Western AF Symposium 2023: Session 13 Roundtable

Screening for Atrial Fibrillation: Should it be a Health Care Standard?
 

© 2023 HMP Global. All Rights Reserved.

Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of EP Lab Digest or HMP Global, their employees, and affiliates. 

Featured is the Session 13 Roundtable entitled “Screening for Atrial Fibrillation: Should it be a Health Care Standard?” from WAFib 2023. 

Western AF Symposium 2023: Session 13 Roundtable 
Screening for Atrial Fibrillation: Should it be a Health Care Standard?

Moderator: John Mandrola, MD – Baptist Health Louisville 
Moderator: Jeff Healey, MD – McMaster University 
Discussant: Edip Gurol, MD 
Discussant: Jodie Hurwitz, MD 
Discussant: Narendra Kanuru, MD 
Discussant: Melissa Middeldorp, MD, PhD

John Mandrola, MD: Welcome. This is a roundtable on “Screening for Atrial Fibrillation (AF): Should it be a Health Care Standard? I'm John Mandrola. I will have everybody else introduce themselves.

Jeff Healey, MD: I am Jeff Healey, an electrophysiologist from Hamilton, Canada, who is going to cochair tonight.

Edip Gurol, MD: I am Edip Gurol, stroke neurologist from Massachusetts General Hospital and Harvard Medical School.

Jodie Hurwitz, MD: Jodie Hurwitz, electrophysiologist in private practice in Dallas, Texas. 

Narendra Kanuru, MD: I'm Narendra Kanuru, electrophysiologist at the Wellstar Health System in Atlanta, Georgia. 

Melissa Middeldorp, MD, PhD: I’m Melissa Middeldorp. I'm a postdoc researcher at Cedars-Sinai Hospital in Los Angeles, California.

Jeff Healey, MD: Great. So, we'll get started with our expert panel in this very topical area of AF screening. I think this is one of these areas where we have multiple different perspectives on this, and it's good to see a very diverse panel here. So, to start off, and maybe John and I will bounce back and forth with questions, but where do panel members think we are with AF screening? Is the evidence there yet? Maybe that's the first question before we get talking about the who, where, and why. Where are we with the evidence? What do clinicians think, what do patient groups think, and what do regulators think? 

John Mandrola, MD: Before you answer, the US Preventive Services Task Force has weighed in on this, and they have said that the evidence is (I) insufficient. So, what do you all think? We've all agreed that we're going to keep our comments short so everybody gets to talk. 

Edip Gurol, MD: Of the patients with embolic or ischemic stroke admitted to stroke neurology divisions who either had previously known AF and were diagnosed with AF during hospitalization, the second category of being diagnosed with AF during hospitalization make up about 20%-25% of patients. So, if you could identify these patients and treat them with either long-term anticoagulation or left atrial appendage closure, we can probably prevent stroke in some of these. As a stroke neurologist, that is what I think I would start with.

Jodie Hurwitz, MD: Thanks. This is sort of hard to answer very shortly. I think screening is definitely here. We have to figure out what we're screening for, whether it's for stroke prevention or morbidity or mortality. We have to figure out who is the highest risk population and separate all of that out. So again, lots of different ways to answer that question.

Narendra Kanuru, MD: I think in some ways, the marketplace has overtaken what we talk about in terms of academic evidence. I mean, people are coming in all the time with these wearables. The technology is, for the most part, pretty solid. So, I know we talk about screening, but individuals are screening all the time right now, and we've got to deal with people coming in just for that reason.

Melissa Middeldorp, MD, PhD: I agree that it's insufficient. I think that we probably have some of the tools to undertake screening. I don't know that we're using them to the best ability that we could be using them. We also need to factor in the cost-effectiveness of this as well. So, looking at the correct endpoints is important.

Jeff Healey, MD: It's a great opening statement from everybody and it reflects the diversity of opinions. We also have to set our frame here, too. When we talk about clinical evidence and randomized trials for treatments, we have one bar. Screening is a very tough game because you must find populations where disease exists. You have to search it out and find it. Then, you have to get them on therapy, and that therapy has to work and they have to stay on it. So, instead of when we talk about numbers needed to treat, for example, in heart failure of 4 to 10 people, even things we greatly accept such as screening for hypertension, and then, the number needed to screen to prevent a stroke in hypertension is still in the 40 to 70 range. We can compare that where we are with screening for AF. But they're very difficult trials, requiring tens if not hundreds of thousands of people to move the needle on the hard outcome. So, the question is, do we need to show a reduction in stroke? Let's confine ourselves to stroke as we talk about this. I think it depends whether you're dealing with screening the population or screening a subset of people. So, maybe people could give a comment on what they think the bar should be in these different contexts of population-based screening and targeted screening. 

