good morning everyone it's nice to see you and I am so delighted to be here and introducing Arun shradhar um and so happy to see you come back and visit um a room was born in Chiang Mai India and did his mental degree training at the Gandhi Medical Center and then he immigrated here to the United States to go to the University of Alabama for a master's in epidemiology I followed that with training in Internal Medicine at Georgetown and then went to University of Kansas for his cardiovascular disease fellowship and I think that's where he really fell in love with EP working for a really wonderful group doing great research he went to the Cleveland Clinic for EP fellowship and I remember in 2017 when we were working so hard to recruit him here talking to my friends at Cleveland Clinic and um you know Arun he just seemed so humble and um just so kind and and in initial presentation really just laid back and at the Cleveland Clinic they were like this guy is so brilliant and they couldn't stop talking about how great he was in the lab with his technical skills and um how he scored the highest on this and training exams and he was great with patients and great with teaching but really really great with research and really Innovative really thinking about novel ways that we could develop new tools to make obligation and safer in the EP lab and um and I do want to highlight something that I thought was really interesting about everyone's work which was during the pandemic so the pandemic starts and I remember I don't know about all of you but um you know I was really focused on how do I find the PPE how do I put it on when do I gel my hands and and everyone figured that all out he managed to continue to grow the EP program at Harborview despite some lack of resources and um and also managed to publish 18 papers in the past couple of years work on five consensus documents and a guideline and this research work was a lot of it was things you would expect from his CV things like what are the outcomes of patients undergoing afib ablation and hypertrophic cardiomopathy how can we develop a new tool to protect um patients get undergoing am ablation from atro esophageal fistulas um but also some really novel and Innovative stuff like how do you do a totally remote trial looking at QT intervals of patients treated with hydroxychloroquine using ECG technology so that patients don't have to come into the clinic during a pandemic and how do you use Smart speakers contactlessly to measure patient's heart rhythms and I I really think that it was the pandemic that really sparked his interest in wearable Technologies and the promise of artificial intelligence to use all this physiologic data to really solve some intracting problems we're having in delivering credit their care and I would be remiss I don't want to go on and on if I did not mention that these projects um that he does have almost universally been with trainees um both cardiovascular fellows from our own program as well as bioengineers some of which who won awards for the work that they've done with Arun and even medical students which I think says a lot about who he is so I am really excited to learn from him and hear his talk from hype to reality navigating the limitations and challenges of wearable Technologies in cardiac arrhythmia care welcome thank you so much Chris for the extremely kind introduction uh I have to say like Chris has been one of the greatest mentors that I've had like including like all the all the colleagues that I've had like have been incredible are queued up uh nazam Genie they have an incredible mentors thank you so much for everything so um let me start with disclosures so I'm actually on the Advisory board for our live course so there is some data that I'm going to show which which has to include our live core because it is one of the uh one of the Technologies which which works in this space but nothing I showed today will be actually promoting or advocating any of the products um so I I thought I'll start with the brief history on the variable Technologies so it's surprising to like when I was looking at variable Technologies and the history I was actually surprised to find that the history of wearable technology starts back in 1700s so uh Lewis parallett he was one of the uh Swiss uh watchmakers so he uh created an automatic self-finding watch in 1777 which uh which actually like uses body movements to regenerate the energy and never has to be like wound so then he went out with all of that technology further and he created the first pedometer in like 1780 and this was a huge hit a lot of aristocrats started started using this speedometer in in 1800s and a lot of this uh um rich people started using this to measure their step count and keep themselves healthy and if you look at the ECG history of ECG Technologies and tied with variable Technologies like you can see that the first parameter started in 1780 and in 1900s is when the first ECGs came into clinical views uh Anthony created the first clinical ECG machine and from then on like there has been a lot of developments in ECG technology over the last century and the wearable Technologies really started becoming more and more like sophisticated in the in the 21st century so 1960s and 1980s it was the time of even monitors and halters and then in 1990s there are ECG monitors which with uh which could actually speak to the wireless handheld devices and there was like Wireless transmission and in 2000s the first patch monitor was created like the zero patch and then Fitbit and Apple watch came in 2000s and in 2014 in the first mobile ECG device got FDA approved and that was a live call cardio mobile and in 2018 is when there was a big revolution Apple Watch series 4 was came about and it had an ECG technology been built into the watch and in some ways I do think that this was actually the beginning of their wearable Revolution so during this period of consumer mobile Health Revolution um cardiac arrhythmia field was also growing up like in exponentially and it found itself in a very unique situation so um like wearable technology started incorporating ECGs and heart rate monitors and cardiac field was very well suited to harness this and create tools to monitor arrhythmias and uh and you know like manage arithms so there were multiple large-scale investigations done in 2010s and uh um like many of these were company company driven but these kind of started the started the revolution in terms of like using wearable Technologies for clinical care uh in this talk I was going to focus on the limitations and challenges of the the Technologies rather than the glory of the technology because we keep hearing about the glory a lot but I think it is very important to take the implementations and challenges into consideration if you really want to make use of it in a clinical sense and um so most like I'm going to focus a lot on the technical considerations the accuracy of these devices and what hampers the accuracy of these devices then the clinical applicability of these devices so okay so we can detect arrhythmia but can we really make a difference in patients life so we're going to look