51 © Springer Nature Switzerland AG 2021 A. B. Bhatt (ed.), Healthcare Information Technology for Cardiovascular Medicine, Health Informatics, https://doi.org/10.1007/978-3-030-81030-6_4 Chapter 4 Building the Digital Heart Center Ameya Kulkarni 4.1 Introduction Norman Rockwell did not need technology to capture the soul of doctoring in his iconic painting, “Doctor and the Doll.” The painting, which depicts a doctor using his stethoscope to listen to the doll of a frightened young patient, reveals the best of what it means to be a physician—the chance to heal, but also the chance to be present during a patient’s most vulnerable moments. There is sometimes a fear that technology, and in particular remote care technology, diminishes this gift our patients give to us. However at its best, technology offers the chance for deeper connections that cross the bounds of time and space. Virtual care can allow us to take care of even our sickest patients from their homes with minimal interruption to their daily lives. Even for those that see this opportunity in telemedicine, the gap between a belief that virtual care will be helpful and the actual implementation of technologies to support that cause can feel very wide. Narrowing the gap requires a process to define the problems to be solved, navigate the myriad of new technologies, understand how these technologies can help our patients, and create systems to evaluate whether they actually did. Connecting these pillars with the core values of your practice are critical to building a successful and effective digital heart center. A. Kulkarni (*) Mid-Atlantic Permanente Medical Group, Rockville, MD, USA e-mail: amey.r.kulkarni@kp.org
52 4.2 Defining the Problem Cardiologists love technology. Our interest in quickly adapting new tools in our armament has allowed us to innovate care at a rate unparalleled in other fields. While this is often an asset in the care of our patients, when it comes to virtual care technologies, this can turn into a “shiny object trap.” The risk of trying new tools without understanding how they help our core practice or advance our core values results in expensive solutions to small or non-existent problems and eventually a graveyard of discarded shiny objects. Defending against this trap requires deliberation. It starts with a decidedly non- technical conversation. Prior to incorporating new technologies into our practice, we take a moment to define the problem we hope to solve with the technology. This helps identify whether a potential technology will be useful to us, but also helps us guard against the shiny object trap. Clearly establishing the problem to be solved is challenging. It requires input from people who understand the core values of the practice but also people who live the status quo. 4.3 Understanding the Technology Once the problem(s) have been defined, the next step is identifying the technologies best suited to address identified needs. Technology is constantly evolving, but most tools for virtual care fit into one of three categories: communication, data collection, and patient initiated. Table 4.1 Four questions for stakeholders Defining question Crowd-sourced answers Influence on platform development What do we do now that makes our work easier? Use of telephone and secure messages to communicate with patients When developing a digital platform, having an existing modality of patient communication, will negate the need to build one directly into the application What are “the pebbles in our shoes” (the things that make our job harder)? Prior authorization, (CV staff meeting) What do we wish we didn’t need to do in this workflow? Manually collect and review vital signs The platform has to not only collect data (ie, HR, BP, weight) but also present it in a way that reduced the burden of sifting through them What are the core issues in our workflow that negatively affect the patient experience? Long patient wait times and limited access to timely data for an initial visit. Virtual care schedules should be designed to improve access and align with image sharing programs. This allows clinical teams to have timely and clinically meaningful conversations. A. Kulkarni
53 4.3.1 Data Collection Tools The modern heart center requires more than discussions with patients. Home measurement of blood pressure and heart rate for patients with hypertension and coronary disease, weight and activity for heart failure patients, and even food journals for patients in virtual cardiac rehab programs are valuable adjuncts to routine care that can significantly improve the way we take care of cardiac patients. But collecting this data in a way that protects patient privacy and presenting it in a way that is useful to clinicians can take time, money, and some understanding of the technical and legal framework surrounding remote monitoring technology. At its simplest iteration though, minimal work is required. For example, a blood pressure collection system that asks the patient to measure their own BPs on a home machine and relay the data to the clinical care team via telephone or email offers a simple, inexpensive solution. If the problem to be solved is one of simplified and inexpensive data collection on a small scale, this solution will be adequate. As the scale or complexity of such a project increases, the work of verifying adherence to self-reporting and tracking becomes exponentially more difficult [1]. For these situations, more complex solutions are required. For example, when building a large remote monitoring program for patients across an enterprise with uncontrolled hypertension, there are three key values (Mid Atlantic Permanente Medical Group) that we considered mission critical: 1. The data collected has to be collected, transmitted and reported in a way that protects the patient’s privacy from end to end. 2. The process of signing up, transmitting data and dis-enrolling from the program should be easy for the patient and clinician. 