Edip Gurol, MD: I'm going to mention what is probably the most important target: patients who sustain an embolic or ischemic stroke without a clearly identified source and without AF. These patients really need screening and we know these patients also have a very high risk of ischemic stroke recurrence. So, we use external monitors for 2 to 4 weeks followed by insertable cardiac monitors for up to 2, 3, 4 years. It's a very important group of patients. There is a very high risk of stroke occurrence in these patients, in whom we can make a difference if it's identified as AF with either anticoagulation or left atrial appendage closure.

Jodie Hurwitz, MD: This is a really great question. I think the idea of monitoring screening patients after they've had a stroke to find etiology is something that we're all well aware of. In terms of screening patients who fall into high-risk groups for stroke, I think one of the big things is to try and figure out what you're going to do with that data. Does it imply that they should be on anticoagulation for a long or brief period of time? We have a lot of questions that are still out. I think that's where long-term screening is going to help us.

Narendra Kanuru, MD: I'd like to ask a general question: when we talk about screening, what does that mean? One-day monitor, 2-day monitor, 14-day Holter, or 30-day event monitor? Implantable cardiac monitors increase the cost dramatically. So, the idea of screening the general population is important to think about. What exactly are you going to do for these people? Even with implantable cardiac monitors with the trials that have been done, they're very effective in patients who have had a cryptogenic stroke already, and I think that's clear. But in the general population, what level of screening is adequate, and should we use, as you mentioned, a higher risk patient? So even without AF, if they have a higher CHA2DS2-VASc score, should maybe those patients be up for more aggressive monitoring?

Jeff Healey, MD: That's an excellent point that gets into our other question about how aggressively one screens. I think one thing we should all discuss is the setting. A recent trial, VITAL-AF, done by very well-run primary care clinics in and around Boston, used a single time point screening and really did not move the needle even on AF detection, let alone stroke, probably because the background medical care in this setting was incredibly good. You take that same tool and deploy it in a rural area in a low-income country where rheumatic heart disease is highly prevalent, and all of a sudden you may have the most single most cost-effective screening program on planet Earth. So, what about context and background medical therapy?

Melissa Middeldorp, MD, PhD: This is a difficult question, because in Australia we've undertaken some screening which has been done in GP practices or pharmacies. It's looking at AF, not stroke prevention, but AF and then working towards stroke prevention. Once again, the uptake is low, you're not getting high numbers of patients. So, I think we have to really focus in on targeted populations in targeted settings to be able to see a difference and make it worthwhile.

John Mandrola, MD: Excellent, targeted population. Let's take that and run with it. So, we have a randomized, controlled trial. The LOOP Study, which randomized patients 75 years old, high CHA2DS2-VASc risk—you couldn't pick a better population to screen for AF—and they put in loop recorders, which you don't have to rely on intermittent recordings and is probably the best screening device that we have. When it came out, there was a 20% reduction in stroke. However, it didn't meet statistical significance, and it was countered by, nearly the same, a 26% increase in major bleeding. So, LOOP finds that you detect 3 times more AF, you start anticoagulants 3 times more, and you don't move the needle. So, I would love to hear why that is. Did LOOP not target the right population, and can we explain this?

Edip Gurol, MD: Thank you so much for asking this—this is extremely important. I think the problem here is our obsession with lifelong anticoagulants as the only preventive measure for AF. Yes, 20% risk reduction is good, but 26% major bleeds is horrible. So, the second step is and should be to identify who is at high risk of bleeding and use long-term anticoagulants, preferably DOACs, in people who are not at high risk of bleeding, but use left atrial appendage closure in people who are at high risk of bleeding. That way, you can cut down considerably on the risk of bleeding in these patients. So, I think this is an extremely important point. If you do not stop this obsession of seeing lifelong, full-dose, systemic, anticoagulant treatment as the only treatment modality for stroke prevention, we are not going to succeed, especially with these low-risk populations.