at some hard data so most of the data on technical considerations and clinical applicability belongs to belongs in the role of atrial fibrillation and in a lot of ways the field of wearables has focused on afib for a lot for a long time and it is it's actually going Beyond afib right now but but but still it is very innocent most of the data most of the meat of the data is on atrial fibrillation then look at some of the other arrhythmias and see if we can differentiate other arabias then I was going to focus on some psychosocial considerations of wearables and then like some impact on Healthcare burden this is by no means a comprehensive and exhaustive list of limitations there are lots of other limitations but I decided to focus on this because this is uh like going Beyond this I wouldn't be able to focus on the specific details of some of these issues so let's start with the patient anecdote so this is a 59 year old man um he's relatively healthy he has like hypertension but otherwise no other heart diseases his wife gives him an Apple Watch for his birthday four months after he starts noticing multiple notifications of a regular heart rate he notifies this primary care physician then hold her finds recurrent peroxisome atrial fibrillation he's immediately put on anticoagulation and rate control and is referred to a Cardiologist and uh and he's undergoing like uh evaluation for advanced therapies and see if he's a candidate for ablation and things like that so let's contrast that with another story so this is a 66 year old retired physician he gets an Apple Watch and monitor uses it for fitness and gets notifications of your regular heart rhythm and visits first one cardiologist and then a second cardiologist he gets three holders three event recorders and then a loop recorder two years in no atrial fibrillation is found only occasional Pacs and short 80 episodes so these two cases illustrate the contrast between the variable devices like the the outcomes available devices in the first case you can clearly see that the patient benefited and potentially we you might have been saved from a stroke uh we don't know but the second physician is a second patient who was a physician he really got hurt by the wearable device so he got like an extensive workup he got a loop recorder which is a which is invasive procedure all the small procedure he got a loop recorder and nothing was found except and and what the the eventual outcome was like a lot of healthcare expenditure and a lot of anxiety but we didn't find it happen so to make sense of this we need to make sense of the sensor Technologies and their limitations so I'll go over few technologies that exist in the current world and let's see like um what are the limitations of the current Technologies so if you broadly classify the variable Technologies it's basically cardiac electrical activity sensors and non-electrical activity sensors so electrical activity sensors are easy to understand for for us as cardiologists like these are basically like any ECG device handheld or a 1280 cge or or any any other kind of ECG device which measures the electrical activity and when you look at non-electrical activity sensors these are cardiac sensors which measure the heart rate and heart rhythm but they depend on some kind of a mechanical impulse to do that the non-electrical activity sensors can be of two major types so like you can either have an arterial impulse sulcers which which uh which sends the peripheral arteries or the central arteries and figure out how many times the heart is beating like how fast it's beating and how slow it's beating the most common devices which do that are Optical PPG devices and then you have like a laser photo platysmography devices then you can also have like blood pressure monitors which do the same thing and there is something called video cardiography which can actually look at the arterial impulse directly so then you have the cardiac impulse sensors the cardiac Empire sensors are sensors which actually detect the impulse of the cardiac motion directly so they look at they focus on the apical motion and there are different ways it uses different ways to pick up the apical motion of the heart directly so these can be thrown out based hot motion sensors radar based hot motion sensors and then there are other Technologies like gyrocardiography and seismo cardiography and ballast or cardiograms we barely use those things in clinical clinical world because these are usually experimental and research devices and the accuracy is not that great so um I will come I will touch on the accuracy of these devices a little bit later in my talk again but I want to contrast that with the arterial impulse sensors these are usually more accurate but you can see that these are more intrusive they have to be on the patient's finger or they have to touch and most of these devices on the other side are are basically a non-contact kind of sensium devices so when you look at PPG sensors again there are two different types that are invoke that are mostly used in the Technologies and many of the mobile devices that we currently use have two different PPG sensors in built in them one is the infrared PPG so this is the red light that you get which which chronically measures the heart rate when you wear it and this is a low energy light and it is prone to motion artifacts but it uses less battery then there is a there is a second technology called reflected green light PPG it is it is basically a green light laser it has a higher Fidelity but it uses a lot of battery so it is not always on so there have been multiple studies conducted comparing the green light PPG versus infrared PPG and almost all almost all the studies show that green light PPG is much better in picking it up with the sensitivity and specificity both are very good for green light compared to Red Light PPG the disadvantage of Greenland tpg is that uses a lot of batteries so you can't keep it on on your Apple watch or on your wearable so the green light PPG usually kicks in when you start exercising and you you tell the watch that you are actually exercising so it starts measuring it more accurately so that's how the mobile mobile device manufacturers like the console battery by not keeping it on all the time so why PPG when you can have an ECG right so like like now the Apple watch 4 4 and onwards has ECG and you have so many mobile ECG sensors so why care about PPG anymore the the ppga technology has like major advantages and that's why they use it in variables so one of the first first big Advantage is that it is easily integrable into any variable I'm not talking about just the Apple watch but you can also incorporate it in like a best there is like a textile variables and then there is like sneakers you can put it anywhere and then it's pretty non-inclusive and you can monitor patients heart rate and Rhythm without patient engagement the patient can wear it and forget it a patient or a participant or a user so they can wear it and forget it and they can still be getting sampled and the heart rate can be monitored and PPG technology also allows for continuous sampling because it doesn't