3. Given the volume of data that would come to clinician dashboards, the reporting workbench had to be easy to navigate and contain “passive alerts” – which meant the system would signal the clinician when a reading was abnormal or when the patient was non-adherent to measurements, so that clinicians would not have to continually check dashboards for hundreds or thousands of people. Defining the problem ahead of the technology allows prioritization of what is needed for the minimum viable product (MVP) [2] and sets expectations for clinical teams. Once a roadmap is outlined, clinicians are willing to wait for iterative versions improving workflow (See example, Fig. 4.1). Of course, between a manual collection of blood pressure readings and a complex automated and integrated BP monitoring system are hundreds of technical solutions with varying degrees of complexity. Starting with the problems and following with the technology offers the chance to fit the solution for the right situation. 4 Building the Digital Heart Center
54 4.4 Implementation Building a strategic framework and defining the technologies set the theoretical groundwork for a digital heart center. The next phase, implementation, requires careful attention to regulatory considerations, clear definitions of what success and failure is, a willingness to iterate solutions as new information comes about, and an accounting for the unique personality mix among the clinicians in your practice. Sample Clinical Workflow for Clinical Telemedicine Visits This diagram demonstrates a sample clinical algorithm to enable delivery of virtual medical care and highlights opportunities to streamline workflows administratively to unburden physicians. New Patient Intake RN Triage No No No Source: Massachusetts General Hospial, Division of Cardiology Yes Yes Yes In-person visit scheduled by admin Is the patient appropriate and safe for a virtual visit? Telephone visit facilitated by admin Are there modifiable barriers to care? • Lack of Patient Portal access • Patient unsure of Portal set-up • Patient unfamiliar with virtual care • Patient is hard of hearing or has a language barrier • Admin registers as patient in Portal • Admin assists patient in Portal set-up • Admin familiarizes patient with visit protocol • Admin enables closed captioning or books interpreter as needed Admin facilitates access to virtual video visit Zoom virtual visit completed with clinical provider Is the patient amenable to a virtual video visit? Patient receives secure email instructions on virtual visit setup In-person visit completed with clinical provider Technical assistance for patient to reach virtual waiting room on day of visit provided by admin Patient receives automated email with visit link on day prior to visit Patient contacted by clinical staff on day prior to visit for medication reconciliation and COVID-19 screening Longitudinal provider Triage Established Patient Fig. 4.1 Sample clinical workflow for clinical telemedicine visits A. Kulkarni
55 4.4.1 Regulatory Considerations The most challenging aspect of offering virtual care is navigating the regulatory framework. At the local, state, and federal level, policies governing privacy, delivery of medical care, and handling of collected data are complicated and in the COVID-19 era, rapidly changing. The reality is that it is nearly impossible for clinicians to navigate these complexities. You will need help from institutional partners who are experts in technology risk, medical law, and compliance. In our experience, involving these experts early in the discussion, even as we are outlining the problem, helps bound our solution in what is possible, making the process significantly more efficient. These partners also help us favor simplicity, which means digital health tools can be evaluated without significant initial investment. 4.4.2 Monitoring and Evaluation A common trap in the early phases of building a digital health program is to start trying a technology without deciding what defines a successful tool, and what warrants abandonment of the idea. To combat this instinct, we begin implementation with the following questions: 1. How do we know if this new tool is a success? 2. How do we know if this new tool is a failure? 3. What is our criteria to expand? 4. What is our criteria to abandon? A critical step in answering these questions is to be as specific as possible. It is also important to be realistic. With new technologies, and in particular remote monitoring devices, there is a deep belief that technology can result in clinical improvement. Unfortunately, the data has not borne that out [3–5]. Empirically this makes sense, as remote monitoring tools don’t offer new information, but rather make that information easier to obtain. So defining success by other targets (such as increasing healthy days at home or improving patient satisfaction) may be more meaningful goals for the technology. As important as defining success is defining failure. Once deep into a pilot project, there is a natural inclination to “make it work,” often by adding financial or other investments. By defining a priori when the costs of the project or the results favor abandonment, it creates an honest arbiter of failure, and in the long run reduces the costs of those failures. 4 Building the Digital Heart Center