Jodie Hurwitz, MD: I totally echo a lot of what you just said. Some of the interesting things in LOOP is to look at the fact that in some of these patients long term, they develop less AF later on in time. So, we have to figure out who are those patients and if they should be on anticoagulation all the time. The other issue is, and I keep coming back to this, is it actually AF that causes the stroke? If we can figure that out better, maybe target anticoagulation for a period of time, it's a really high-risk patient in a high-risk situation.

Jeff Healey, MD: So, the different technologies have different features. Single time point is very inexpensive, it picks up largely persistent and permanent AF, which, by the way, also has the highest risk of stroke. There's roughly a doubling of the absolute risk of stroke when one moves from paroxysmal to nonparoxysmal AF in our clinical trials and registries. So, for picking up people who are in AF all the time, it is quite good. On the other side, you have continuous monitors or even implantable monitors which are very sensitive for picking up all the AF that may exist, but from what we understand from pacemaker studies, this has a lower absolute risk. Perhaps you need to consider the zeal with which you go after finding AF and really frame it in the setting of how high risk is the population. Post stroke population, maybe that makes sense. Primary prevention or community screening, maybe that doesn't make as much sense. These are the things that one has to trade off with the various technologies.

Narendra Kanuru, MD: That's a great point. We have long-term monitoring in a large population, which is our pacemaker patients. So, we're following some huge number of pacemaker and defibrillator patients in our clinic. We pick up AF all the time, and anybody who's on an EMR gets alerts about this daily about something that's happened. In our practice and our institution, if somebody has an elevated CHA2DS2-VASc score, we will start them on anticoagulation with 6 minutes of AF. Based on what you've mentioned, I have to wonder if we are doing them a disservice. 

Jeff Healey, MD: It’s going to be an interesting year. We'll have NOAH AFNET present probably at the European meeting in the summer, ARTESiA present probably at the American Heart Association, and ARCADIA also coming out this year. We know that short-lasting episodes picked up with long-term monitoring in pacemaker patients are associated with increased stroke risk. This is true, but it's a 2-point fivefold increased stroke risk, not a fivefold stroke risk that we see typically with AF, and the treatment is going to be quite interesting. NOAH was stopped for futility, and so I think the world may change a lot in the next 6 to 12 months.

Melissa Middeldorp, MD, PhD: Are we screening to look for burden, though, or are we screening to look for the diagnosis? I think that's the question that comes into that, because then it depends on what type of device you are using. With pacemakers and loop recorders, we have to get some form of burden to be able to look at anticoagulation. But when you're screening a general population, you're looking for a first diagnosis. So, you've got to determine what the point of the screen is.

John Mandrola, MD: Before I ask another question, I want to push back on Dr Gurol. I don't agree at all that LOOP was not a positive trial, because they didn't use left atrial appendage closure. But be that as it may, one of the interesting things that I do want to ask the panel about is that Soren Diederichsen published an interesting post-hoc analysis of the LOOP study, in which they found significant amounts of bradycardia. Then, when they looked at the bradycardia in the screening arm, they found a 53% increase in the rate of pacemakers. This was asymptomatic bradycardia, but when they looked at syncope or sudden death, they found no difference. So clearly, this was an overuse of pacemakers. I'm reading Wilson and Jungner's criteria about screening, and one of the things that comes up is what are the harms of this? Should we consider that, and how much should we consider that? I guess one harm is we've already talked about the increase in major hemorrhage. But I guess there are other harms as well.

Jodie Hurwitz, MD: I think that you bring up a really good point, because we've all seen the worried well walk into our offices or gotten inundated with phone calls. As you alluded to the fact that everybody's wearing watches now, or that those of us who are in charge of device clinics are constantly getting messages that there are a lot of people receiving very short episodes of asymptomatic and brief AF or heart rate episodes, and what do we do with this. One of the things is screening. As I said, screening can be looked at in a lot of different ways, and I think one of the things that would be very helpful for us to is to really come up with some better guidelines or consensus document in order to make sure that we support what our patients are doing, we support the information they're getting with their watches, wearables, or whatever they have and that we teach them so that they can be better help to us.