need the patient engagement so PPG can monitor passively long term with minimal user engagement and that's why it's so popular among the wearable Technologies so the problems with EPG so pulse deficit so when you have a fast heart rate or when you have PVCs your heart impulse doesn't correspond to the arterial impulses so you can have what is called as a bratisfygnia where your arterial impulse lags behind the heart rate and PPG can get full like it can underestimate the heart rate and or it can actually like have a false negative for atrial fibrillation when you have tachycardia then skin tone affects accuracy this has gotten better over the last few years with the laser PPG and like the green light PPG but in general darker skin tone has lesser accuracy compared to lighter skin tone so if you take African Americans or like the darker the skin tone actually the accuracy actually falls the other issue is like PPG requires longer episodes longer episodes of atrial fibrillation for detection so if you have a 30 second afib or one minute afib it won't be able to pick it up it has to last for two or three minutes before it can confirm it reconfirm it and say it is atrial fibrillation so bottom line PPG can be temperamental and almost always all the technologies that we use they say you should compare it with an ECG once you come once you have a if your alert on on PPG okay so why atrial fibrillation why is there such a big focus on atrial fibrillation from from the mobile world like from the from the wearable device manufacturers there are good reasons like one is like high prevalence uh like there is a high prevalence of atrial fibrillation among the U.S population and the world population in U.S afib affects about like 8 million people it can be asymptomatic so picking it up early has value and it has a significant morbidity risk but I think the most important reason the device manufacturers are focused on atrial fibrillation is because it is a low hanging fruit it is very easy to pick up using diagram using PPG as compared to some of the other rhythms which actually hurt the heart like hurt the patient and kill the patient like ventricular tachycardia so it atrial fibrillation is a bit more easier to diagnose with PPG and um with irregular heart rhythm so and it is actually um more common so you can get more customer base and you can create you can create a better user base for for ppgs so that is actually something very important that the device manufacturers rely on so let's go to the atrial fibrillation detection studies so these are the three big studies that have happened in the last uh one uh last few years like last decade but mostly in the last five or six years so Apple heart study and Huawei heart study and Fitbit heart study all three of them were device uh manufacturer funded and all of them use different devices so basically they have a similar uh workflow they all use a mobile based participation request and they recruit the patients based on mobile alerts and then once the patients once the participants are recruited the monitoring is through PPG and once they have afib alerts then they ask them to get a confirmatory ECG monitor or or a patch monitor or or some other kind of ECG confirmation so all three of them very very successful and showing that atrial fibrillation can be detected using mobile variables they were also very successful in showing another another big issue so the feasibility of mobile recruitment uh Apple study was one of the first studies we showed that you can recruit 420 000 patients just by using mobile alerts and like like you can use a mobile mobile technologies to gather recruitment so this was important and it will come in some of the other later studies that are ongoing and some of the future studies that are being planned so the positive predictive value of these studies were anywhere from 84 to 98 but take that with a grain of salt um the the reason is because the compliance with the patch monitor after PPG detection of atrial fibrillation was not ideal about 35 to 62 percent of the people followed through with a patch monitor so the denominator is is the people who actually follow through with the patch monitor not the people who did not follow through so when you actually like uh follow through with the patch monitor like the people who are more likely to follow through with attach monitor are those people who have multiple alerts like if they have like multiple alerts of uh like a PPG detected afib they are more likely to follow up with a patch monitor so that affects the positive predictor value so all in all the the studies did show that it is it is possible to pick up atrial fibrillation the major issues are it is not randomized there were no control arms and it was heavily biased towards younger and healthier population for example um the Huawei study um the mean age of participants was 35 years and in Apple watch study um only 5.9 what only six percent of the population was a cohort was beyond 65 years of age so it is heavily biased towards younger and healthier population and the most important issue is that the impact on clinical outcomes is not clear from from these studies so so if you do manage to find more afib does that mean anything like like if you put them in anticoagulation earlier does it actually result in like a hard clinical outcome you're talking about heart failure and stroke and things like that does it actually make a difference so there are some other small studies which have been done for similar Technologies like TeleCheck AF study rehearse AF study and vital day of study most of these studies like like Chris was mentioning happened in a covet like covet kicked off a surge of wearable Technologies and using them for different reasons so um study was based on a 60 BCG platform and here's the app Study was based on single adcg and why delay of study was either with a device or a physical exam they they actually randomized and they actually showed that you have similar rates of detection if you do just a physical exam on patient bottom line is like all these studies show that it is it is possible to pick up afib uh but the big question is does screening translate to better outcomes this has been studied in a couple of Trials and the the answer is not out yet it's ambiguous still when in 2021 there was a stroke stop studying they used they randomized the patients to two groups handheld ECG screening versus control group which was routine care and they did intermittent ECGs for 14 days they saw that they were able to diagnose three percent of the patients with atrial fibrillation on the handheld ECG group and almost zero percent on the other group control group on a five-year follow-up there was a very small but significant statistically significant but questionable clinical relevance in the combined endpoint like 5.48 versus 5.68 the combined endpoint was stroke and heart failure if you take the loop trial Loop trial was a trial which randomized patients to ilr versus usual care and they showed that when you place a loop recorder you can have a like you they caught more airfare so 31 percent of the patients in the ilr group had a