Good morning, everyone, and thank you for coming to our heart failure session here. I hope you're enjoying your stay here in Orlando, Florida. I have the distinct honor and pleasure to introduce to you Dr. Andrew Wilson. He is visiting us from Brooke Army Medical Center. He is a cardiology fellow as well as an assistant professor at the Uniformed University of Health Sciences. And I think during the past couple of years, we've had to be very innovative of how to treat our heart failure population, especially with follow-up, because we couldn't really do those face-to-face interactions and we're very limited in resources. And quite frankly, this population is very, very large, and trying to get your grasp and hands around that. So how do we do that? And I wanted to share, and we all wanted to share, again, shared best practices with you of what's been actually successful. So with that, it's my proud pleasure to introduce you to Dr. Wilson. »» All right, well, good morning, everyone. And I just want to extend a sincere thank you to the ACC Quality Summit for having me here and allowing me to present as a first-time attendee. And I'd like to thank you all for your time, for coming to this session. So I'm very excited to be sharing with you the findings from our Quality Improvement Initiative titled, The Viability of Virtual Heart Failure Clinics. As a disclaimer, I would like to state that the views expressed in this presentation are my own and do not represent those of Brooke Army Medical Center, the Department of Defense, or the U.S. government. And importantly, we have no conflicts of interest with the electronic service and database called Current Health that we used during this project. So the focus of this talk is going to be on managing heart failure with reduced ejection fraction or HF-REF. And I want to take a moment to talk about why this diagnosis is so very challenging to manage. Up front, we know that there's increased mortality with HF-REF. For patients that are hospitalized with HF-REF, their estimated median survival is just over two years. Their five-year survival is just at — or sorry, their five-year estimated mortality is 75%, very staggering. In addition to that, their re-hospitalization rates over five years is about 50%. So a very morbid diagnosis for sure. The other challenge is that early follow-up is highly recommended for these patients after hospitalization. So the most current guidelines are going to tell us that within seven days of these patients being discharged from the hospital, they should be followed up either in an in-person or virtual format, which takes a lot of careful coordination to pull off. And while the strategies are well known as far as GDMT, most providers are well aware of the agents that need to be started for these patients. But the exact strategy for how to implement these agents and how to titrate them all up to their target doses is less clear. The guidelines would tell us that it's reasonable to increase the doses of guideline-directed medical therapies every one to two weeks, but that the strategy itself should be individualized to the patient. So whether or not we start two agents and increase those doses to max or start all four therapies up front, what we call quad therapy is up to the individual provider. It truly takes a village to take care of these patients as well. A multidisciplinary team serves these patients the best, including cardiology providers, nurses, clinical pharmacists, and often primary care managers, social workers, and perhaps even behavioral health experts if they have those comorbidities. And I want to spend a moment, too, to talk about how important effective self-care is for the patient. These patients have a very steep learning curve to understand their diagnosis of heart failure and the morbidity and mortality that's associated with it. While we encourage them to participate in regular exercise, to restrict their sodium intake, get their vaccinations, the patients themselves have to learn how to take care of themselves, to monitor for the signs and symptoms of decompensation, and to understand how to appropriately react to those changes. These behaviors, this self-care is really something that needs to be learned by them and must be sustained if we're to be successful. And so the challenge at our particular institution, and maybe some of you can relate to this, is that we have a clinical staff shortage that limits our ability to provide adequate and timely access to care for these patients. And what that has resulted in at our institution is that some of these patients with HFREF diagnoses are sent to the network. We lose our chance to care for these patients, and they're sent elsewhere. And as you can imagine, that negatively affects our graduate medical education mission. Those are lost learning opportunities for our fellows, for our residents, for our interns, our medical students, very much a ripple effect. And so sometimes the most challenging of times produce some of the best solutions. And if there's anything that we learned from the COVID-19 pandemic, it was that virtual health capabilities and platforms became a very sustainable model for caring for our patients. And so the idea started to materialize from our department that maybe we could use some of these same virtual health platforms for the care of heart failure patients. This resource had already been purchased by our institution and was actually going to go to waste as the COVID pandemic was coming to an end. So we figured we better use what resources we already had available. And so the idea of this virtual heart failure clinic, as it started to materialize, we had in mind creating a virtual platform managed by a multidisciplinary team. And our objectives were as follows for our virtual clinic. We wanted to, one, improve access to care for patients with HFREF. We wanted to allow for effective and safe titration of