Well, I want to welcome everyone to the session at HRX. This is a section focused on a pretty significant clinical problem, the management of cryptogenic stroke in contemporary practice, and we're very lucky to have an esteemed group of panelists today, and I think we're going to have a very invigorating discussion from now until 1115. My name is John Pacini. I'm a cardiac electrophysiologist at Duke University and have done some research in the cryptogenic stroke space, but I'm very excited about this session and look forward to learning a lot from our panelists, and we're going to have each of them introduce themselves in turn. Ravi, do you want to go ahead and start? Yeah, thank you, John, for moderating. Pleasure to be here with the panel here. My name is Ravi Shah. I'm a cardiac electrophysiologist at Atrium Health down in Charlotte. My background is in digital health innovation health policy, and I started the, I guess, AI research core at Penn State and hoping to bring some of that stuff down to Charlotte. I'm Brett Meyer. Thank you, John, for having me. I am not a cardiac electrophysiologist. I'm a stroke neurologist. I'm the co-director of the Stroke Center at UCSD. I'm the clinical director for the Enterprise for Telehealth, and I'm well-researched in the space of technology and stroke and now moving into the electrophysiology and cardiology application world. Great. Welcome, Brett. Lynn? Good morning. My name is Laura Flynn. I'm a nurse practitioner at Valley Hospital. I've been there for a while, and I'm excited to join this roundtable. Yes, and I, Laura, you've wanted to be a nurse since the third grade, and now you're an MP, so I'm really glad that we have the nursing perspective captured because I think a lot of the key challenges for the patient journey, that perspective is, you know, can't be overstated, so thanks for joining us. Jonathan? Hey, John. Thanks for having me. Jonathan Su from University of California, San Diego. I'm a cardiac electrophysiologist working with Brett over in San Diego, and a pleasure to be here. Really looking forward to talking about cryptogenic stroke. You know, I think there's so many things we could talk about, but I'd like to, like, wade into the deep water immediately and kind of dispense with the easy questions. So I think I'm interested in getting an appreciation from all of you what you guys feel are the key barriers. I certainly know what they are at my center, but I'm curious to get the perspective from around the table and from each of our different perspectives and our background about what the key challenges and barriers are to managing these patients. I guess I'll kick it off. So the first barrier is recognizing that you actually have a barrier, and so in our experience, that really was the first and the biggest hurdle, is that if you would have asked me a year ago, is there a barrier to the care of, in my case, stroke patients who have concomitant electrophysiology or arrhythmia issues, I would say no, everything's fine. We see our patients, we see our very, very small cohort of patients who have a stroke due to an underlying arrhythmia, and we manage them just fine. And then, you know, in reality, obviously that's not true. We have a very large slice of patients who have an underlying arrhythmia problem, and we do not manage them well. So I think that the first barrier is recognizing that you have a problem. Once you recognize you have a problem, then the question is, you know, how are you going to solve it? And that goes to, you know, everything from lean principles and change management. So we can talk more about that as we go, but recognize you have a problem is number one. That's great. Ravi. So I'll just make one more point to it. I think of cryptogenic stroke, we think of it as an inpatient problem, but surprisingly, you know, there are a fair number of patients that we see in clinic who come in with TIAs and having had strokes at different facilities, and they've never been screened for atrial fibrillation. They had a cardiac monitor on for two weeks, 30 days, but again, that's not the standard of care. So getting those patients in, getting them loop recorders, again, if it's justified, is a significant challenge, and the reimbursement for that is a challenge as well. I think also talking about barriers and the cryptogenic stroke population, we have to think about, like, how do they even get to us as cardiac electrophysiologists? We see ourselves as the end referral aspect, but we see a sliver of the population, right? I mean, there's bias in those individuals that get to us as a referral, and so to Brett's point in regards to barriers and realizing that we're not seeing a large population of these individuals just because they're not getting to us. So understanding that there is this issue, I think, is really important. That's what we found at UC San Diego, is realizing that just the numbers that Brett is publishing in regards to understanding that this population exists and wasn't having the right care taken care of in regards to external monitors and loop recorders is in and of itself emblematic of that. And Laura, you probably have the best boots on the ground perspective out of all of us, given the volume of your practice. What are the key barriers from your perspective? So the biggest stuff for us is the patient care, and when a patient comes in with a stroke, they meet a ton of people, they're scared, and then we get called at different points during the hospitalization. We can get called at the beginning where the person's just getting that CAT scan or the MRI to diagnose them with a stroke, and then we also have to tell them that maybe you have this, but we're not sure, so we need to do something. And