FIP and twelve percent in the control group but there was an anticoagulation was initiated as soon as if it was diagnosed but there was no significant reduction in stroke or systemic arterial embolism so the word it is not out yet so there is actually a there is actually a big there are some big ongoing studies which are still looking at this question so this treatment of incidentally diagnosed afib improve clinical outcomes based on different Technologies rtci is a study that was happening in UW like Universe recruiting for it uh Placer and ICD detected Affair so small amounts of afib detected on facer and ICD does it if you anticoagulated anticoagulate versus anti don't anticoagulate does it make a difference then Heartland AF uh is a study based on Apple watch um so they are looking at um like a real real world endpoints like stroke and uh like a embolic episodes and other arterial embolisms and Liverpool Huawei study is using Huawei watch do the same thing and then there is a safer study which is using a singularity CG to do the same thing bottom line is all these studies are looking at detection to stroke impact uh like when you do earlier detection does it make any foreign what are the issues with screening why does screening fail so the problem is like uh screening works if if patients are compliant there are two big issues so duration of monitoring is very important in afib detection and compliance with with wearing is also very important so if you have a if you have a condition which is very paroxysmal it's paroxysmal and it is rare then the amount of time you can monitor increases the chances of detection of atrial fibrillation now continuous monitoring versus intermittent monitoring the continuous monitoring will always succeed compliance with that is also important if you have a patient who wears this Apple watch for a whole day and then like puts it on charge in the night and has a vaguely mediated effect in the night you're not going to pick it up we're going to lose that detection because he he removes it for me for for most important part of his dream the conundrum of mobile technologies is that the more intrusive the technology the more accurate it is but it is less less the patient compliance is less so like ECG Technologies are obviously more more um accurate but the patients have to engage with it and it doesn't it doesn't help with like you know continuous monitoring the less inclusive the technology the more compliant it is but it is less accurate so if you look at our study which we did with smart speakers to contactlessly monitor the heart rhythms it was an extremely non-intrusive study so you can just have the patient in the room and like you can you can detect it but the problem is it it starts failing if the patient is like more than seven feet away so it is a less intrusive study but it is much less accurate compared to a study which would be more intrusive and needs patient engagement so that that problem uh exists for all sorts of technology which are which are less intrusive and mobile world the sensitivity and specificity of any um any of the current Technologies vary between 68 to 100 depending on the patient compliance so the question was can we use uh can we use any sophisticated algorithms to increase the diagnostic yield of of ECGs so there were two uh there was a big Landmark study done by the Mayo group uh where they looked at uh sinuses and ECGs of patients who had like peroxisomeletrial fibrillation detected at any point in their life you're not taking the afibcg you're taking the sinus rhythm ECG and you're trying to see if there are any signatures of atrial fibrillation that can be detected using artificial intelligence algorithms and they claim that they were able to find um atrial fibrillation patients purely using sinusoid them ECGs and they're also subsequent study where they showed that there is a five-fold increase in detection of afib compared to standard monitoring if you use AI enhanced ECG detection we try to replicate the same um same study using mobile ECGs and we used alive core ECG 73 000 patients and 260 000 alive core ECGs and we showed that we can also um predict events but but the accuracy was much lower compared to the Mayo grouping um so and uh like we showed that like we can predict atrial fibrillation occurrence retrospectively as well as prospectively like if you have a sinuses in ECG today and if the patient is going to have any A5 episode two days from now we can actually detect it there is a higher incidence that we can detect it or it can also predict if that occurred like two days prior so the problem is the accuracy was not great the mayor group had a very high level of accuracy but that has not been replicated by other groups which have tried to do this in what is the current status on screening so USB stf currently says that the evidence is insufficient to assess the benefits and harms of AF screening and they do not recommend AF screening on the other hand the European guidelines are slightly more favorable but they don't recommend wearable or routine ECG used for screening they say that internet and ECG screening for patients on an opportunistic basis is is favorable compared to like routine screening so if a patient above greater than 65 years of age comes to your clinic get an ECG from him get an ECG for him and do a pulse file patient or physical exam but don't go and find these people who are a boy age 65 not like colonoscopy so where you start recommending screening once they start like once they come to a certain age they say they don't go on screen them do it on an opportunistic patient if they actually come to the clinic so let's look at a different scenario the pivoting from screening to management so this is a 55 year old very athletic commercial pilot with porosis by neutral fibrillation his status post ablation and doing well he prefers minimizing inpatient in-person clinic visits and Clinic checks and occasionally has calculations uses portable ECG and makes televisit as and when needed for abnormal ECGs only so does this work so management of patients with no atrial fibrillation is probably the most likely immediate impact for wearable Technologies although most of these Studies have focused on AF detection and AF screening although we although there are no formal guidelines like on an online survey by done by a patient group it is noted that 71 percent of the respondents like they said like they were already using some kind of digital Health device to manage their atrial fibrillation so the patients are already doing it now this could be like a completely biased survey because like they were they sub-selected the patients who are engaging with the survey but but there are patients who are already using it regardless of whether our regulatory bodies have said that it is okay to use it or not I do think that mobile platforms have enabled modern novel patient directed algorithms where the power is with the patients right now and you can have like on device analysis of ECG the patients can get an ECG and they can get an immediate report of whether it is afib or not or you can have a cloud-based analysis using AI