guideline-directed medical therapy. We wanted to reduce patient symptom burden, prevent heart failure-related hospitalizations, and allow for effective patient education and development of, again, that high-quality self-care for these patients. How exactly we planned on getting that done was, like I said, we formed a multidisciplinary heart failure virtual team that included three cardiologists, four nurses, one clinical pharmacist, and two fellows in training. And the way we meant to enact those objectives previously described, we wanted our clinic to be able to provide expedient and efficient follow-up for patients hospitalized with heart failure on a virtual platform and adjust that frequency of how often we were seeing them based on how high risk they were for rehospitalization. We wanted to provide the patients with a digital toolkit that could measure continuous health data that would then be relayed to us and allow us to make clinical decisions on their behalf, mainly altering the doses of their medical therapies and keeping them out of the hospital. We also wanted this same digital toolkit to have the capability to measure symptom burden for these patients, so to allow a lot of input from their perspective to guide their patient-related outcome measures. We also wanted to comply or validate our clinic based on quality metrics that are already available. So we reached to the International Consortium of Health Outcome Measures, the ACC, AHA Quality and Performance Measures. We wanted to comply with all those as a metric. And then finally we wanted to spend quite a bit of time with patient education and developing those effective self-care habits. So with that in mind, with kind of our vision statement, we started to set our sights on our target patient population, which really came from two fronts. We wanted to recruit from the inpatient side new diagnoses of heart failure that were hospitalized or those that already carried the diagnosis of heart failure and had had a recent rehospitalization within the past six months. And most of these patients were either on the primary cardiology service or being taken care of by the hospitalist service with a cardiology consultation. We wanted to also recruit patients from the ambulatory setting as well. So from our very own cardiology clinic we recruited patients with known diagnoses or new diagnoses of HFREF that were not already on kind of maximum-tolerated GDMT. And so our exact inclusion criteria for enrollment into this clinic was again HFREF with a reduced left ventricular ejection fraction of less than or equal to 40%. And we preferred these patients to have either ACC Stage B or C ischemic or non-ischemic cardiomyopathy. So we tried to really broaden the different phenotypes of heart failure included. Again these patients would either be hospitalized with acute heart failure syndrome or a repeated admission within the past six months if they already had an established diagnosis. We wanted to also take on patients in particular that there was difficulty titrating GDMT in the past due to either patient adherence issues or side effects associated with it. For our exclusion criteria we thought this was definitely important. We know that not all heart failure patients are created equal. And so if the patient had another underlying disorder that could be better served by intervention aside from just medical therapy, we did exclude those patients. So patients that had severe valvular disease that was either amenable to surgery or percutaneous intervention were excluded. Those with severe ischemic cardiomyopathy that was amenable to revascularization either surgically or percutaneously, we did not include patients with heart failure with preserved ejection fraction or infiltrative cardiomyopathies or hypertrophic cardiomyopathies. Those with uncontrolled atrial fibrillation, other similar arrhythmias we did not include, as well as those that were already on maximal tolerated medical therapy or had documented intolerance that was persistent because of side effects. If patients were already on maximally tolerated therapy or had appropriate devices placed to prevent sudden cardiac death and then not had any acute hospitalizations within the past six months, we didn't think they would benefit as much. Those with persistently depressed LVEFs despite medical therapy we did not include. And those that had severe renal impairment on hemodialysis that would prohibit us from starting many of our GDMT therapies. And then lastly, because of the virtual nature of our clinic, we did have to enroll patients that at least had some enthusiasm to learn how to use a virtual platform and have some kind of social support at home. Those that maybe lived at home alone and had no one to help them troubleshoot some of these systems were kind of less favorable. Those that were discharged to a skilled nursing facility or inpatient rehab also interfered with our clinic. And again, if you didn't have access to home internet, cellular service or telephone, which is the case in some rural areas of San Antonio where I'm from, those patients unfortunately could not be included. Or those that were very functionally disabled and could not complete ADLs. So that was a mouthful. That's all of our inclusion and exclusion criteria. How would this actually look enrolling these patients? Well it all started with a face-to-face initial encounter. And the way that we would do this is usually on the last day of the patient's hospitalization we would actually set them up with a face-to-face encounter in our clinic or shortly thereafter their discharge. And so at the initial visit we obtained consent. We exchanged contact information with the patients. We reviewed the structure of what the heart failure virtual clinic would look like. So we'd usually tell them something like, you know, plan on being enrolled in this for at least two, three months. Plan on getting calls about