false
Catalog
Inside EP: Insights From Clinical Decision-Makers ...
How mHealth Will Impact AF Diagnosis/Treatment
How mHealth Will Impact AF Diagnosis/Treatment
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
So, let me talk to you about mHealth and how it will impact AF diagnosis and treatment, and these are my relevant disclosures. So, what is mHealth? A lot of people use mHealth and digital health interchangeably, but you may see some of these other terms on the far right. But essentially, it's sort of the intersection between healthcare and digital technologies that are designed to either acquire, collect, manipulate, or share healthcare data. And I think that there are three ways of looking at this type of digital health. One is sort of doctor-facing. This is where the information resides with the traditional closed-loop medical establishment infrastructure, and I think the electronic medical record, implantable cardiac devices, and the heart monitors that we prescribe are all sort of examples of doctor-facing digital health. The patient-physician communication digital health is telehealth, right, where we can remotely assess a patient, or the patient portals like MyChart, where the patient communicates with their healthcare team. And then lastly are the patient or consumer-facing technologies, and in general, these are technologies where the data resides with the individual, and they could choose to share it with their healthcare team as they want. And the wearables that we're talking about, or the self-monitoring that you do with the blood pressure, or ECG, a smartphone adapter, or internet searches are all considered sort of patient or consumer-facing. So actually, HRS last year sort of put together a group of us, and I was part of it, to write a paper in HeartRhythm about a transparent sharing of digital data, a call to action. And I'm going to show you some interesting sort of data where they surveyed patients with implantable devices and who use wearable devices, these are HeartRhythm patients, and ask them what you want from this data. And interestingly enough, patients who have implantable devices, right, it's in their own body, but through remote monitoring or in-person visits, they have no idea what's going on in their own body, right? The data comes from them to us. But actually, right, the most common response is that patients want to know what arrhythmias they had in their implantable devices. They wanted to know information about how their health status was changing. They wanted reports on their own bodies at sort of predefined intervals. And they wanted the results expressed in a non-technical language that they can understand. When it came to wearables, they want to know, how do I show my doctor, right? How do I get this from my Apple Watch to my healthcare team? And interestingly, what do I do with the results at 2 o'clock in the morning? Which is really fascinating because, you know, many of these patients have a known history of atrial fibrillation, are on anticoagulation, right, they're our patients to begin with, and their watch tells them they've gone into atrial fibrillation and they panic, even though they have no symptoms and are, in theory, protected against a thromboembolic event. So they want that reassurance. We then asked physicians, right, electrophysiologists, what they want to give their patients. And I think, unanimously, we all agree that patients deserve the right to know about what's going on in their own body, right? We want them to have that information. What we can't agree upon is how much information they should be given, right? We believe that they should have a high-level summary, and we believe that they should have more granular information should they want it. But we sort of touched upon this earlier. The disaster would be that we give everyone everything, right? And you're going to be called for 5-ohm changes in impedance once a week. So we have to find that sort of comfort zone, but there are going to be some patients, typically engineers by training, who want to know everything, and there are some patients who just sort of want to know the high-level stuff. When it comes to consumer wearables, that's a bigger issue, right? We want to explain to our patients how to transmit the data, right? Is it going to be through MyChart? Is it going to be some other patient portal owned by industry, right, where we can access that information in the cloud? Is it going to be through texts or emails? We want to define how frequently the data is transmitted. We certainly do not want to know your heart rate every five minutes, right? We also want to understand, excuse me, that's my watch calling me. We also want to understand how frequently it's transmitted, and it may change. If we just cardioverted someone, right, we may want to know what their rhythm is for the next few weeks, but then sort of stop, right, and know at some infrequent interval. And then, and very importantly, we need to establish expectations. I must confess to you that I have handled this poorly because I have given my email out to many patients. So last night, someone emailed me at 11.45 with their, you know, a flutter at 160 beats per minute saying, please call me. Well, I've been asleep for two hours, right? So patients need to recognize that lo and behold, their doctors sleep, and just because you send the information to me via email, it's not reasonable to expect a response at all hours of the day, all days of the week. So monitoring has changed, and we have been the beneficiaries of a lot of the progress in sort of this consumer grade and wearable technology because you can measure the rhythm of the heart relatively easily. Does anyone know who and what this is? Shout it out. You're correct. This is Norman Holter. Norman was not a physician. He was a tinkerer, and he was interested in designing a device that can measure the rhythm of