my connectivity rate is 95.6%. Well, Sana, we've had similar experience that Martha's articulated, pretty much patients now realize the benefits of connectivity. We've actually been very active with messaging patients or device data through their personal health record. Our enterprise EMR is epic. And so we've spent a lot of time integrating customized reports on a particularly complex devices such as CRT devices, as well as ICDs in messaging patients. And so the heart failure patients particularly, we would actually ask them to download at least monthly so we could actually see what their sensor data was looking like and then giving them feedback, particularly on the percent CRT pacing. So patients start to understand precisely the importance of CRT pacing and the percentage. And so it's done a world of difference in engaging patients in heart failure care, particularly with the complexities of CRT device for patients to understand the importance of why their device needs to be pacing effectively for them to benefit. And that's been transformative and we're going to scale that as much as we can. We've studied different cohorts of patients and what they want to see specifically. That's great, Mike. I do have a couple of questions on which I'd like you, Alan, to comment. But before we do so, again, in terms of what's relevant to EP, obviously not every patient we take care of has a device and we have an increasing number of our patients who are using consumer products, watches, monitors, things like that. A lot of them actually transmit these data to us, upload them into their chart where we can look at the data. We were doing that before the pandemic, but based on your experiences, have you seen more of those come through now with the pandemic and have you been able to establish workflows in your own practices to be able to attend to those requests? So I'm just going to say from my standpoint, we have encouraged our patients to sign on to the LifeCore Cardia app. Some of them have purchased, but they never signed up to the website. So one of the upsides for us was that when they are calling the office and we are blinded to it, we just guided them to how they can just join the website and we have a robust, maybe not as robust as I want it to be. It should be more. But again, it's all education in this community for patients. And so I have seen an increase in what I did see from a particular few patients was the anxiety that this provoked. They would send a lot of those cardia, a lot of those rhythms because they thought they were in a fib, they didn't, their heart rates are going fast. And when we made those follow through phone calls and basically it's just the anxiety of what's happening. And the other services that I was able to just immediately green light was the Zio patch or the Brady DX patches. We also put in place their home monitors. Patients wouldn't have to come to the office. You know that they can, those companies would send the patches to their homes and the patients would place them on because the companies are walking them through it. And they were able to just send it back and we were able to get those reports. I haven't implemented any of the, what Dr. Miro was, excuse me, talking about uploading some of those reports into the EMR because the challenge for me in this community has been to give them access to the MyChart in Epic. That has mushroomed. Now they understand why they have to give me their email address, which they were reluctant to do. And so the, it was very little education to say, well, you know, in order to do what you want me to do, which is provide your medical information, if you could kindly just provide us with your email address. There was no question. We have, I would say just exploded in the way that patients are now signing on to the MyChart through that Epic EMR. So I think that that's a plus. Yeah, our experience has been similar to Mark. I think patients historically, PHR adoption had been slow because they really didn't receive meaningful data or, you know, temporally appropriate data. So they had no reason to log in. So they would forget their login information and they would find it to be very cumbersome. But once they are getting routinely getting valuable information through their personal health record, they start logging in all the time. And actually we've actually published a lot of data around this. And actually the biggest adopters and users are patients that are over the age of 70, interesting enough. They just need a little education, but once they get it, they're very engaged with looking at their data. And that's been important. As far as wearables, it's been a challenge since they're all over the place. With the Apple Watch, for example, or some of the other devices that detect potentially AF, for example, we would typically ask the patient to, you know, give us their data either electronically, but we do not incorporate that in the EMR. We actually then order a device, a wearable device that they could then use to see if they do have AF. And our experience has been similar to what Martha. We would partner with a company like iRhythm, which has a Zio patch, which could remotely send the patient their device. They could place it on their chest and they walk them through it and then they mail it back. So the patient never really comes to the office. That's a perfect solution. It's completely done virtually. So if a patient calls in because their Apple Watch indicates that they have some heart rhythm disturbance, we would then order a commercial device that would be able to actually detect it. One of the issues with some of the recordings, by the way, the Cardio Mobile being a great example, PDF files. So the more PDF, how many PDF files are going to put in your EMR, right? Is there's already a blizzard of data in there anyway. So you have to curate the data in your EMR. So that will be a challenge going forward. I think the wearable platforms will have to figure out some way, some standardized workflow that would allow us to curate the data so that we can really only put the meaningful data for that patient in our electronic health record. Otherwise, it's just so much. One potential solution, of course, is what Alan touched on is machine learning and AI. When you actually want as much data as you can get, right? But the AI tool will shrink that down to a