five times a week from our nursing staff. And obviously complying with the virtual toolkit that we gave to them. And if the patients already maybe didn't have the bandwidth for that or that was kind of beyond what they were able to accommodate, they could disenroll at that point. If they moved forward we did a detailed medication reconciliation of course to make sure there was no errors with their medication list, updated the EHR, made sure there was no harmful drug-drug interactions and moved on. We took some time to review their care plan they had received thus far. So if they had any questions about their recent hospitalization and the results of some of their tests, their diagnoses, we went over that. We collected baseline vitals, labs, things of that nature during this visit. And then very importantly we talked to them about what their goals of care from this clinic were. And we got many responses, things like they wanted to be able to attend their grandson's wedding or they wanted to be able to attend their daughter's graduation or perhaps just walk every day without having debilitating heart failure symptoms. So we kept those goals of care in mind. We set individual alarm parameters during this initial visit. So we educated the patients on alarm symptoms to watch out for, changes in their vital signs, increased heart failure symptoms that should tip them off that they may be decompensating. And we also provided for them a very comprehensive self-care educational book, which we were proud of. We used the free ACC resource CardioSmart, which provides a lot of great patient education resources. And in particular, we compiled all of the heart failure patient education and compiled it into one manual. And then most importantly, each patient left with a digital toolkit as part of their virtual clinic experience. And so what exactly did this digital toolkit look like? Well, we can see it pictured here. It came with one digital blood pressure cuff in three different sizes pictured in the upper left of the screen. It also came with a digital wearable pulse oximeter, heart rate monitor and pedometer to track continuous movement. It came with a digital scale that could weigh up to 450 pounds. And we counseled the patients that only they should be weighing themselves on this, not their friends or their kids, because that would highly confound our data. And then they were provided with one electronic tablet as well. And this was primarily used to communicate with us through video teleconferencing capabilities and also to fill out all of their patient-related surveys and questionnaires, including symptom surveys, the KCCQ, the PHQ-2, and medication diaries and such. And then a modem was included to basically sync with their Wi-Fi and relay all this information. So the way that this would work, right, is that the patient would wear 24 hours a day the pulse oximeter, heart rate monitor, but that was the only thing they wore all day. The rest of these items, like the blood pressure cuff, they only had to wear just briefly in the morning to take their blood pressure. And then the scale, of course, they would step on once a day. And as soon as they were within range of their modem, it would transmit all of that health data wirelessly. And so we had a great tech support team to support us. So as soon as the patient got home with their kit, they would be walked step-by-step through how to set this up in their home through a call from our tech support system. They would verify the equipment was functioning. If it was malfunctioning it would be replaced in a timely manner. And we made sure to go over these outcome questionnaires such as the KCCQ before they went through it themselves just so they could understand what they were being asked. We had, of course, reliable contact information from the patient. And we provided a treatment plan for the week for them. Most of the time it would be as simple as just checking their blood pressure and weight every day, picking up their medications from the pharmacy and filling out their daily symptom survey assessment on their tablet. Here's a snapshot of what the digital platform looked like. This is kind of what the dashboard showed. And it basically displayed each patient as kind of a tile. so I don't know how well it shows up, but you can see the patients here are de-identified. You see a quick snapshot of their vital signs, which this was all continuous data collected and in real time, and the great part about this was when we had more and more patients enrolled, this was a great way to quickly screen through many of their vitals and see if any alarm parameters had been activated, which I'll go over what those were. But you could get a quick idea of who was maybe kind of thriving and who was at higher risk of decompensation. This was all HIPAA compliant, and again, video teleconferencing capability was available. You can see that was displayed there where they could chat with us, FaceTime with us if they wanted to, and again, the way that the health information was displayed, it was continuous, real-time monitoring, all time-stamped and able to be displayed with trends and kind of graphic graph formats as well. And then this was what the heart failure specific questionnaires looked like for the patients. You could see that it was time-stamped as well, and it was a simple way to basically fill out all these questions they could select from different choices here. So that's a little bit of what the virtual platform looked like from the patient's perspective. Now I'd kind of like to put you in the shoes of the patient of what day-to-day life in the clinic was for them. So they would wake up, and at about seven in the morning, if they had their tablet close by, it would remind them to put on their blood pressure cuff, be in range with their heart rate monitor, and take their daily weight. And then that would be transmitted to the database. It would