Jeff Healey, MD: That is a good segue point. We have to have an approach for what's being done, even if the USPTF doesn't submit a recommendation, because people are being screened and the consumers, the wearables, and the patient-facing diagnostics are out there. I'm not sure who has ever gotten more than 200 AliveCor transmissions in a single day, but it is definitely happening. So, I think we need guidance on how to manage these things. It’s important. One other quick point. Most of us are EPs or interventional cardiologists. We're used to doing stuff in our intervention trials that look a certain way. So, I would just point out that if people are looking for a 25% to 40% reduction in stroke with a screening intervention, you're going to be looking for a long time, right? Treatment effects in these types of trials are measured in 3%, 4%, or 5%, and again, if you take a benchmark of something such as hypertension, remember, most of our data is imputed for hypertension. We have big meta-analyses of treatment trials, but if you randomize people to measure blood pressure, those numbers get pretty small, even though it may still be a good intervention. So again, it's worthwhile looking up the screening literature. What are the types of things that are out there? What sort of numbers did they deliver?

Edip Gurol, MD: A very important issue, again, in trying to prevent embolic stroke is not to disproportionately increase the major hemorrhages into several hemorrhages. So, data from DOAC randomized controlled trials clearly show us that. It’s DOACs and warfarin across those 4 trials. Endocardial causing major hemorrhaging was 2.5% to 4% per 100 patient-years, and non-major hemorrhages occurring 8.7 to 15 patients per 100 patient-years, and any hemorrhage was about 25 patients per 100 patient-years. So, these drugs increase the risk of hemorrhages tremendously. We know that less AF means less embolism. That is very clear. No AF, you don't get any benefit from all anticoagulants. So, in order to help these people who have relatively low baseline and embolic stroke risk, you need to have modalities that do not increase the risk of hemorrhage. But this is why I like left atrial appendage closure so much.

Nassir Marrouche, MD: This is a question for Jeff. We're dealing with this today and now. When all of us go back to our clinics, the patients are coming to us with AF on their watches and on the tools they have today, which we did not have 10 or 15 years ago. So, I would like to have everybody's input on how to deal with that patient. What does it mean in terms of outcome? What I'm struggling with now is when a patient comes to me and says they have 5 or 10 seconds of AF on their watch. He's my patient. He has an AF diagnosis and I have to deal with that. How do we manage this? I'm calling it the “Apple Watch” or “smartwatch” AF. We're struggling to know what to do. Obviously, echo, Holter monitor, stress tests, and lab work. But a patient will be at home watching TV and record 10 seconds of AF on their Apple Watch, FitBit, or PPG. Before we get to the trial you're talking about, we have other trials that are ongoing right now, but this is a population that everybody in this room is dealing with now and we are struggling what to do next. So, what should we do with these patients who are in the clinic today? What should I do with this patient? Send them back home? 

Jeff Healey, MD: It's an excellent question. I don't know if you're a sailor, but clinical trials are like marker buoys. You don't have them lining your river or lake. They're there for guidance, and most of your decisions on the water are made by yourself with your general knowledge. The pacemaker/ICD and AF is a very interesting analogy to what we're seeing right now with consumers. We didn't ask for that. We just turned up in the pacemaker clinic and there's all these logs and electrograms, and we didn't know what to do with it. I think we will have a much better idea what to do with it by the time we're ringing in the New Year in 2024. In terms of the wearables, I think we're less clear. But we do need to make decisions in the meantime, and these decisions may be correct some of the time. I think you have to use your best judgment. You must weigh the clinical risk factors. You have to weigh the burden of the disease, and you have to make your best clinical judgment when we don't know. But I think the most important thing is we have to accept and admit that we don't know. We're willing to do trials, keep an open mind, and learn more, because we're right about half the time when we predict the future. 

John Mandrola, MD: Let's hear from the end of the table about what to do with these Apple Watch/smartwatch episodes, which Nassir referred to. 

Narendra Kanuru, MD: I try to make sure they're referred to my partners. But be honest, a basic workup is what we do. I struggle with patients who are asymptomatic who pick this up. A lot of devices now have ECG monitoring and pick up irregularity. For symptomatic patients, I feel like there's a bit more of a proper path for them. But for asymptomatic patients, it's a bit of a challenge. The idea of a burden in these patients is also a little more difficult to get. I mean, we're not getting the same kind of burden out of a watch as we're getting out of a pacemaker, for example. To that end, even when I see somebody with a pacemaker who has 15 minutes of clear AF on the EGM, we'll start into migration. So, then we go a year and they've had no other episodes. Do I have the guts to take them off anticoagulation at that point? I don't know. Especially on a low CHA2DS2-VASC score, of course, but on a high CHA2DS2-VASC score? I'm not sure that I’d do that, because we're always afraid of sins of omission, taking them off anticoagulation and then all of a sudden, they have an episode that they're not aware of and have a stroke. It’s the same with these smartwatches that come in. I'm not sure what to do with them in general. If I see somebody with that, if they're candidates, I'll put them on anticoagulation. We'll try to sort through the symptoms and make a decision on what's to be done about it, but it is challenging.