digestible form for the clinicians. Thank you both. I think we actually only have two to three more minutes left and in those remaining minutes, it would be great if we can have some closing remarks from each of you about, the pandemic has been pretty bad for all of us, but at least the silver lining is all the progress that we've made in telehealth and using AI and what have you. Going forward, how do we build on that momentum to effect more change? If each of you can actually just chime in with a few thoughts, that would be great. Starting with you, Alan. Great, yeah. It's been great to hear all these experts talk about the field. I think the biggest opportunity I see is that as we gather all of this data, albeit from clinical symptom intake or remote patient monitoring or EKG or other diagnostic tools, the real value in the AI is using that data to make a prediction about the patient's outcome or their risk for developing a complication or going down into a clinical scenario that's not readily addressable. I think if we can use the data in a meaningful way to make these predictions, it can enable our health systems and our physicians to anticipate which of their patients might need earlier intervention and prevent kind of a catastrophic event, especially in this field. And because so many patients have delayed care or put off going in to see their doctors or even getting an in-person visit, there's still gonna be some gaps in information, but hopefully we can use the data that we've gathered through platforms like telemedicine and remote patient monitoring to make meaningful predictions to enable our health system to respond to the needs of these patients. Great, thank you very much. How about you, Mike? Yeah, I think the silver lining, if there is one, in a pandemic is it's been transformational for medicine into a digital health era. There's no question that with the explosion of telehealth visits, now we're surrounding patients with telehealth solutions and digital health solutions that could integrate into the system. That will enable these more advanced AI tools to digest that data and as Alan articulated, then start predicting events. So I think that we'll see a new era of medicine. We know the amount of VC money going into digital health solutions has exploded. I think last year there was $9 billion in VC going into digital health companies. And that will now accelerate even faster with the transformation that's been catalyzed actually by the terrible pandemic. But it will be best for patients in the long run. They'll be more connected to their clinicians and their health systems. And it will have patients who did not have access to healthcare gain access to healthcare, particularly with rural broadband. Great, thank you, Mike. And Martha? So I just wanted to say that I would hope that our present manufacturers, you know, device manufacturing companies, at least give a little read to the paper by Dr. Ursula Weiner. I think that you were one of the coauthors of that paper, the transparent sharing of digital health data, a call to action. I reread that paper in anticipation for this meeting because I think it's just simple. AI should be utilized the way that it was intended to, but it needs to start from the patient's perspective. Because that is who we're trying to serve here. And for me, AI, and I'm looking at it, you know, from the nurse practitioner that stands at the bedside with the patient every day, and if not the bedside, at least by handholding these patients when they come to clinic. You know, AI should be precise. I think companies can do that. It should be proactive. Read that paper. I think that all those CEOs would be served well to read that paper. The simplicity of what we need is just, you know, outstanding, nothing better than to start there. And I think it should be used in preventative manner from all that all the other experts have shared. I think that we can harness the power of AI because it will help us serve all of our patients and just minimize that healthcare disparity that exists that it was one of the highlights from this pandemic. You know, I work in the Bronx and this community was so hard hit and the disparity is so apparent. So if AI can be utilized in any way at all from where we're coming from, this is it. We are here for our patients. Well, thank you. And thank you all very much for your excellent and very insightful comments. And I agree, I am as excited as all of you are and really hoping for the best. So thank you. Thank you. Thank you, everyone.
Video Summary
The panel discussion titled "How Digital Health Has Transformed the Delivery of Healthcare to Patients with Heart Rhythm Disorders" focused on the role of AI and telehealth in improving patient care. The speakers highlighted the impact of AI in pre-visit, peri-visit, and post-virtual care settings. In the pre-visit stage, AI can assist in taking patient histories and triaging patients based on their symptoms, providing timely and appropriate care. During the visit, AI can act as a virtual scribe, helping clinicians with documentation and allowing them to focus on patient interaction. After the visit, AI can be used to follow up with patients, checking on their symptoms, medication adherence, and scheduling future visits. The speakers also discussed the increased adoption of telehealth during the pandemic, with remote monitoring and virtual visits becoming more common. Patients have benefited from the convenience and efficiency of telehealth, and providers have found ways to integrate wearable devices and patient-generated data into their workflows. The speakers emphasized the need for data governance and education to ensure the successful implementation of AI and telehealth in healthcare settings. Moving forward, they believe that AI will play a crucial role in enabling predictive analytics and improving patient outcomes, ultimately leading to a transformation in healthcare delivery.
Keywords
Digital Health
AI
Telehealth
Patient Care
Remote Monitoring
Wearable Devices
Data Governance
Predictive Analytics
Heart Rhythm Society
1325 G Street NW, Suite 500
Washington, DC 20005
P: 202-464-3400 F: 202-464-3401
E: questions@heartrhythm365.org
© Heart Rhythm Society
Privacy Policy
|
Cookie Declaration
|
Linking Policy
|
Patient Education Disclaimer
|
State Nonprofit Disclosures
|
FAQ
×
Please select your language
1
English