also display a reminder for them to fill out their daily symptom assessment, which was a quick three-question survey of just how they had been feeling in relation to days before, if they felt better or worse. And again, those questions were all taken from the CardioSmart resource on ACC's website. They would then be contacted about an hour later by our nurses who were monitoring the database, and they would review any alarm symptoms with them or alarm parameters. They would remind them if they hadn't taken their weight or blood pressure, give them a gentle reminder that they needed to do that still. And if there was any alarm parameters met, that would be immediately sent to the physician on call for our Virtual Heart Failure Clinic, and then there would be a subsequent response intervention, whether that be just altering dose of diuretic or just monitoring more closely, or perhaps even advising them to check into the emergency department. And so the standardized alarm parameters that we set for our patients were these. We set a five-point reduction in the KCCQ score from baseline, that being a validated threshold for increased risk of cardiovascular mortality and increased risk of re-hospitalization if there was a five-point decrease. A weight gain of five pounds or more in three days, three back-to-back days of worsening heart failure symptoms, worsening vital signs, respirations, heart rate going up, symptomatic hypotension, or just reduced levels of activity on their movement tracker. Those were all parameters that we would be reaching out to the patients about to make an intervention. And then once weekly, we would have our multidisciplinary virtual rounds. And so that would be where all 10 team members mentioned would log on to Microsoft Teams, and we would basically, our objectives would be to review the individual patient data trends, to make recommendations on how to augment their GDMT, to go up or down on medications. We would place the prescriptions for those medications during these visits. These would all be documented in the EHR so that other providers could see what our thinking, what our strategies were for each of these encounters. And any appropriate referrals that needed to be replaced were also done at that time, whether that be cardiac rehabilitation, behavioral health providers, if we found depression on their screening, or perhaps even advanced heart failure if the patients were failing some of our therapies and needed a higher level of care. And the average duration for these rounds was anywhere from one to three hours. As we enrolled more and more patients, you can imagine the time duration went up, but for sure, this was much more efficient than face-to-face encounters. It would have taken much longer to do each one of these encounters face-to-face. And so we've talked in depth now about the structure of the clinic. I'd like to now focus our, or shift our focus on who our patients were. Our patient population was enrolled from July of 2022 to February of 2023, so about a seven-month enrollment duration. We had 41 patients in all enrolled, and 40 were able to complete the pilot clinic and one withdrew. And the average enrollment duration before these patients graduated, we called it, or were discharged from our clinic was just under 12 weeks, so 11.2 weeks. Our average age was 62 years. We had a predominantly male population, 27% female, 50% Caucasian, and 30% underrepresented minorities. Our average BMI was in the obese range, and our average baseline weight was 206 pounds. Some of our most common comorbidities with our patients, not surprisingly, were hypertension. 78% of our patients had that. 45% of them had diabetes, 35% with CKD, and then other comorbidities that were less common, but tobacco use, prior MIs, lung disease. And it's important to note here that we chose these exact comorbidities because this is what has been specified by the International Consortium of Health Outcome Measures as risk variables for heart failure outcomes. And so our patients were a pretty good mix of 43% ischemic and 57% non-ischemic cardiomyopathies, so a good mixture there. They certainly had reduced ejection fractions, an average of 29%, and the average pro BNP at baseline was elevated over 2,000. Their baseline KCCQ scores were, for the overall and clinical, were 66 and 69 respectively, which put them into the fair to good quartile. And just to remind ourselves a little bit about the KCCQ, it's an FDA-approved, well-validated clinical tool that helps to assess symptom burden over seven domains, 23 items, and also correlates very well with heart failure outcomes, cardiovascular mortality, and re-hospitalization. Scored from 0 to 100, and as you can see here, yeah, our patients were in kind of the third quartile, so fair to good as our starting point. Also at the starting point, we did encourage providers that were enrolling patients in this virtual clinic to try and at least start them on some form of GDMT to give us a bit of a head start, so we encourage providers to try and start them on two, three, or even four of our basic GDMT agents. So at the time of enrollment, 98% of our patients were on some dose of beta blocker, 90% were on a dose of ARNI, and 10% were on a dose of an ACE or an ARB. And you can see kind of our scheme for how often we up titrated these agents. Anywhere from five to 14 days was our typical timeline of increasing the doses on these. 