trying to convince them that they need to have something done is pretty hard sometimes. You know, loop recorders, patients don't like things implanted in them if they don't think they need them. So there's always that patient care perspective that's a big barrier for us. Yes, so you bring up, I think there have been a couple of commonalities, right? There's a lot of heterogeneity in how these patients are managed, and you just brought up implantable cardiac monitors. I was over at ESC in London and I was just skimming through the new AFib guidelines from the ESC, and their recommendations for cryptogenic stroke are really interesting. So there's a class one recommendation for prolonged arrhythmia monitoring, but there's really no specificity to that. And then the other one that kind of piqued my interest is there's a class three of upfront anticoagulation for those individuals as, you know, completely contraindicated, yet often we'll see someone with severe left atrial enlargement and runs a PAT, and you know, I think that patient's extremely likely, if not guaranteed, to have atrial fibrillation. I'm curious to get your guys' perspective on what prolonged arrhythmia monitoring means, and what's optimal, and is it the same for every patient? So I'd like to answer that, kind of dovetailing to a comment that you just made about invasive monitoring. So in our analysis, CONNECT, which really was our foray into assessing what getting further monitoring, we actually ended up leaving it more to the patient as to whether they wanted, this goes to your point about people don't want invasive monitoring, we actually left it up to the patient. I, very paternalistically early on in my career, said, well, you'll get an event monitor and then eventually we'll do an implantable loop recorder. And then with our CONNECT program, which allows us to do both event monitors or implantable loop recorders early in the hospitalization, we actually started to have more of what we call a vis-organic or CONNECT organic conversation with the patient, saying, you know, I can do both, or you can have both, and let me tell you about event monitors, let me tell you about ILRs, and interestingly, in our first analysis, over 60-62% of patients, not 10%, 60-plus percent, decided that they wanted that implantable loop recorder early on, like prior to hospitalization dispo. Since that time, since we published that, that number is going up, and so patients are, you know, I'm old enough that I remember the commercial of set it and forget it, that toaster or whatever it was, patients just want to set it and forget it, and I thought that they would want a non-implantable, and more times than not, 70% these days, they're picking the implantable. So what is prolonged monitoring? Well, the first thing is, what will patients do? Okay, the second thing is, what is prolonged monitoring? With Stroke AF and Crystal AF, my answer to you is, as long as humanly possible, get me that data. I want that 10% at one year, and 20%, 30%, I want that three years worth of data. I'll take what I can get, if their insurance will only cover 14-day monitor, that's what we'll do. I want as long as I can possibly get, because we're finding in the stroke world that it's not that patients aren't having the strokes due to AFib, it's that you just didn't look, or didn't look long enough. It's an air fryer now, and not a toaster, Brett, just just so you know. But in the same vein, so, I mean, the beauty of this shared decision-making that you're talking about, basically empowering the patient, is that there are data to support whatever they decide, right? Embrace, Crystal AF, all published in the same issue of New England Journal, supports exactly that. And so, having a shared decision-making is all the rage right now, right? But the model that you all are employing in the neuro-stroke world, I think, if it can be codified in either a consensus document in some way, would really have some balance, and I think give credence to some type of monitoring in this regard. So, I don't know, that's just kind of my feeling on, it's beautiful when there are data to support what we're talking about, right? Well, hopefully that data will be published today. I'm waiting to hear if it's published today, so let's see. So, I kind of want to take this in a slightly different direction, since we're quoting trials, right? And when we look at trials, like Navigate and Respect, right? Across the board, empiric anticoagulation was no different than aspirin. Great. But then you have other trials that try to look at, you know, were there factors that were thought to predispose people to having stroke, right? So you have Atticus, you have Arcadia, right? Both were stopped for futility. But then when you look a little deeper, you have to think about what characteristics they identified. So, you know, P wave terminal force, left atrial diameter, I mean, these are things that we see and we can codify into the schema, but I think what we should be looking at is left atrial cardiomyopathy, right? Mechanical dysfunction. Our group has shown that, and we've been able to show that, you know, left atrial mechanical dysfunction increases your risk for recurrent stroke after cryptogenic stroke. Having atrial fibrillation in itself is a risk factor, but there are people who we find don't have atrial fibrillation and still will have recurrent stroke. And you have to look at the mechanics. And if we do these trials again, and one of the abstracts that I'm presenting today has to do with an AI algorithm to identify patients who will have atrial fibrillation. But if you go back into the NAVIGATE trial, right? If you go back into RESPECT and you pull their echo data, you pull