Melissa Middeldorp, MD, PhD: I'm obviously not a clinician, so I don't treat patients, but I have a question for the clinicians. If you do pick up short-term monitoring on a watch or AliveCor, do you then refer the patients for longer term monitoring? Are you putting a patch on or are you thinking about implanting a loop recorder? What would be your standard of care?

Jodie Hurwitz, MD: This is a fantastic question. Right now, the definition of AF that we treat is on an ECG or 30 seconds on a monitor. It's not 5 or 10 seconds on a wearable. But I think we've all heard over the last day and a half about treating the patient and not just the arrhythmia, and so perhaps what it does is after we work the patient up, what we do is we say these are your risk factors and these are the things that you need to change. We'll continue to follow you. If I had somebody who had a high CHA2DS2-VASc score and had other morbidity and mortality, I would definitely put longer screening monitors on them. 

Edip Gurol, MD: I love the question about 10 seconds of AF. I can tell you what's happening right now. So, this person goes to his or her primary care physician (PCP). This PCP prescribes apixaban thinking that bleeding complications are no more than aspirin, which is wrong, because the on-treatment analysis showed 56% elevated risk of major hemorrhages. So, this person comes back home. If they have access to the internet, they will tweet “the Apple Watch saved my life” and tag Apple and Apple likes it, which has tens of millions of followers. Then, this person feels like a king. Later, this person starts to have some nosebleeds or whatever. The physician says, “this is the price of using this.” In 5 to 6 years, they come back with major bleeds. They go to their PCP and are told that was your fate. So that is what's happening right now. We definitely need randomized control data. We need good evidence. Unfortunately, these things are being directly advertised to end users. Look, I have one. By the way, don't get me wrong. But there are a lot of problems there.

Jeff Healey, MD: I'll just say something to provoke a little bit. First, would we even be having this discussion? Remember, ASSERT was done and ARTESiA was designed in the pre-DOAC era. We almost couldn't get the trial off the ground, because people would never contemplate starting warfarin for 6 minutes of AF on a pacemaker. Does the presence of an easy-to-use drug, perhaps safer, does that modify this in any way? Is that coloring the way we decide this?

Narendra Kanuru, MD: I don't think there's any question that it's made it easier. Starting somebody on warfarin with all the issues that we're all very familiar with compared to the DOAC era—there's no question. In my practice lifetime, I've seen that transition from VKA to DOACs. There’s no question that it has probably driven a lot of our decision to start anticoagulation on such small amounts of AF that we detected on devices. 

John Mandrola, MD: Before I take this question, I want to take a moment. What I'm hearing is this tension. On the one hand, we have patients come in with these short-duration episodes that may lead to harm in treating them. We may have created harm, but on the other hand, the tension of screening is that some of these patients will have had permanent AF and we probably will have helped them by finding it. So, in an overall trial like LOOP or STROKESTOP, there may be very small effect sizes. But within that, that's the thing about screening that makes it so challenging is that we will help some people. 

Audience question: Just to state my bias, in addition to ablating AF, I'm a Watchman implanter. So, we find AF, we put in loop recorders. I'm pretty aggressive about it, and we find a lot of things. Then, you talk about the high bleeding risk and the increase in bleeds. I think we all have a moment of pause when you come up with something and you're looking at the patient thinking, “They don't fall. They haven't had a blood transfusion. I don't have a great reason to do a left atrial appendage occlusion on them.” So, I sometimes wish I did have a better reason than we're offered. For high-risk populations, high CHA2DS2-VASc score people who have had TIs and strokes, absolutely. In terms of prevention, we're going to have a lot of high yield. There are sick patients. Do you see the guidelines moving toward allowing left atrial appendage closure more for risk than actual history of having declared themselves with bleeds and falls and things like that?

Narendra Kanuru, MD: I'll take a jab. I think there's certainly a push, and I can't blame industry for wanting to go in that direction, that the treatment for AF in patients who are at risk for stroke is left atrial appendage closure. That is a colossal sea change. I mean, it is an overall safe procedure, but it is quite invasive. Like you mentioned, I obviously do ablation and I'm also a Watchman implanter. That's tough to say. I'm not sure I'm ready to tell everybody who comes up with AF, “Here's the solution. You have an invasive procedure and a device left behind as well.”