85% of our patients had a prescription for SGLT2 inhibitor, such as Jardians, and 58% had a starting dose of an MRA, like spironolactone. And again, every 14 days we would go up on that medication, and then 93% of the patients enrolled were at least on some dose of maintenance diuretic, which we titrated as needed based on symptoms. And so just to give a little flavor for what our workload looked like for these 40 patients over that average duration of 11 weeks, we had 250 medication adjustments without any in-person visits, we had 106 virtual encounters for diuretic adjustment, we had over 7,000 virtual points of contact, either calls or virtual messages that was taken care of by our wonderful nursing staff, we had over 3,000 alarm notifications for some of those parameters we talked about earlier, and we had over 4,000 virtual encounters generated in our EHR. So a lot of those would be very quick kind of documentations of how we address some of these alarm parameters, but as you can imagine, took a lot of bandwidth for these 40 patients to provide this care to them. And so how did we do with our GDMT implementation? I can happily report that by the end of that 11 weeks, we had 85% of our patients on quadruple therapy at the time of discharge. So 100% on beta blockers, ARNIs, SGLT2 inhibitors, and we had 85% of our patients on an MRA. And for the 15%, there was certainly a contraindication here, it wasn't for lack of trying. We had tried to initiate these agents and due to things like hyperkalemia or reduced kidney function, they just could not tolerate it. And I think it's important here to gain some perspective on what's been seen in the past versus what we were able to accomplish in our clinic. So I would direct your attention to the CHAMP Heart Failure Registry. This was a registry of outpatients in the US with chronic HFREF done in 2018. And what this observational registry sought to characterize was the baseline use of GDMT agents like ACE, ARBs, ARNIs, beta blockers, and MRAs. And they just evaluated really the trends of GDMT implementation during this timeframe. It was prior to the use of SGLT2 inhibitors, so we can't hold that against them. But this study showed there were significant gaps in the use and dosing of GDMT. And I'm gonna display this in this bar chart here. So I would focus your attention on the red here. What's color-coded red is the large percentages of patients that were indicated for each one of these therapies but were not started on them for whatever reason. And our rates of, we know that it's not enough just to start these agents. So kind of going back to our data, we know that the full mortality benefit is really realized when we achieve 100% target dose for each one of these agents. So unfortunately, some of the other pie charts aren't showing up here, but I can tell you that we had a 90% adherence with full target dose beta blocker therapy. We had a 75% dose adherence with 100% ARNI dosage and 100% SGLT2 target dose for our patients, as well as 78% of our patients reached target therapy for our MRAs. To compare that again with what was done or what was seen in the CHAMP Heart Failure Trial, we can see large proportions of these patients with less than 50% of target therapy during their study. Sorry. I think now it's showing up. So those are our rates of 100% target therapy for our patients. And then this is our rates compared to the CHAMP Heart Failure Trial. So just to show you some contrast on what we were able to achieve. You might ask, though, what about an actual randomized control trial? Because we didn't actually have a control to compare ourselves to. This was more of a proof of concept or pilot study. Well, the Strong Heart Failure Trial did something similar. This was a randomized control trial that targeted patients that had been hospitalized for heart failure, either with HFPEF or HFREF, and they randomized this population to two different treatment strategies that being an intensive treatment strategy in which they up titrated all GDMT up to target dose within two weeks, so very aggressive, and then followed these patients up over four visits after that, or they just received usual care in the control arm. And their primary endpoint in this study was 180-day readmission and heart failure or all-cause death. And what they found in this was surprising. They found that in the intensive treatment strategy group, there was reduced symptoms, improved quality of life, and reduced risk of 180-day all-cause death. So they actually stopped this trial early because of there being a much greater than expected effect on these patients. So that being said, we wanted to compare our rates of GDMT up titration to full target dose with that of strong heart failure. So strong heart failure is displayed in orange here. Our clinic is displayed in blue. And you can see that on average, we did just as good or better than their rates of full GDMT up titration at 90 days. And so that being said, how did our patients kind of shake out at the end? So we were happy to report that by the end or discharge from our clinic at 90 days thereabouts, we had 95% of our patients reevaluated for LVEF improvements. About half of them underwent echo and the other half received a cardiac MRI. And we certainly saw a meaningful increase in their LVEFs. We started at about 29% and we ended on average with an EF of 38%. So overall a 9% increase, which is not incredibly surprising. GDMT definitely works. 38% of our patients received an appropriate ICD to prevent sudden cardiac death and 10% of those were appropriately given CRT devices if they met criteria. What about our safety record? That was really one of our main objectives in addition to just GDMT implementation. So our safety record included seven hospitalizations over the course of clinic enrollment. Three of those were for symptomatic hypotension, one for hypoglycemia, two for syncope and one for chemotherapy initiation. And while it's always discouraging to hear about any hospitalizations, we were happy to report that none of these were for heart failure decompensations. Five of our patients were appropriately referred to advanced heart failure and advanced heart failure center of excellence because of inability to tolerate some of our therapies and two patients in all were referred to behavioral health for severe depression that was discovered on their PHQ-2 screening tool. Some other safety endpoints that weren't already discussed, again, 0% cardiac failure for our patients enrolled, 7.5% symptomatic hypotension overall. 