algorithms to say whether it is afib or some other condition and there is also um there are also like vendors coming up with third-party like clinician analysis of ECGs so it could be you could be the doctor you you could have a patient and the patient could be sitting at home but they do an ECG and it goes to a third party and they say that there's a clinician looking at that ECG and saying that you have afib do you want a consolidation and they actually offer a consultation so that is also happening so there are different models of uh um if a management happening already regardless of the um like um regardless of the guidelines and it is a matter of it is it's a matter of us to catch up with these things and figure out what is the best uh best kind of technology that you can use in your clinic and whether it is appropriate for your patient or not so when you look at management of non atrial fibrillation there are four different things that uh that are um that are the major issues the first one is symptom correlation when you have a patient with afib and you have managed them you can use these mobile ECGs to figure out if the treatment is working well and you can also use it to titrate a date and Rhythm control there was a study which was done there's actually a TeleCheck study and then there is a televost study which is going to come up um this is a proposal which is uh where they are trying to implement the study but uh it's it's in Europe so the concept is that if you have a patient who has atrial fibrillation and who comes to the Ed instead of cardioverting him give them a device and send them home and then like put them on rate control and Rhythm control and try to manage it remotely if they don't get card averted they can come to the EDM data class 1C pharmaceutical agent and get um get cardioverted using a antiarismic agent so basically using remote remote monitoring to reduce the cost of healthcare by preventing them from coming to the Ed when you use uh when you use these Technologies to try trade medications you need to make sure that your patient selection is good like the patients have to be facile with available technology and they should be comfortable with their medications and they should be able to understand their medications on how to retreat them and the patient should be able to understand the implications of false negatives and false positives so the second issue is correlation with risk factors so more and more Technologies these days are incorporating sleep detection and exercise detection so the technologies have become very good at sleep sensing um like so much so that they can not only detect when you sleep and when you wake up but they can also detect your oxygen saturations and correlate it with your apnea episodes so none of them are FDA approved yet for Osa monitoring and oversee detection but I think it's it's in the Horizon so like you should we should be able to see it in the next couple of years Osa monitoring on Osa diagnosis so Osa has a big impact on afib and if you're able to correlate Osa with afib by time points you could probably tell the patient um how to impact a safe like by by giving him a CPAP and making sure he is responding to the CPAP so the next issue is burden assessment so can we do burden-based anticoagulation this is a big um this is a big uh question so let me uh let me illustrate a clinical scenario so there's a 48 year old patient with atrial fibrillation status post oblation he's doing great uh he's an extreme sports athlete he's a ski jumper it's a very high risk of Falls and very high risk of bleeding he has a chance to ask of two but he wants to use Eliquis on a PRM basis he wants to use it only when he has a fit and not all the time and he uses his Apple watch to monitor his AFib I think we've all met this patient like we all met this patient who wants to like take his Eliquis as and when needed so does that work so intermittent fill in Pocket anticoagulation the concept of as needed anticoagulation is not new so this has been tested before there have been three feasibility studies at least as far as I know which have used different Technologies to see if they can feasibly do as and when needed anti-coagulation so react com was a loop recorder initiated study where they looked at 59 patients and saw the feasibility of acetated anticoagulation versus a continuous anticoagulation tactic AF used Pacer and defibrillator detected afib and eye caref was daily ECG is mobile based Apple watch ECGs and they were they looked at it so all these studies were small like I don't think you can make any big clinical um uh you can't make any clinical decisions based on these studies but the last study of note did it was a 20 month follow-up and it showed that the GI B was higher in the continuous anticoagulation group compared to intermittent and anticoagulation group but there is a big study that is going to start so react Afra so on YouTube is part of part of this trial um and become my current Institute policy is also a part of this trial so react AF trial is an Apple Watch based study it's an ambitious project it is funded by both NIH as well as apple and the the goal is to recruit 5000 patients and follow them over a seven year period and randomize them to standard of care versus Precision targeted brief anticoagulation using Apple watch so the concept is you have a patient with a child's mask of two or more with afib you give them an Apple watch ask them to be off the back and they keep monitoring their Apple watch the day they have affair they start taking their duvac and continue the do act for 30 days and if they don't have any further a few episodes they stop the duac so it is brief anticoagulation regiments based on Apple watch monitor so we'll see how that goes there's a lot of money going into the study and if it if it turns out to be a positive study I think it's going to be a game changer for a lot of if patients so the last thing I wanted to touch was drug loading so we use a lot of drugs for atrial fibrillation and um the most powerful drugs are also the ones with the most side effects and that includes certain oil and doped light so QT monitoring is a very important thing for uh sort of law on difficult and most of the centers in the country use inpatient drug loading where they measure cuties on on a bi-daily basis like two times a day and the question is can we use mobile devices for outpatient workloading and keep them home and reduce the cost of healthcare so we need to answer two questions can mobile instances actually measure QTC so that is the first question you need to answer so there have been multiple studies all during the pandemic uh where they have compared mobile acgs to 1280 CGS and they showed that it is actually like equivalent like you can actually get a pretty accurate QTC recording using a mobile ECG and it is pretty comparable in uh like accuracy to uh to a 1280cg this was our trial that we we uh ran during covet there where we used mobile ECG monitors for QT measurement and we used it with mayo Mayo was the ECG core and they were they were like measuring the cuties for us but we were using it through a mobile device and