their ECGs, can you run it through an algorithm and then do a subgroup analysis and say, hey, these patients actually did benefit? And I would invite anybody to do that study with us. Yeah, I mean, I think it's a great point. Not that there's a good thing about having a stroke, but I think one of the things that makes management easier for us physicians and the medical community is that these are individuals who had a stroke. So I agree with you. I don't think it's what was the cause of the stroke, although that it's an important question. They've had a stroke, so they're a secondary prevention patient. So the other way to look at it is just identifying which persons who have suffered a stroke have atrial fibrillation as a secondary prevention target. Because even if that wasn't the mechanism of their first stroke, we certainly won't want them to have another stroke. And I think that's what I took away from Stroke AF and CRYSTAL. And I think many people have pointed out, well, all of these trials are identifying a certain percentage of atrial fibrillation. How do we know it just isn't background noise or the epidemiological load of the arrhythmia in the population? And, you know, my kind of answer to that is, well, does it really matter? You know, we're trying to optimize everyone's health regardless of what was the specific cause of that original stroke. John, you bring up Stroke AF. I struggle with that study a lot, right? I mean, speaking of background noise and the numbers at six months or one year, whatever you choose in regards to the outcome for ILRs, but I don't know what the Brett Myers, the stroke neurologist, feel about that study. But to me, do we just put it in everyone? And is it just that AFib is so predominant that if you monitor longer, you find it? And how do stroke neurologists feel about that study specifically in small and large vessel disease? Honestly, that's the biggest question. What does WNL stand for within normal limits of writing your note? It stands for we never looked. So WNL, we never looked. So when it comes to arrhythmias, we need to look more. The stroke neurologist read Crystal AF and said, yeah, cool, great. But then we read Stroke AF and we said, oh my god. So there is so much AFib that we are just covering our ears or our eyes or our mouth and we're just not looking for. The under-assessment is issue number one. Issue number two is build the program that allows the communication, which we're going to talk about, that builds the communication between the strokeologist and the electrophysiologist to improve that pathway to get those people monitored more. Because if there's an inertia in the system or if there's a boulder up the hill moment, you're not getting those patients monitored even if you wanted to. So we are under looking and we need to look more. And correct me if I'm wrong, but then that paired with Navigate and Respect for ESUS specifically, I would imagine for a stroke neurologist, so an electrophysiologist looked at the data and were like, okay, sure, you don't anticoagulate those individuals, but I'm sure that that difference between detecting something and changing management is probably really important for a stroke neurologist. Agreed. In short, I agree with John that if you find the AFib, whether or not they have other cryptogenic issues, you may have an AFib patient and you just didn't find them yet. And then one-to-one, you find AFib, it's going to change your management in stroke, period. You're going to anticoagulate that patient. Doesn't really matter what else you find, you're anticoagulating. So go find it. So I definitely agree, but the the loop study might suggest otherwise, right? I mean, it's atrial fibrillation increases your risk of stroke, we know that. But there are a lot of modifying factors and so across the board, is anticoagulation the right answer? Yes, I think eventually it is. I think everybody has those risk factors that are going to put them at higher risk, but I think maybe we need to look a little more deeply. Yes, so Laura, I'm curious, you discharge all these patients and marshal their care. So we're all saying, yes, if there's AFib, we should treat it. But how much, how much AFib in your guys' practice in a patient who's had a stroke do you require before you then go ahead and treat them? I can only speak for like the hospital side, because I don't have a lot of outpatient experience. But you know, there are people who do get loop recorders who stay in the hospital for like a week or two, and we found some AFib, now what? Usually from our neurologists, they don't want us to start AC for a certain amount of days, depending on the type of stroke and size of the stroke. But eventually that person will get put on it, put on some kind of blood thinner at some point. But again, it's all patient individuality and what they, you know, what type of stroke and where and all that stuff. I disagree. I think any AFib equals anticoagulation, period. I think that you're right about the point about inpatient assessment, you know, is it going to be a week before I put them on anticoagulation, but any AFib equals anticoagulation in my book. Especially when you add in things like good old-fashioned stroke data like WARDs, which were for an aspirin recurrent stroke study, comparing anticoagulation to antiplatelets. Anticoagulation is a good choice. You find a little AFib, you want to quibble about one minute, three minutes, or whatever. When I have AFib, for my wife, who will be the person who finds my AFib and sees my TIA in stroke, please put me on a DOAC as soon as possible. And even if it's 30 seconds of AFib. Hopefully it doesn't end up with a stroke, but I know Brett's cutoff is two minutes because some