35% of our patients had hyperkalemia at some point during their clinic enrollment, which was appropriately treated. 27% of them had some form of renal impairment, likely as a consequence of adjusting our GDMT dosing. And any adverse symptoms was reported at some point in 65% of our patients. But these could be very minor symptoms all the way in leading up to reasons for hospitalization. What about quality of life improvements? I think that, you know, GDMT implementation and keeping these patients out of the hospital is very important. But if these patients didn't feel better, I don't think they would be very motivated to continue on with this care. And so just to remind ourselves, our overall clinical KCCQ score at baseline was just over, was in the 60s. And we can happily report that by the end of this clinic, these patients, for the most part, had a meaningful increase in their KCCQ score, anywhere from 12 to 13 points, which moved them into the next quartile of good to excellent, with averages either being 79 or 81, depending on which score you use. And again, just to highlight that every 5% increase in the score is associated with a 10% relative risk reduction in heart failure mortality and hospitalization. So a very meaningful outcome. Put another way, 70% of our patients had a greater than five point improvement from their baseline. 23% of patients either had no improvement or a deterioration of just five points from their baseline. And a small minority of 7% had more than a five point decrease from their baseline. What about our outcome measures? That was certainly something we wanted to validate our clinic against. So again, we reached for the International Consortium of Health Outcome Measures, their 2020 statement, which is designed to improve kind of quality and the level of care provided to our heart failure patients. So they came up with a 17-item outcome list, outcome measure list over four domains. And we're happy to report that we tracked every single one of these. We tracked mortality for our patients. We tracked their functional status through the use of the KCCQ score. We looked at their psychosocial impact of their heart failure diagnoses by screening for depression and other health, other behavioral health disorders with the PHQ-2. We also assessed their burden of care, of course, the financial burden to these patients for their medications, their complications from their treatments, medical stays, length of stays. We tracked all of that. When it comes to the ACC 2020 Heart Failure Performance Measures, we also adhered to this with a very high compliance rate. 100% of our patients had LVEF assessments. At enrollment and discharge, we tracked symptom management and activity assessments. Our GDMT was talked about in great detail of what rates of compliance we achieved with that. For MRA specifically, we had appropriate lab monitoring at standardized durations for that particular therapy. When it comes to Performance Measure 11 here, I'd like to clarify on that. Bidyl or hydralazine isosorbide dinitrate is indicated for those patients that self-identify as black or African-American. But it's only when they have Class 3 symptoms despite maximum tolerated GDMT prior to that. So only 50% of our patients actually had an indication for this and were appropriately prescribed this medication. 100% of our patients were counseled on ICD implantation to reduce or prevent sudden cardiac death. And those that were appropriate for CRT received that as well. For our quality measures, we certainly spent a lot of time with patient self-care education, measuring their patient-related outcome measures with the KCCQ score. And again, we had 75% of our patients with sustained or improved health status in their heart failure. For structural measures, just a few of these we assessed. We don't currently participate in registries at our particular institution, unfortunately, because of security and cost reasons. But that may be changing in the future. And when it comes to cardiac rehabilitation, which is highly recommended for all patients with heart failure, we reported 80% adherence rates with this. So 80% of our patients enrolled were actually able to complete the full three months of cardiac rehab. For the other 20%, they either refused or did not show up when their referral was placed. So in summary, our value proposition for a virtual heart failure clinic model is this. We saw improvement with access to care in a very expedient manner. We leveraged technology to manage very complex patients with a very lean and small team of providers. We achieved optimization of GDMT safely and efficiently. We had improved patient outcomes as evidenced by our GDMT implementation rates, improvement in EFs, reduced hospitalizations. Again, no heart failure hospitalizations during this clinic duration. And also improvement in symptom burden measured by several surveys. We had improved patient satisfaction as well. I can personally attest that I had several of my own patients enrolled in this clinic. And they were very pleased with the care that they received. They felt a very close relationship with the cardiology clinic during the clinic enrollment. And they felt much safer that they had a very close watch over them. We had improved heart failure performance measures as well, as outlined by many of the criteria from the AHA, ACC, and International Consortium of Health Outcomes. So questions that are maybe posed for the future as we think about sustaining this model. The questions that we have in our mind is, how does the heart failure clinic model virtual platform perform with a larger patient population? We only treated 40 patients here. And you can see all of the alarm parameters and EHR visits that were generated just for these 40 patients. So you can imagine how high that