56 4.4.3 Iterations Technology companies often define the first version of any product as the minimum viable product (MVP). This concept gives permission for a new idea to not be perfect at first attempt, but also helps identify what needs to be perfect first, before other aspects of the product can be worked on. In designing digital health tools, a similar framework is necessary. Often the first version of a product will only solve the basic technical and regulatory challenges. Walking through the markers of success and the minimum requirements for each iteration are important, but most important is to acknowledge that there will be multiple iterations before a fully functioning technologic solution works as intended. 4.4.4 Gaining Support for Widespread Adoption Even the best of ideas need to be used by clinicians to be successful. Like technology adoption in any realm, digital health tools have an adoption curve. Innovators and Early adopters are willing to try new technology and will be willing to engage and re-engage, even when a tool is not functioning perfectly. Mainstream users (early and late majority) need a polished product but will tolerate some challenges. Late adopters require a nearly seamless experience and a well established workflow (Fig. 4.2). Part of any implementation strategy should include identification of which clinicians fit each of these categories for the technology being deployed. Since comfort with new tools is not monolithic—someone may be very comfortable with video visit technology but may not feel as comfortable with a remote monitoring platform—where clinicians fall on the adoption curve should be assessed with every roll out. Once the groups have been identified, roll out should follow the curve, with early adopters getting their first chance at trying the new solution and offering feedback on what doesn’t work. Feedback from late adopters will be useful as well, because when the solution takes a life of its own, the feedback from late adopters will help identify where the potential gaps lie. Central figure: Key steps in Building a Digital heart center Define the problems to be solved Identify technologies that will help solve the problem Implement the solution with a plan for defining success and iterations Fig. 4.2 Key steps in building a Digital Hearth Center A. Kulkarni
57 Case Studies In the following case studies, apply the framework from this chapter to design a technology based solution to problems commonly seen in modern cardiology practices. The discussion accompanying the case study is not meant to reflect a “correct” answer as there are many possible and viable approaches. Case 1: Your cardiology practice has noted a poor adherence rate to a traditional hospital based post myocardial infarction cardiac rehab program, with completion rates at 7%. Your team decides to design a virtual home based cardiac rehab program based in your office. Question 1: What core problem do you hope to solve in this program? Question 2: What technologies are required to accomplish your goal? Question 3: How will you know you are successful? Question 4: Design a program that would work in your practice. Discussion: In this case, the core issue was adherence to the program. While patients were being referred at a high rate and were attending their first sessions with consistency, by week 6 of the hospital based program, the rigors of needing to go to the hospital three times a week made adherence difficult. When we took stock of what worked well already and where the pain points were, we discovered that patients liked the guidance with exercise and regular contact with their rehab team, but felt that the rehab program was disconnected from the office practice and most importantly, they wished for an exercise program they could do at home. So the core problem to solve was to design a rehab program that allowed patients to safely exercise at home. 4.5 Summary Conversations on building a digital heart center often focus on the technology to be deployed. It is the fidelity of the ECG tracing on an Apple Watch, or the computing power of machine learning that dominates the discussion. But when a more systematic approach is applied, starting with identification of the problems to be solved, bolstered by careful selection of technology, and executed through an implementation plan that accounts for success measures and the clinicians responsible for using the technology, something bigger happens. The technology becomes a vehicle to solve the problems that all of us face in busy modern cardiology practices—how to diagnose and treat disease as quickly and efficiently as possible, how to streamline data collection so that treatment plans can focus on the patient needs, and most importantly, how to leverage modern tools to build deep and meaningful connections with our patients. At its best, the digital heart center gives us a chance to repaint Rockwell’s doctor on a modern canvas. 4 Building the Digital Heart Center