loop recorders, that's the cutoff, so I know it's not six minutes based on assert or other type of data, but what I struggle with, with the cutoff question, John, is left atrial appendage occlusion, right? So a lot of these individuals, whether they have an MRI with susceptibility and they don't wanna, they worry about, stroke neurologists worry about bleeding into the stroke, et cetera, but what is that cutoff for doing an invasive procedure, which we feel as electrophysiologists or cardiologists, that's gonna benefit the patient to get them off anticoagulation long-term, but doing an invasive procedure, and so I battle with this in clinic where we're seeing these patients that have two hours on a loop recorder because they had cryptogenic stroke, and then I'm talking about an invasive procedure where the patient's not gonna feel any better, and to me, that's where that cutoff, that question of time really burdens me in regards to what is that cutoff to do an invasive procedure. Yeah, I think left atrial appendage certainly is interesting. For me, I just, I don't wanna talk out of both sides of my mouth, but my overall view in these patients is the mechanism of the stroke isn't necessarily the end all and be all of their management because if they've had a stroke, we wanna prevent all forms of future stroke, but with left atrial appendage occlusion, we are assuming a certain mechanism and likelihood of where the stroke is coming from, and so I think that's, I agree with you. I struggle with that a lot in our patients who may not be an optimal candidate for oral anticoagulants with cryptogenic stroke, but Ravi, just to step back a little bit, what do you think about the time question? How much is enough in a secondary stroke patient, a secondary prevention patient to pull the trigger? So, I mean, a lot of people have asked this question. That's where Artegia came from, right? And I think there is a distinct benefit to anticoagulating even for very short times, and again, it's been a while since I reviewed that trial, but the findings showed that there is a net benefit to anticoagulating even for very short amounts of time for a atrial fibrillation, but the way they phrased it wasn't as conclusive, and so I'll need to honestly go back to the study and look at the results again. Yeah, well, I think what we're all saying is it's short, it's minutes, it's not hours, and I think hopefully we'll refine that over time. Let me ask, Brett, please. Sorry, I just want to change the direction at the risk of offending, and if it's on your list to get to, I apologize. You know, to me, John Sue mentions the issue of left atrial appendage closure, right? So LAA is a big issue, and we're tackling it with our own processes, but from my chair, you're not even seeing these patients yet. These patients are coming through, maybe they're coming to cardiology for whatever reason, but many of them are coming through the stroke world, and we're not even giving them to you. We're not even sending them to you, so there's a whole pile of patients that need an LAA closure that John Sue doesn't even get to because they come in my clinic and either, you know, I don't recognize, hopefully I do, or the process for me to get more patients into your clinic is kludgy, to say the least, and I'd love to talk to processes, but so from LAA or cryptogenic stroke, we've solved that problem. We have improved solutions to get the patients through the system. We can debate all day long about, you know, who should be closed and how much AFib, and we can spend hours talking about that kind of stuff. I don't think we'll come to an answer today about those kind of things, but we could all agree on one thing. There are more patients out there that need to be assessed and need to be shunted into the system, and then there's a lot of data that goes to deciding what that answer is. I just wanted to get that out there early. No, I, but to John's point, does that LAA closure that then at six months, whatever your cutoff is, becomes aspirin 81 milligrams, does that patient vex you in the stroke world in regards to that being the only therapy that they get lifelong? Because I wonder about that patient specifically and what a stroke neurologist feels about that, a closed appendage with aspirin 81 as background therapy for arterial type. Yeah, well, this goes to the age old question, which is why I think strokeologists are important, which is, you know, please make sure that you are comfortable with what the cause is and the treatment is, because whatever choice you make for a treatment is a choice against the other thing, and so that's why, John, we talk quite a bit about let's do that ILR a little longer before maybe our interventional cardiologists close a PFO or before we do a LAA closure, just because the act of treating is preclusion of something else. But yes, we're all frustrated because your hands get tied at one point. So I wanna go back to the access discussion because I think it's really important. I think we all come from a place of privilege as far as how we practice medicine, right? We have tertiary stroke centers, right? You have all these internal referral pathways. I was talking to a lot of folks yesterday, and what you realize is that there is a huge number of patients that exist outside of academic medical centers, who are at rural hospitals, who only see an outreach clinics once a year, maybe once every five years, right? How do you get to those patients? How do you identify those patients who might have atrial fibrillation or some other cause for having a stroke that we're missing? And I think there's a really good discussion to be had about the use of artificial intelligence. Can you do a 12-lead ECG remotely, run