would be with a larger population. What is the accurate cost of a virtual heart failure clinic model? And how does this compare to an ambulatory visit model? So again, we didn't have a true control to compare ourselves to. And again, we were doing this on a fairly small scale. So the cost really of not only the virtual software and platform, but also all the staff needed to kind of monitor that. And then what is the appropriate patient-to-virtual staff ratio for patient monitoring and weekly rounds is on our mind. We did this in a manner that really most of the providers involved with this, this was all volunteer time. This was not built into their clinical schedules. And this was done on top of all their regular clinical duties. So I think this would look a lot different and would be a lot more sustainable if this was actually built into the schedules of several of our providers. So with that, I would like to thank you for your time. I'd like to especially thank my mentor, Dr. Fintanis, who was really the creative influence and inspiration behind all this. And I just want to thank him for recruiting me to be part of his team. So I'd be happy to take any questions. Again, thank you for your attention. »» So if you have any questions, you can put it in the app. And I think he maybe have addressed some of them during the conclusion of his presentation. Dr. Wilson, do you know how much those digital kits were? That's one of the questions the audience has. And you definitely saw obviously the cost savings in it with the reduction in your readmission. So that's a question our audience has. »» Yeah, that's a great question. I figured that would come up. So I can't tell you the exact cost of an individual membership for this. I was told by my staff that our institution had purchased like a 1.5 million contract that provided the electronic resources with current health plus the critical care trained nursing staff to monitor it. And that was for a few hundred licenses for it. And so we only took 50 of those spots from it. And we only used 40. So I can only tell you just the total bill for a year's worth of those. »» On top of that, obviously you know you are in the outpatient setting and they were seeing you for three months. How did they transition now back to their primary care or cardiologists? »» Yes, that's a great question as well. I can tell you that they still maintained their appointments with their primary care managers and their cardiologists. And they probably were seeing them less while they were enrolled in this clinic because there's the risk of too many cooks being in the kitchen and kind of trying to make too many changes independent of one another. But I can tell you that they were seamlessly discharged back to the care of their cardiologist or primary care manager. And they were well aware of all the strategies that were being employed during this clinic because of the extensive documentation. So it was a seamless transition for them back to their in-person care. »» One of the questions is again with the digital technology that's out there and triaging your patients, how were your alarms set? You know, if somebody's blood pressure was way up, if they were readmitted, what were those triggers for you guys and what you set those parameters at? And let's see, virtual, yes. Okay. I think I captured that. »» That makes sense. We had like I said those standardized parameters. And for the vital signs, most of the time that was really I guess individualized to a certain point. We really relied heavily upon symptoms for the patients. More often than not we had a lot of episodes of hypotension as you can imagine going up on a lot of these agents. And we were sure to very carefully monitor for symptoms associated with the hypotension rather than just the numbers themselves. So if we were able to make contact with the patients and they still felt well with lower blood pressures, we felt more kind of assured by that and didn't make any big changes. But if there were symptoms associated with high or low blood pressure, it usually resulted in a visit to the ED for them. »» Regarding transitions of care, I mean you're working from the inpatient setting to the outpatient setting. How did you educate the hospitalists or those attending physicians that this program was available? »» Yeah. You know, I think this was mainly advertised to the cardiology department. So I think that many of the primary care managers maybe saw these notes or the patients were able to tell them that they were enrolled in this clinic. But the main patient enrollment efforts were really sent out to our own department. So there was, I think most of the providers would just have to read our notes to understand what they were enrolled in and rely on the patients kind of telling them that they are part of this virtual clinic. »» A new question regarding, well the newest guidelines that just came out with the SGLT-2s, how did you overcome any potential barriers with that? And were there any cost issues in acquiring those medications? »» Yeah. So that was one advantage I will say that we certainly had in our kind of single payer system. So I'm part of the, we accept TRICARE patients. And so fortunately for them, SGLT-2 inhibitors and ARNIs such as Entresto are actually all formulary now for our patients. So there's actually no copayment for them. So I understand that certain systems have limitations on these medications. But for our particular patients, these were widely available and at no cost to them. »» Regarding staffing, so you mentioned that, you know, your team also had, were seeing patients as well. How did that kind of work and what percentage of time maybe they would devote to the virtual clinic? »» Yeah. So as more and more patients started to enroll in this clinic, it did take longer durations of time to address all of them. So at any one time I think we at the busiest of times maybe had like 20 