the heart outside the hospital setting. The good news is that Norman achieved his goal. The bad news is that this is the device. It weighed 85 pounds, had to be carried on your back, and I think could record the rhythm of the heart for almost 24 hours. We have come a long way. There's a large number of devices out there. Many of us have moved away from sort of the wire technology and more towards sort of patch technology that can monitor the rhythm of the heart for periods of time. But remember that the ability for someone to wear a monitor, right, is inversely proportional to the time you want them. So the compliance with a 30-day wearable monitor is probably 50%. It just can't be done. The other issue, right, is that these are expensive and need to be prescribed, and I have no doubt that the future of AF monitoring will be smartphone-based. I show you here a picture of the Pope's inauguration in 2005. If you look closely on the bottom right, there is one individual with a device capable of taking a picture and making a phone call. This is that same exact scene eight years later. There is not one person in the audience who does not have a device capable of taking a picture and making a phone call. There are now more cell phone accounts on the face of the earth than there are human beings on the face of the earth. And in the United States, this is cell phone use and this is smartphone use, and actually now more than 80% of Americans have a smartphone, and that cuts across all sort of socioeconomic and demographics. I mean, smartphone use is almost ubiquitous, particularly in people sort of under, I would say, 85 years of age or so. So how can we turn our smartphone into an ECG monitor? Well, there are kind of two ways of doing it. The first one isn't an ECG at all. It uses technology called photoplasmography, or PPG. You could download an app, put your finger up to the lamp, and it will measure your pulse for 30 seconds, and it could tell a regular pulse here from an irregular pulse here. And in highly controlled settings, it's pretty sensitive and specific, but once you let it out in the real world, people apparently can't hold their finger up to the light for 30 consecutive seconds, and therefore the sensitivity gets diminished. Still, in the year 2020, we don't diagnose atrial fibrillation by patterns of PPG. We require an ECG, and you could turn your smartphone into an ECG device for a very small amount of money on the internet. And this is a device called Kardia. It's a single-lead ECG that you put your fingers on, records a modified lead one, and the algorithm will read this as either afib, sinus rhythm, or indeterminate. And actually, it doesn't even need to be on the phone. It just needs to be near the phone, and it transmits the information with high-frequency ultrasound, which actually I didn't know until I was taking care of a young individual who had children and a dog. And every time he used his Kardia, everyone would go berserk, even though he couldn't hear it, because it's so high-pitched that an older individual can't hear it, but the high frequency is actually quite annoying to your animals and perhaps small children. So this is pretty accurate, but it gives no diagnosis in about 15%, often due to sort of compliance issue with noise. There is now a six-lead device. You put your fingers on the top, put the bottom on your leg or your ankle, and it could record a six-lead ECG, and you could send that to your physician. And I use this all the time. I believe that this technology has saved my patients' ER visits and doctor's visits, and it allows me to manage my patients remotely. Here's a patient who I ablated who was traveling internationally. He complained of palpitations, was panicked. He sends me his ECG, and he's in sinus rhythm with single extra beats or PACs. Great. Here's another individual. I was using this technology to increase her dose of an antiarrhythmic drug called flecainide. And even if you don't read ECGs, you could see the widening of her QRS, right, consistent with toxicity of the drug. We were able to bring her in, wash out the drug, and watch her before something bad happened. And while these devices are quite accurate, remember, you still need to overread them. So here's a patient of mine who flipped out when she saw this tracing. This is clearly normal rhythm with noise, creating artifact, which the machine misread as a very rapid and irregular heart rate. Well, all of the technologies I showed you are problematic because they only provide a snapshot of an ECG, right, a 30-second snapshot. The problem, of course, is that AFib can be highly intermittent. And this gets into this concept of AF density. Now you could read the definition here. I could show you the formula here. And truthfully, I don't really understand either of them. I will, however, take you through an example, and I think it explains it nicely. These are two patients in whom they both had dual-chamber devices. They both had atrial fibrillation. And this is sort of a reconstruction of their AF pattern. The patient in the red has daily episodes of atrial fibrillation lasting minutes to hours. The patient in the blue has one long episode of atrial fibrillation lasting a few weeks. They both have 21% AF burden, right? But if I want to find the atrial fibrillation, if I did a single 24-hour Holter monitor randomly, right, throughout the year, my sensitivity for the patient in the red would be 100%. It is literally shooting fish in the barrel. The sensitivity for the patient in blue would be only 23%. If I had done random 30-day monitors, right, in the patient in the red, the sensitivity of finding and diagnosing that patient with AFib would be 100%. The patient in the blue would be 26%. So if we want to find an intermittent rhythm like atrial fibrillation, doing a snapshot, doing a snapshot every day is probably insufficient, right? What we need is a