58 Interestingly, the technology to achieve such a goal was minimal. A twelve week program with an initial discussion of an exercise plan (after a stress test to establish baseline exertional tolerance) and regular check ins via phone solved the problem. Initially we discussed including an app based rehab program that collected activity and BP/weight data, but understanding that while interesting, that technology did not solve our core problem meant that we held off. The reduced technical lift meant that we were able to quickly deploy and expand a virtual home based cardiac rehab program with minimal financial investment. Defining success in this case stemmed from the core problem, which was completion adherence. We knew if we had a program with a higher rate of completion than 7%, we would have a successful program. Amazingly, after 1 year, our completion percentage was >70%. One lesson for me in the design of our virtual cardiac rehab program was that technology should fit the core solution. In this case, the most appropriate initial tool was a telephone, despite the availability of shinier objects! Recognizing this early meant more rapid deployment at a significantly lower cost. Case 2: The electrophysiologist in your practice has been wondering about a new way to consult on patients who need primary prevention devices. He is interested in minimizing the amount they need to travel to see him, and because he notes the physical exam doesn’t add much to a consultation to discuss device placement, he has been wondering about virtual methods of consultation. However, he says that the device battery itself is an important and useful part of the consent discussion, and so is concerned about the use of virtual consults. Question 1: What is the core problem? Question 2: What are the potential communication tools that can be used to solve the problem at hand? Which one will work the best? Discussion: In this circumstance, understanding more detail about the core problem helped elucidate the right solution. In asking the electrophysiologist what about the device was useful, he said that it was seeing the device that gave the most information. Understanding the dimensions of the battery offered patients who were anxious about having the implant some comfort. Evaluating virtual communication tools through this lens made the answer obvious—while a secure message with a picture would solve the visual problem, it wouldn’t allow for a conversation/informed consent. A telephone call would lose the visual portion, so a video consult was the optimal solution. By doing video consultations for new devices, patients could comfortably understand the risks of their upcoming procedure without being burdened with a drive to the office. A. Kulkarni
59 Case 3: Your traditional cardiology practice has been disrupted by a global pandemic of a highly infectious disease. You are particularly concerned about patients presenting to hospital emergency rooms with rapid atrial fibrillation, as these patients are most vulnerable to the infectious disease. Your organization is supportive in helping you build a home based hospital program for patients with rapid atrial fibrillation but has asked you to design a solution. Question 1: What are the core problem(s) to be solved? Question 2: What technologies will be required to build such a program? What other resources will be needed? Question 3: How will you measure success? Discussion: A highly relevant problem in 2020. The answers to these questions will vary by practice and by population, but for our practice, the core problem was safely managing patients with rapid atrial fibrillation outside the hospital. What this meant for us was the ability to (1) Be sure the patient was safe to be home (2) monitor heart rate periodically (3) Administer oral medications intermittently and (4) Connect with the patients virtually 1–2 times per day as needed. These four considerations structured the problems to be solved and thus the solution. To begin with, we needed to define who would be safe to go home. Because rapid atrial fibrillation is a condition of symptoms in most cases (and not a life threatening arrhythmia), we could use entry criteria for a home based hospital program to set safety guardrails. Patients with hemodynamic instability or evidence of heart failure, for example, would not be candidates for a home based program. Delving deeper into our heart rate monitoring requirements, we realized that medication titration really only required a measurement of pulse and BP—parameters that a patient could be taught to do in the ED prior to discharge. Doses of oral medications could be sent home with the patient for titration under the guidance of a physician and virtual connections could be established using our existing virtual tools (telephone/video/secure messages). Definition of success for us was solving the patient’s problem without requiring hospitalization. In this case, having an existing framework to design new solutions meant that we could quickly adapt to a seismic shift in the landscape of care in a safe and measured way. References 1. Bhandari A, Wagner T. Self-reported utilization of health care services: improving measurement and accuracy. Med Care Res Rev. 2006;63(2):217–35. https://doi. org/10.1177/1077558705285298. 2. https://healthcare.mckinsey.com/why-evolving-healthcare-services-and-technology- market-matters 4 Building the Digital Heart Center
60 3. Perez MV, Mahaffey KW, Hedlin H, et al. Large-scale assessment of a smartwatch to identify atrial fibrillation. N Engl J Med. 2019;381:1909–17. 4. https://twitter.com/venkmurthy/status/1042142046402891776?s=20 5. Chaudhry SI, Mattera JA, Curtis JP, et al. Telemonitoring in patients with heart failure [published correction appears in N Engl J Med. 2011 Feb 3;364(5):490] [published correction appears in N Engl J Med. 2013 Nov 7;369(19):1869]. N Engl J Med. 2010;363(24):2301–9. https://doi.org/10.1056/NEJMoa1010029. A. Kulkarni
RkJQdWJsaXNoZXIy MTYzOTI3MA==