patients enrolled at any one time because there were some that were entering, there were some that were discharging. But near the end it was taking approximately three hours to get through these virtual rounds to talk about all of the patients. And so again, this had to be done outside of our clinic time and our clinic duties. So we had to basically just kind of move our schedules around and at times it was tough because not all team members could be there. But the fact that we did these rounds on the Microsoft Teams platform allowed us to very easily kind of tune in and jump out of the conference if we had to. But there was always enough providers involved in these rounds that they could fill in the other ones if there was any, if they had missed any time. »» And again, regarding the alarms, you know, 7,000 points of contact with the patient. I mean, it seems very overwhelming in the 3,000 alarms. But you were able to adjust 250 medications. That's a lot of intervention in the outpatient setting. I'm sorry, this is jumping on me here. So is there anything that you could give advice on, on what would be those, again, particular triggers that you would look at and, you know, that would really triage up to the top or maybe down to the bottom? »» Right. Yeah. We've all heard of alarm fatigue. And, you know, when there's too many kind of alarm notifications, I think we tend to tune out sometimes. So I think that as you got to know these patients better, you probably would realize that some of these alarms that persistently kept happening, such as, you know, hypotension that was not associated with any symptoms may be kind of less important for some patients. So I think establishing maybe a trend of persistent alarm parameters per patient and knowing when to take those seriously or when to kind of just more closely monitor those would be key here to kind of signal a noise, I suppose. »» And also, I think you mentioned that the data that you did collect was also documented within your EMR. So was that able to pull, you know, the data? We have a lot of abstractors I know here. I want to know how to get that data. »» Yeah, absolutely. These were all coded under their appropriate ICD-10 codes and documented as a full, you know, as an encounter. So I think that they would be able to, I don't see a problem with kind of abstracting or mining some of this data because it was coded properly. »» And then during that time you said that you also referred your patients for cardiac rehab, again with COVID and everything else. Do you have a virtual cardiac rehab center or did they all have to do in person at this time? »» You know, I don't think that we have. I have not heard of any virtual cardiac rehabilitation. But I think we're kind of just so happy to see people back in person. We're kind of enjoying that. We have a cardiac rehab center of excellence at our institution. And if it's too far for our patients because of where they live, then we're able to find a network provider for cardiac rehab that's closer for them and works for them. But all in person as far as I know. »» Great. And everybody of course, they see the great outcomes with this project and they apply it to everybody else. Do you think this would be a great project for HFPEF patients now as well? »» It could be. It could be. You know, the challenge with HFPEF of course is really the care, the management strategies for HFPEF are definitely not as well validated. There's really not much in the way of standardized GDMT for these HFPEF patients. So I think it would be great to start a clinic including them. But the management strategies I would think would not be as standardized. But absolutely, I think it could be applied to them. »» Okay. Probably time for two more questions. One is about cardiac MEMS or Barostim. Did you ever consider including those in your program? »» That's a great question. Because of the remote monitoring capabilities, absolutely. We unfortunately did not have that capability. Our institution a few years ago did because we had an NP that was specifically responsible for monitoring all those values. But we actually just hired a provider that may be able to, that will be able to accommodate that. So I think that is absolutely something we'd like to add in the future and restart that program at our institution. And I could see the importance of incorporating that into a virtual clinic for sure. »» The question about equitability. I'm trying to get that word out of my mouth. Okay. So those with the least and maybe the sickest and that don't have internet or devices, any thoughts on maybe how we can expand to this population? I know you kind of addressed it a little bit. But anything innovative that maybe you can share or have kind of already thought of? »» Well, we could move them into the city, I suppose. That is a challenging barrier, I think, to overcome. Perhaps with more satellite and wireless capabilities, perhaps those services may become more available as technology progresses. And we would love to include them. Because as you say, they are the highest risk. But I think we'll have to wait on the technology to address that. But I think it's a great question. Technology to advance and allow that. »» Yes. I think we're all going to learn at our session on Friday that does discuss that as our closing plenary session. So again, I want to thank you for sharing your story with us, sharing your best practices. I think we've all learned a lot. I will go ahead if you guys want to come up and ask us questions about accreditation, about heart failure, please come up. But thank you again and enjoy the rest of your day. »» Thank you, everyone. I really appreciate your attention. Thank you.

Dr. Andrew Wilson

Brooke Army Medical Center

quality improvement initiative

virtual heart failure clinics

heart failure with reduced ejection fraction

HF-REF

managing HF-REF

multidisciplinary platform

digital toolkit