we showed the feasibility so the second important question is okay so you can measure QT using a mobile ECG but is it safe to keep the patients home and monitor their cuties what happens if their QT is long do they do you bring them back immediately or you wait for the next dose how safe is it so there have been very few studies and very extremely small studies for social all loading you can see like three patients with sort of all and 12 patients or all and they showed that it is safe to use them nobody has been daring enough to do it on the feet light and as we all know the fitted light is is much more of a QT prolonging agent compared to sort of law so I don't think this is prime time yet so but there are a few studies in Horizon um and they're trying to build up some studies to see if they can do the fit like monitoring at home okay let's go beyond sinus for them and atrial fibrillation for a second and look at other rhythms so this is the patient this is actually an interesting story so 38 year old patient with chest pressure in a transatlantic flight and they are talking about like diverting the flight and all that stuff and they're reaching the nearest airport there happens to be an ER doctor on board who uses an Apple Watch to check the Rhythm so this is the Rhythm you can see it is an SVT and you do have some St depressions so he reassures the patient and asks him to do like multiple wall servers a co-passenger shares is not a problem and he takes that and then he converts to sinus so clearly you can see that it was SVT and Apple watch picked up the SVT so this is another patient with this is a published report so patient with Recon palpitations and dizziness so picked up torsars on Apple watch ECG monitoring and this is another patient who picked up BT so you can clearly see that there is a like our PR dissociation and it was creativity so where are we with uh Beyond atrial fibrillation where are we with SVT and VT discrimination there are case reports on case series there are social media accounts but these are difficult to systematically study because of the Rarity of the condition and inability of patient to engage with these devices when you are in one of these bad tachycardias um but if you but there are studies which have been done using retrospective um ECGs no there are not studies these are like um company algorithms which are generated using retrospective ECGs and you can like the the technology to recognize why complex uh wide complex rhythms is that like the ability to recognize tachycardia and radicardia yes we have that the ability to discriminate SVT from VT we don't have any algorithm that can truly do that using a mobile platform ECG but I think this is where again like AI can probably help and in the this probably is not so far out so far out in the future like we should be able to get SVT and VT discrimination from mobile platform ECGs in the near future so I'm going to Pivot and talk about a completely different topic here a slightly different topic actually not not completely so psychosocial impacts of wearable monitoring so how does it impact the patient and how does it impact the relationship of the patient with the physician okay let's look at another question and I told so 32 year old female with Osa BMI of 32 she gets diagnosed with hypertension she gets a she gets a wake-up call she decides to change lifestyle buys a Fitbit joins a runner's Club picks up running and trains while watching her heart rates successfully sheds 35 pounds in a year and completes her first half marathon next year he's weaned off anti-hypertensives and for her own saying she's in the best shape of her life so 28 year old physician assistant student persistently elevated heart rates on her Apple watch goes to the PCP undergoes extensive workup including halter even monitor till table test referral to a Cardiologist for inappropriate sinus tachycardia ultimately started on beta blocker and then evaporated also incidentally diagnosed of anxiety during this time and he started on anti-anxiety therapy so the first example is a great example of everything good about a available device where it motivates you on like where it makes you change your lifestyle positively and the second example is how it can initiate or or perpetuate anxiety in certain users so the psychological impact of consumer variables uh the recognition of both positive and negative psychological effects of utilizing variables is very important both for a clinician when we are recommending a variable as well as for the patient when they go on by one um but there are no not a lot of big studies so there are a few studies though the small studies and uh but but we need large scale studies so the first issue is station engagement so there is clear data that wearables and the gamification of of these um these variables like encourage and incentivize healthy behaviors the variables also give individualized risk scores and they provide personalized patient education and support and this helps with patient engagement on the other side you have the somatic three occupation phenomena so some people are like exhibit what is called as a worried well phenomena they're fine but they keep focusing on their symptoms and they measure their heart rates often and they end up with pathological anxiety so what leads to the first kind of patient what leads to the second kind of patient it needs further study and we we don't really understand but there might be a component of underlying anxiety which perpetuates once you start wearing a wearable the second issue so connection to the healthcare provider so there is a perceived connectedness when you wear a wearable and this has been studied in in small studies and the patients get get a sense of being connected to the healthcare providers and it enhances the patient to provide a relationship and it might actually increase the sense of safety for the patient under satisfaction with care because they feel that there is some kind of a backup for them on the flip side it can also lead to unnecessary and frequent contact with the medical system like you can have patients who are worried well and like who go into the anxiety Loop multiple phone calls multiple messages unscheduled patient visits and uh for for small and normal fluctuations in heart rate this creates an adverse relationship with Healthcare Providers and as much as we um claim to be impartial like there is always a psychological impact on us like when we deal with them such scenarios um the third issue is real-time feedback so feedback from variable sensors and biosensors is immediate and that might actually improve adherence so if a patient notices that he starts wearing his CPAP and next day he does have he has less burden of a web that improves adherence and it it also uh like it also gives the opportunity for self-management like rate control medications you can get like immediate feedback and then you can see you can control Untitled medications on the flip side of that you have false reassurance issue so if you have insufficient education among the patient group they might feel that they don't have fa but they could have afib but it could