Devi Nair, MD: I want to add on to that. I completely agree that left atrial appendage closure is not first-line therapy for stroke prevention. I think right now it is class IIb indication, and it is for patients who cannot tolerate anticoagulation and should have a genuine reason why they cannot take anticoagulation. But I think we will have those answers when we have results from CHAMPION and CATALYST, and hopefully, there are robust trials that are going to look at appendage closures for first-line therapy. So, I guess time will tell. In about a couple of years, we'll have answers from OPTION, CHAMPION and CATALYST. 

John Mandrola, MD: One of the problems that I see with this whole thing that we're talking about, and I think it's a reasonable scientific question, but since 2016, there has been 50,000 implants of left atrial appendage closure in the United States and there are zero randomized clinical trials. Imagine a scenario where we have randomized, even at 20%-30% of those patients, we would have had an answer. So, I find it a disappointing thing. By the way, I'm not anti-Watchman because I hate Watchman. I'm anti-Watchman because I don't think the data is compelling. If there was data, I would change and learn how to do it tomorrow. 

Edip Gurol, MD: What do you think about PRAGUE-17? 

John Mandrola, MD: We're not going to adjudicate PRAGUE-17 here. Let's move on. 

Jeff Healey, MD: We're probably drifting off our primary goal about AF screening, which is thorny enough. 

Tom Deering, MD: First, I appreciate all your comments. We live in a gray zone. We need more clinical data to figure out where we go from research. Also, we have to be comfortable being honest and being fair and telling our patients we're not sure, and all of you said that nicely. But I have another question for you. I am very concerned about the false positivity that is manifesting with the use of these devices, whether it be algorithmic for AF where there's artifact and PACs, or whether it be bradycardia for a gain issue or another thing. I think you mentioned it earlier, John, that we're seeing too many pacemakers inserted and we're seeing too many people treated for diseases that are not really there. So, what do you recommend that we do, and how do we make that a systemic issue to prevent those overutilizations, which have significant consequences potentially, especially where they're not indicated?

John Mandrola, MD: I would add, that's one of the reasons why the USPTF found insufficient evidence, because they looked at benefits and harms of screening. So, it's a very good point. 

Jeff Healey, MD: I think that there is a phase that all new technologies have to cross. I run a genetics clinic. I have a whole clinic day just seeing variance of unknown significance. That's the downside of clinical genetic testing. I mean, it's a happy day when no one has anything. But it's the same thing with procedures and surgery, right? You have to bump through several years before you can get enough people in the world good enough at doing whatever it is to possibly execute a trial with any kind of effectiveness. So, I think we just have to have our humility and soldier on through the period of evidence and be prepared to generate it.

Jodie Hurwitz, MD: I would echo exactly what you just said. I mean, let's be realistic. These are here to stay. So, it's very important for us to say this works, this doesn't. This gives us important information, this gives us nonsense, this helps us. This scares the patient, and these are perhaps the better ways for us to gather data. So, we need to be a little more proactive. That's a great question.

Melissa Middeldorp, MD, PhD: I agree. I think we really need to hone in on the purposes of all this. I think what Jodie says is right, we need to create some really quick guidelines on how we're going to manage this. Are we going to continue mistreating patients effectively, because we are putting in pacemakers in patients who don't need them? That's also a cost issue. You're treating people with anticoagulation when they probably don't need it. We really need to define this much more specifically and how we manage these people.

Audience question: So, I want to come back to the comment about a patient coming in with their watch showing AF. We're so focused on stroke as an outcome, but maybe we can take that opportunity to implement risk factor modification, and Melissa has done a lot of work on that, with exercise and weight loss. Maybe that should be the focus rather than trying to prevent strokes. 

Jeff Healey, MD: It's a very broad lens, and it's a nice perspective. Our patients, for the most part, are not dying of ischemic strokes, right? They're dying of heart failure. These are patients with AF. That's why they die. So, there's a lot of merit to a broad lens.

Edip Gurol, MD: Can I ask my question now?

John Mandrola, MD: Hold on a second. I think this is an interesting point. Does anybody have any answers to that question about off-target beneficial effects of picking up AF? 