an algorithm on top of it and say, hey, you have a 90% chance of having atrial fibrillation, right? Maybe we can't get you a loop recorder, but I can start a DOAC remotely, right? And having that shared decision-making with the patient, I really, really wanna hammer home that we are not seeing so many, so many patients. They're out there. And in clinical practice, you just see how many patients have kind of been neglected, and they're all new referrals. But you ask yourself, where are they coming from? Well, they've always been there, right? We just don't have kind of a vision into their world. So Ravi, you brought up a bunch of points there, and I was attending the session before us, and Kevin Thomas was talking about the monumental access problem. And I agree that I think AI can be a solution there. But you highlighted the case of, well, what if we have an AI algorithm that identifies someone has a 90% probability of having atrial fibrillation? And that's not the future, that's today. We have publications that show that. So as someone who's really versed in this, I'm curious to hear what other folks think. What probability do you need to then go and put someone on a medication that could leave them with an annual major risk of bleeding as high as 2.5% or more? For me, it's way higher than 90%. But I'm curious to hear what people think. So I'm gonna share a story. So I did my training in Hershey, Pennsylvania. We have a very large Amish population. And there was a patient who was Amish who came in, and they're all self-paid. They have a collaborative payment, kind of their insurance. But we talked to them about Loop Recorder after he had a cryptogenic stroke, and he said, I can't afford it. Said, okay, that's understandable. Looked at his echo, his left atrium was huge. And he had some features on his telemetry that suggested to us, like you said, this guy probably has atrial fibrillation. We make these decisions already, right? I mean, we're putting this burden of proof on an algorithm that's been shown to have pretty good discrimination, right? 90% is no joke. But what's important to understand is also that it's not just the atrial fibrillation signals that is picking up, right? There are other signs of changes in the cardiac structure that that algorithm is picking up. We don't fully understand that yet. But Mayo's done a couple of really good studies about, and again, this is true, this is kind of unrelated, but it has to do with predicting morbidity and mortality, right, in patients with seemingly normal ECG. And you're able to predict risk of mortality over a span of a few years with a seemingly normal ECG. So there are features there that we're not picking up. In the same way for atrial fibrillation, it might be the same thing. I would say 90% in somebody who has other features, right, structural changes in their left atrium, PACs, it's completely reasonable. John, to answer your question, I think that number is high. I wanna answer it directly and say like 98%, because I think, you know, whether it's the oath that we took at Do No Harm or what have you, and oral anticoagulation is highly litigated, whatever the reason may be, I feel like that number is high. I think we're also a little bit jaded clinically by false positives, right? So when we prescribe or a patient has a wearable and we know that specificity is in the 70-some percentage range and we see these false positives, right, in addition to implantable loop recorders even, where that should be, it is the standard of care, but still there are false positives. I think that drives that number higher and that's why I think with the treatment is a pill that they have to take once a day or twice a day and can result in life-threatening complications. I do think that that number is high in the 98% for me range. The 70% just doesn't cut it in specificity with wearables from my perspective. So we have about 10 minutes left and there is a question that I have been dying to ask you guys when I knew this session was coming up and maybe Laura, I'll start with you and then Brett, you can give your take after. So we, very strong evidence that prolonged monitoring is very beneficial for the patient who suffers a cryptogenic stroke. What does that look like in a world where everyone has a watch on that's monitoring their heart rhythm continuously? How do you envision this whole field evolving over the next couple of years as all these wonderful technologies we're hearing about today continue to progress and evolve and get us increasingly adopted? I think there's always a question about whether or not that watch is actually gonna work, whether the person is wearing it all the time. We could have 15 minutes of charging your watch and that 15 minutes you could have atrial fibrillation and then for the next 23 hours and 45 minutes, you don't. There's always that risk out there that there could be something. Because these aren't 100%, I have a little bit of a fear about that, but is it better than having nothing? Yeah, a loop recorder is probably ultimately better no matter what, but yeah, that's my always thought. My always thought is are you wearing it 100% of the time? Yeah, I love this question and I also love the fact that I'm gonna slightly disagree again and because we're having fun together. The first issue is this, the wearable. I am incredibly concerned about the false sense of security that someone might get. Every time a patient comes into my clinic, they tell me, I'm fine, my watch says I'm fine. Well, okay, but you're not, but I'm fine because I didn't feel, in the olden days it would be like this, I didn't feel my palpitation so I don't have AFib, right? We finally got patients off of that. I don't have