be a false negative and that leads to a false sense of security and safety and they might actually miss their anticoagulation dose like like in that instance of fill in the pocket anti-coagulation strategy so um if the if the afib is not picked up due to whatever reason on the PPG that can lead to a false sense of security and a real hazard so recognition of positive and uh negative psychological effects is very important large-scale studies are required and hopefully we should we should see some big studies come up in the next few years because this is being recognized more and more in the uh not only in the cardiovascular world but also in the psychology uh world of psychology so I'm going to touch on burden on healthcare system for a little bit so this is actually a very important issue because this this uh needs to directly to clinician burnout so let me give you an example clinical scenario so 72 your neurologist AF studies post ablation very worried about his risk of stroke monitors his rhythm with Apple watch and cardiac monitor daily sends multiple ECG strips weekly to confirm normalcy leads to adjudication by nurse then by an MD each time leads to a clinical action such as patient phone calls or secure messaging then clinic visits and documentation burden hi this is actually one of my real patients and I had to have the patient come to the clinic and I told him like you know this is what is happening on on the back end like when you send these ECGs he was very sweet patient and he immediately understood what like what I was saying and like he said like he'll do his best to cut down on the number of his ideas that he sends and but this is not an uncommon scenario this happens quite often because patients don't realize what kind of a burden it creates for the Healthcare System the biggest issue is the biggest issue is the false positives so when you have a patient with very small burden of atrial fibrillation the amount of false positives is always going to exceed the true positives so there was a study there are studies which have shown that like when you take a patient when you take a population with low prevalence of atrial fibrillation and put like wearable monitors of them you're going to have more false positives than true positives and that is just the nature of the problem so if you if you have a patient with high risk High degree of afib it's not used it's not useful in that patient because he needs to be on pills and medications anyway and if you take a patient with low risk of you're going to have false positives then inconsistency between the device grants that is a big issue because a lot of device brands use their own proprietary algorithms like sensitivity and specificity differ among the different brands and that leads to issues then electronic medical record integration and workflow do not exist currently they are under development but that is a big issue because we need to be able to see it and seamlessly like uh like read it and like once we we don't have that we are doing all of this work on the back end with with a poor workflow so lack of reimbursement or poor reimbursement is another issue and that needs to be addressed so this is a this is a very interesting uh um uh like news article that came about like two about three months ago so Apple watch released iOS update where they would actually detect car wrecks and they said like if you have like we can detect car wrecks and when you have a car wreck we will notify 911 immediately so like skiers in Nebraska it started detecting uh like a car wrecks because these people were skiing and coming to an upright part and it detected it as a car wreck and like started calling 911 immediately and they say that the cops were buried under an avalanche of false alarms so this is one of the issues like it goes to show how the technology companies do not understand the clinical workflow and how this impacts the clinical workflow on the Healthcare System so data Deluge is a real problem and that needs to be recognized and proactively solved we need appropriate integration and dashboard dashboards to like measure these things and like um sorting sorting needs to be done you need to have like no alert system high alert system and things like that so when we did this study during covet we had Neo do our um core lab for ECGs and they created a dashboard for QT monitoring so they created a dashboard and the dashboard worked worked in a couple of Fashions so the first was an AI screener so the AI screener would look at the ECGs and it would put the put them in different buckets like it would have a yellow bucket and a red red bucket the yellow would mean that there is a small Duty prolongation and the red bucket would mean that a clinician needs to see the see the ECG immediately so the yellow buckets would go to the ECG technician who would confirm the QTC and if the QTC was a false alarm he would actually put it put it as normal but if the red bucket red bucket directly meets physician and the Physicians would read it so we need that kind of seamless Technologies so that we can have an AI sorter which sorts out some of the issues and then brings the real problems to the clinician otherwise the clinicians are going to burn out so when you look at when if you're hoping for a mobile Health technology integrated arrhythmia care in the in the um Clinic you need integration from you need like um all these takeout stakeholders need to come to the come to the table and like contribute their uh bit so the device manufacturers need to come out and have consistent standards and develop need based Technologies not ease-based Technologies they shouldn't be developing what is easy to develop and then like um putting it on the clinicians to figure it out they should be working with the clinicians to develop need-based Technologies and they should be a responsible labeling of clinical utility you can't claim to figure out car wreck when like when it can actually have a false positive it's just keying so you need to have responsible labeling of the clinical utility for these devices when you take clinical researchers and Physicians there is a lot of responsibility on us we need to welcome the mobile technologies into clinical trials and that's how we're going to be able to assess the accuracy and the clinical utility and in a lot of ways it is our responsibility to work with the industry and tell them what is important and what is not and guide them towards achieving the right goals and then we also need to develop and design files and studies which focus on hard outcome focused evidence from a Healthcare System standpoint EHR integration should be a priority whether we want it or not we are going to face a future with variables and we are going to get a lot of information from the patients from with variables this is only going to increase as the millennial age group are like ages and start having cardiovascular problems everybody is very Savvy and they can they can use variables so we are going to get this the data Deluge is going to happen so we need to be prepared for that and we need to invest in personal training and then Pathways and infrastructure to