continuous, wearable, opportunistic device that intermittently checks the pulse for atrial fibrillation. And that device obviously is widely available. If you've ever looked on the back of your watch or your Fitbit, there's a green light that goes on, right, when you're exercising and intermittently throughout the day. And the purpose of that is to intermittently measure your pulse. And this device can see PPG, which looks like this if you're in normal rhythm, and looks like this if you're in atrial fibrillation. So your Apple Watch is a passive, opportunistic AF monitor. And in fact, the earliest iteration was not made by Apple. It was made by this other company, Cordia, that used an earlier version of the Apple Watch that did not have this algorithm in it and put this sensor on the band. And basically what the watch would do is it would monitor your heart rate, heart rate variability, and level of activity. And when it saw that your heart rate was faster than you were moving, it would give you an alert to touch the band and record the ECG. The good news about this watch was that that green light was on all the time, which means that it was always checking your pulse, which also means that that battery wouldn't last 24 hours. You needed to charge it throughout the day. So the question is, how accurate can a watch be for the detection of atrial fibrillation? Well, we bought a whole bunch of these, and we experimented with them. And we gave them to patients with a known history of atrial fibrillation. So here's a patient. You could tell when they go into AFib, not only does their heart rate go up, but their heart rate variability increases. They touch their band and record a very nice ECG from their wrist. This is an ECG recorded during sinus rhythm. You can see the heart rate and heart rate variability particularly is slower. This is their activity in the green. And not only can you tell, does this patient have AFib, you could tell how much time they spent in AFib throughout the day. And you could tell when they took their watch off because they have no heart rate and no activity during that 9 a.m. to 10 a.m. period. So we actually gave 24 of these watches to 24 patients with an implantable cardiac monitor in place. And one of our fellows, Jeremy Wasilow, actually looked at more than 31,000 hours of continuous data. He's now medicated, but he's doing well. And this is a 14-hour episode of atrial fibrillation on your $10,000-plus implantable monitor. Here's that same 14-hour episode of atrial fibrillation on your $400 Apple Watch. The sensitivity for episodes of AFib for more than one hour was 97.4%. And the correlation of the duration of AFib on the watch versus the implantable monitor was nearly perfect. Well, of course, Apple has crushed this industry, right? The release of the Apple Watch Series 4 and Series 5 have really brought this to the forefront and has really changed sort of my practice. Not a week goes by where someone doesn't come to me because their Apple Watch told them that they have a rhythm that they didn't feel. So actually, the story is very interesting. Apple did not set out to define or design an AFib monitor. They set out to design a device capable of accurately tracking the heart rate with exercise. So they released the Apple Watch Series 1 and 2 with a very accurate heart rate tracker. But then these stories began to appear in the press, and people began to write letters to Tim Cook, the CEO of Apple, or Tim Apple, as our president calls him, and say, the Apple Watch saved my life because the Apple Watch told me I had atrial fibrillation and my doctor put me on a blood thinner. Did it save their life? Well, maybe not, but it may have prevented them from having a stroke. So the Apple Watch has some interesting features. For those of you who have not seen it, it's very simple how to do an ACG. You basically touch that icon, hold your finger on the crown, record a 30-second ACG, and the algorithm will, again, adjudicate it instantaneously as either sinus rhythm or atrial fibrillation. And then you could actually scroll down and add your symptoms. So you could send that PDF to your doctor saying, I had shortness of breath, and this is what my ACG looked like. And actually, not only does it have the ability to look at your ACG, but as I said, it has other features that allows it to passively look for atrial fibrillation. So there are three features, right? One is the high heart rate monitor. When it sees that you've not been moving for 10 minutes, it will check your pulse. And if your pulse is rapid, it will give you an alert. The other feature is the irregular rhythm notification. And what that does is throughout the day, when the watch sees that you're not moving, it starts to check your pulse. If it sees that your pulse is irregular, it will increase the frequency of pulse sampling. And then every 15 minutes, it will check your pulse. And if five out of the six pulse checks show atrial fibrillation, it will give you a warning that you've been in AFib. What that means is that the watch is not going to see an AF episode lasting less than 60 minutes. So you need to be in atrial fibrillation for at least 60 minutes to get the irregular rhythm warning. You heard a little about the Apple Heart Study. One little anecdote that I will share with you is while they signed up this number of patients, they actually were quite slow in releasing the study. They weren't sure how many people would sign up. So Apple started releasing the study, enrollment was going well. They said, be ready for more patients. They sent out one email to Apple Watch users announcing the study, and in a single 24-hour period, they signed up 250,000 people in one day. Now we were all concerned that the release of this product would inundate us with false positives. Now the Apple Heart Study was, remember, very problematic because it used a version of the watch that didn't have