AFib because my watch says I don't have AFib. So I think that this false sense of security moment is a big concern. When the technology gets better, which it will, we'll have a different conversation about that. The second point is about all those patients that are out there with more data that are now coming into clinic and it seems to be easier for us to say just wear your watch and let us know if something's going on. I prefer that if there's a signal, high sensitivity moment, maybe high sensitivity moment, hopefully that's what the data's showing, that then I can click my easy button, which we talked about, and get those patients shunted into electrophysiology because I don't know about you, well, I do know about you, but I didn't go to electrophysiology school. I went to stroke school and I know that this person now has a TIA or stroke or whatever and their watch says that they got a little something, something, I wanna click that easy button and get them over to John Sue. If the old way we would do it, they come into clinic, my answer was, ah, it's hard to get you into electrophysiology clinic, why don't you just keep wearing your watch and tell me if it shows you something. I wanna go the other route. I want more patients to come in and I wanna shunt them into Dr. Sue's clinic. Having gone to electrophysiology school, which was fun, by the way, the problem with, I think, the watch or the wearable question, my first question to a patient when they come in with AFib detected by a watch is did you interact with your watch? And usually they say no, so then I know it's a PPG that detected AFib and therefore the sensitivity's in the 70s and it's usually a false positive. Granted, that's referral bias, et cetera, but that's, I mean, I think that that showcases the issue. I think we need to improve specificity, specifically in regards to the wearables because of this aspect of it having tremendous treatment implications by methods like PPG other than a true ECG, so I don't know how that gets solved. I don't know whether prompting or other ways of doing that, but I do think that the sensitivity, oh, sorry, specificity in the 70% range is a problem. Completely agree. When you interact with the watch, when you get that telemetry single lead ECG, it does help. There are patients who said my watch said AFib. I took the one lead ECG and you're able to review it. You're able to review previous strips. You can have a baseline comparison and you can adjudicate pretty reliably that no, ICP waves. Well, I want to thank all of you guys for an amazing session. We're coming up here on time. I think I'll close this out by recapping some of the things we talked about. We talked a lot about if the patient's not getting evaluated or doesn't have access, that is probably our biggest barrier. We talked about differences in the clinical approach, the prolonged monitoring, and the risks and benefits of each type of technology. And we had some really good debates about how much is enough and how you optimally detect that. And I think one of the challenging things for me is I think we're getting so good at diagnosis and AI has so many capabilities with diagnosis that we're identifying patients And does that necessarily mean that treating those patients is going to yield the same benefits or the same risks as someone who has highly established persistent atrial fibrillation? Again, the nice thing about cryptogenic stroke is that secondary prevention is so important that a lot of those issues are minimal compared to the person who's never had a stroke or primary prevention. But I just want to open it up, everyone, for last words. And thank you again for joining us and participating in such an engaging discussion. I'll take a last word, which is we, John, thanks for having us, but we have two issues. The first issue is access that we just talked about. The second issue is technology getting us better throughput because if these patients aren't being seen by the right providers, then whatever we choose to do, we're not having that opportunity to do it. So we need to improve access and we need to improve the technology or lessen the technology hurdle in order to get these patients seen, evaluated, and therefore treated. I agree and I think we should support wearable technology. It will get better. There will be ways to do this non-invasively. And there's a lot of data that we already have before we put in the loop recorder. I'm a big fan of loop recorders. I think they do wonderful things, but I think there are situations where you have enough evidence to prescribe anticoagulation. And let's start looking at left atrial mechanics more closely. It's a huge area of research. I would encourage everyone to read about it. Ultimately for me, it's just about the patient and what the patient's gonna choose. It doesn't, in the course of things, it doesn't matter to me whatever way, whatever way the patient's going to choose is better than nothing. So I like to just make sure that the patient's taken care of. Yeah, that's such a critical point. An engaged patient is a patient who's gonna do well. Jonathan, last word goes to you. Maybe back to the patient. I think shared decision-making is all the rage right now, but I think that this era of paternalistic medicine is probably not an era, right? So I think involving the patient and having them involved in that decision-making process is really important. So as that becomes more prominent in cryptogenic stroke or in our ethos in general, I think it's important specifically with this patient population. Well, thank you all.

cryptogenic stroke

atrial fibrillation

anticoagulation

Artificial Intelligence

wearable technologies

patient access

shared decision-making