integrate it into our EHR system we need to get ahead of it and otherwise the clinicians are going to burn out and um like is going to be a vicious cycle from a regulator on state or regulator and payer standpoint um we need to push for cross-platform standards which are equivalent like device manufacturers have reasons to uh like patent their algorithms and keep it proprietary but it doesn't help patients in the end or the or the clinicians so we need to have cross-platform standards which are Universal and we need to have faster audit Pathways like uh it should not take a long time for us to uh for The Regulators to audit these Pathways because the technology develops at a very fast pace and creating encouraging reimburse models reimbursement models is very important to encourage adoption and proper use of Technology if the reimbursement is very poor it will not get integrated into the Healthcare systems and we know that from different different scenarios in the past like for EP we are unique like we have had like multiple rounds of remote Technologies like we have had like ICT and pacemaker remote Technologies and there were multiple workflows that were created because there was a data Deluge initially like when the Pacer RPM Technologies came up and then there was an alert based monitoring where a lot of alerts would actually a lot of alerts would actually be sorted out and then like you would actually get like they sorted out alerts or there was actually like a three monthly pay pay a reimbursement model for pacemakers and defibrillators that was to encourage remote monitoring otherwise it was not going to be adopted so they needed to create a rainbows model reimbursement model to encourage option so to conclude variable devices have significant potential in arrhythmia care but the limitation should be recognized the limitation is currently are due to accuracy consistency and burden issues but the psychological impacts are also huge and studies need to be performed for a large scale studies need are needed in this in this space AI based techniques might actually help enhance the quality of care and experience for both patients as well as Physicians but this remains to be seen thank you thanks sir that was fabulous oh Dr diet check has a question so thank you very much I really should be brought up today um [Music] a lot of the data that you imported are reported as full increases that can be extremely investigated you know if you go from the lines and it's still very very rare event and and it would be I think far more informative to report data more and actually the quantities they're in full increases uh by the way if you can encourage to move in that direction then that would be great if this is a big question I have it involves um I mean API misinterpretation where I'm not in contact with Gary but but in terms of this pill in the locker at anti-coagulation wasn't there done a number of years ago that haven't just Associated stroke from episodes of atrial fibrillation detected by monitoring and if that's true for that sort of remove the rationale for that approach that is absolutely true there are Pacemaker and I seeing studies which I've looked at a correlation of atrial fibrillation and the timing of stroke the react AF actually goes against that hypothesis right like but I don't I don't know if we know enough about it so if you have if a burden of like one percent uh does it mean that like the timing of the atrial fibrillation should correlate with the stroke or one person burden actually meets a higher risk factor we do not know that so if a patient goes from one person to zero percent does it actually mean a decrease in stroke we do not know that answer either so the timing doesn't the timing doesn't correlate in in some of the studies but then there are some of the studies which show that when you do um a pill in the pocket based on certain parameters then it can actually have an impact in this these are smaller studies there is one study that I showed I think it was a react a react com which uh which used pacemakers and uh defibrillator based uh detection and it showed a smaller decrease in it did not show a decrease in stroke but it showed a decrease in bleeding already the big caution is that the patient should be able to wear it continuously over a long period of time during the day if you don't wear it in the nights then then the study becomes irrelevant right so like you need you need to have extremely high level of compliance she doesn't [Music] um I was wondering uh what do you think this is Denmark [Music] um I don't know the specific numbers amongst the treatment on those things so you're talking about incidentals which are not which are not the target of the monitoring but you find something else yeah so uh the clinical significance is not clear so um and I think like incidental pickup of those things will lead to increased expenditure and that is a concern there was a subsidy done for stroke stop study to see if these kind of incidental uh issues do lead to increased burden of Health Care and increased cost and it was actually shown that despite these pickups of incidental findings the cost of healthcare actually goes wrong if you do a monitoring using wearable Technologies but uh but I think the debate is still on on that zone last question and one thing I wondered about is sort of medically how this information is kind of being handled and where liability or responsibility is possible by I think is that an example like the students [Music] [Laughter] that's a great question yes so when you do um when you get a lot of these uh so so there is one study a very interesting study which is looked uh done on physician burnout and finish and burnout so what it says there is more the more responsible explanation is the more is going to burn out right like if you actually constantly think about not missing every single diagnosis you're going to burn out earlier because because you you will actually like look at all the data and you you have only so many hours as the other guy too right so like the problem is when you get these reports and there is a pathway for it to come through and if you create a reimbursement model for that and somebody is paying for it there is also like a reduction in the number of contacts with the medical system so that might be actually uh that might actually control the number of visits you do to the doctor or number of a number of reports you send to the physician that doesn't exist right now so when when a patient sends a report to a physician they can send as many as they want and that increases a huge burden I don't know I when I was dealing with it like I didn't have a good way of dealing with it so I don't know if uh [Music] yeah so and that is a problem for us like in like Pacemaker and ICD world too but at least we have some layers of protection in placemakers and ICD so we have groups of device nurses and like our remote monitoring companies which are doing these things and like sorting these data out before it reaches us so at least we don't get alerted for false positives thanks everyone will stick around for a few minutes if he has time thank you