Melissa Middeldorp, MD, PhD: I guess you're right. I mean, we have an opportunity to impact these patients early, and we keep talking about treating patients with AF early. So, if you're getting a patient who's coming in with some short episodes, let's target their risk factors first, start looking at their blood pressure and whether they have diabetes. Assess if there's some sort of underlying heart failure. Start them on risk factor management. Get them exercising. Change their diets. There are simple measures that can be taken that can be quite effective. So, I agree we should be probably implementing that first and then working from there.

Jeff Healey, MD: So, to be provocative again, if they come in with AF on their Apple Watch, do you tell them they should never drink again? 

Melissa Middeldorp, MD, PhD: We say 3 standard drinks a week. 

John Mandrola, MD: To highlight that point about off-target effects. I talked to Soren Diederichsen just before we came on about his theory on why STROKESTOP was a positive trial, it was P of .049 or whatever, and his view, which I think is reasonable, is that the invited to screening group got more care, a sort of beneficial performance bias. It was the same thing in DANCAVAS, where the screening group seemed to get more attention. So, that may be one of the beneficial effects.

David McManus, MD: This will be an easy one for you. So, if we're having trouble with short episodes of AF, as was highlighted by Nassir, and what to do with that, and especially given the prior question, which I thought was a good one, which is there may be benefit to treatment upstream, how are we going to deal with the entity pre-AF as it's predicted based on AI on 12-lead ECGs, which of course, we can do today? I think the screening discussion highlights an opportunity for us to think about classification of either pre-AF or AF without structural heart disease or without stroke risk, and do trials to look at interventions, because I would argue I don't think there's a lot of evidence to support. Exercise got a very limited side effect profile in the negative, but I haven't seen a lot to really show me that interventions matter for prevention of stroke from AF, never mind the anticoagulation discussion. So, your thoughts on pre-AF, predicted risk of AF that's going to be coming for polygenic risk scores and AI on 12-lead ECG?

Edip Gurol, MD: I can take this one if Dr Mandrola does not oppose. The clear thing is that there is no benefit from neither anticoagulants nor left atrial appendage closure. There are a lot of studies, especially involving anticoagulants in patients who might have had PAF, like atrial cardiomyopathy and all these things. ATTICUS was negative. It was an atrial cardiomyopathy trial, and the ARCADIA study was stopped. So, there is really no evidence, and I'm learning that the NOAF study was also stopped for concern of no benefits. So, there is no reason to consider anticoagulation unless you detect AF. I think this is one very important issue. So, regarding how much of AF is enough to treat with anticoagulants or other measures, I'm going to ask Dr Healey. In the AVERROES initial paper, there was no subgroup analysis comparing the effect of apixaban or aspirin between permanent versus paroxysmal AF in patients. We know for sure that again, paroxysmal AF is associated with a much lower risk of stroke. Did you guys ever publish these data? Did AVERROES show benefit among the paroxysmal AF patients?

Jeff Healey, MD: Yes. So, the big paper was probably Thomas Vanassche’s paper in the European Heart Journal, and it was a pooled data set from AVERROES and ACTIVE-A in patients not receiving anticoagulants where there was definitely a graded increase in absolute risk of paroxysmal, persistent, and permanent being the highest, roughly 2x. The individual power if we’re looking at interaction was under power, because of course, AVERROES was stopped early, one year of median follow-up. There just wasn't power. I was going to say, maybe I can make a pointed question. We've got high-risk populations for stroke prevention. Maybe we leave them to the side. Maybe we don't dispute that if a patient comes to our office for angina or heart failure that we should measure their pulse or do an ECG or something, opportunistic case finding. But let's talk where we started maybe about the USPTF and recommendations for population screening. Everybody age 75 and over should be screened for AF. We have many clinical trials are very diverse, over 350,000 patients in them, a plan for a patient level meta-analysis with virtually every study participating. The pointed question for the group is do we need to show a reduction in stroke before the USPTF changes its I rating to something more favorable? Do you need to show you can prevent stroke with population-based screening? 

John Mandrola, MD: Yes, and no harm as well.

Edip Gurol, MD: No.

Jodie Hurwitz, MD: Yes, and morbidity decrease.

Narendra Kanuru, MD: I would say no, because there are other issues other than just stroke.

Melissa Middeldorp, MD, PhD: I would say no. I think we need more data.

The transcripts have been edited for clarity and length.