the ECG capability. So in the study, if you got an irregular rhythm notification on Friday, you maybe got the patch two weeks later. So the fact that the patch didn't show atrial fibrillation doesn't mean that you didn't have atrial fibrillation two weeks before on that Friday. It just means that they weren't done at the same time. So even with that caveat, the positive predictive value was 0.84, which is pretty impressive. More important data is that the percent notified in young people, presumably without AFib, was less than 0.2%. And indeed, most of these notifications were for things like frequent PACs and PVCs. So it wasn't nothing. And then as you might expect, in older individuals, the percent notified was higher. Again, a relief that we're not going to be inundated with very high false positives rates. So briefly, where does mHealth play into the future of EP? I think clearly as a screening tool, easy, patient management, I showed you some examples and research. There are studies that use these devices for screening. And sure enough, the screening population, higher yield and routine care. I'm on the steering committee of what will be or what is the largest randomized trial in the history of cardiology, released just about a week ago. This is called the Heartline Study, sponsored by J&J and Apple. It will randomize 130,000 people over the age of 65 to either get an Apple Watch or not. The purpose is, can we find atrial fibrillation and appropriately anticoagulate those patients who need it and prevent strokes before they occur, given the knowledge that about a quarter of all atrial fibrillated strokes, the stroke is the first manifestation of the disease. And as you might expect, enrollment, from what I understand, is going quite well. Here's a study that we've been actually working on that is hopefully something you'll be hearing about in the not too distant future. It's currently winding its way through the NIH. But it gets into some of the issues we were talking about. If we understand that in some individuals with infrequent episodes of atrial fibrillation and low cardiovascular risk score, can we target anticoagulation only during what we believe to be a high-risk period? Can you use a wearable device to take your anticoagulation only in response to a prolonged AF episode and therefore spare yourself the exposure to an anticoagulant during long periods of sinus rhythm that you've either achieved on your own or through medications or ablation? So the study is called REACT-AF, and all I could say right now is stay tuned. I think the other big advantage of mHealth is in the way we do research. Just like the Apple Heart Study, just like the Heartline Study, we can now do sightless enrollment. We could also do virtual follow-up. Right? Why come back into the hospital to fill out a paper form about what's gone on over the last six months when you could actually get your medical records on your smartphone? And there's a process called geofencing where we can program your phone to let the study center know when you have been in a healthcare facility more than 15 minutes. So we can know whether you are visiting a loved one or having surgery and automatically get that information from that hospital encounter to see if you've had an event. And I think that this will really streamline research and whether it can reduce costs and improve efficiency, I think we need to still study. I think there are, excuse me, limitations of mHealth in the EP space. As mentioned previously, the data is coming from all sources. We don't know how we get this into the medical record. We're not clear how we monetize this. We're not clear how we defend ourselves. Right? That patient who sends me that ECG, you know, late at night on a Friday, and perhaps I don't see it until Monday. Right? Or that ILR that shows AFib on a Friday and the patient has a stroke on Sunday. What is the exposure now that this information is pouring into us? And then lastly, how do we remotely evaluate patients and treat patients when we just have an EKG? Do we need to go into the medical records no matter where in the world we are traveling? Right? Where we're on vacation and make informed decisions? I think all of this are questions we don't have the answers to, but the technology is already here and has outstrept our ability to sort of integrate it. But we will be forced to do so immediately in the not-too-distant future. So I've gone over in time a little, but I thank you very much for your attention. And now we'll open up to any questions. Otherwise, I'll see you outside. Thank you.
Video Summary
mHealth, or mobile health, refers to the intersection of healthcare and digital technologies that collect, manipulate, and share healthcare data. There are three categories of mHealth: doctor-facing, patient-physician communication, and patient or consumer-facing technologies. Patients with implantable devices want to know about their arrhythmias and changes in their health status, while those using wearables want to know how to share the data with their healthcare team. Physicians believe that patients should have access to information about their own bodies, but there is a debate on how much information they should receive. The use of wearables, such as smartwatches, has greatly improved the monitoring of AFib. The Apple Watch, for example, can passively monitor a person's heart rate and alert them if there is an irregular rhythm. Studies have shown the accuracy and effectiveness of using wearables for AFib detection. However, there are still challenges in integrating mHealth data into medical records and establishing protocols for remote patient evaluation and treatment.
Asset Caption
Rod S. Passman, MD, FHRS, Northwestern University, Chicago, IL
Keywords
mHealth
digital technologies
wearables
AFib detection
remote patient evaluation
medical records
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