Use of Human-Centered Design Methodology to Develop a Digital Toolkit to Optimize Heart Failure Guideline-Directed Medical Therapy

Abstract

Background:

Guideline-directed medical therapies (GDMT) improve quality of life and health outcomes for patients with heart failure (HF). However, GDMT utilization is suboptimal among HF patients.

Objective:

Engage key stakeholders in semi-structured, virtual human-centered design (HCD) sessions to identify challenges in GDMT optimization post-hospitalization and inform the development of a digital toolkit aimed at optimizing HF GDMT.

Methods:

For the HCD sessions, we recruited: a) clinicians who care for patients with HF across three hospital systems, b) patients with HF with reduced ejection fraction (HFrEF, EF≤40%) discharged from the hospital within 30 days of enrollment, and c) caregivers. All participants were ≥18 years, English-speaking, with internet access.

Results:

A total of 10 clinicians (median age 37 years [IQR: 35–41], 12 years [IQR: 10–14] of experience caring for patients with HF, 80% women, 50% White, 50% nurse practitioners) and three patients and one caregiver (median age 57 years [IQR: 53–60], 75% men, 50% Black, 75% married) were included. Five themes emerged from the clinician sessions on challenges to GDMT optimization (e.g. barriers to patient buy-in). Six themes on challenges (e.g. managing medications), four themes on motivators (e.g. regaining independence), and three themes on facilitators (e.g. social support) to HF management arose from the patient and caregiver sessions.

Conclusions:

The clinician, patient, and caregiver insights identified through HCD will inform a digital toolkit aimed at optimizing HF GDMT, including a patient-facing smartphone application and clinician dashboard. This digital toolkit will be evaluated in a multi-center, clinical trial.

Introduction

Background

In the United States, approximately 6.7 million Americans have heart failure (HF).¹ The prevalence of HF is expected to increase by 46% from 2012 to 2030, affecting over 8 million adults.¹ Among Americans with HF, roughly 1.3 million were hospitalized in 2019.¹ According to the American Heart Association Get With The Guidelines Heart Failure registry, 46% of patients admitted with HF have a reduced ejection fraction (EF) [HFrEF, EF≤40%].^1,2 Additionally, 3 in 4 patients with HFrEF die within 5 years of diagnosis.¹ However, the initiation of guideline-directed medical therapies (GDMT) for HFrEF is estimated to reduce the risk of cardiovascular death or HF hospitalization by up to 62% compared with limited therapy, resulting in an estimated 1.4 to 6.3 additional years of life.^1,3 For patients with HFrEF, GDMT consists of quadruple therapy with angiotensin receptor/neprilysin inhibitors (ARNIs), beta blockers, mineralocorticoid receptor antagonists (MRAs), and SGLT-2 inhibitors.^1,3,4 Despite the clear benefits of GDMT, less than a quarter of patients with HFrEF have been reported to be on optimal GDMT.⁵

The process of initiating, adhering to, and uptitrating GDMT is a team-based approach. GDMT is both a critical element and result of HF self-care. HF self-care has been defined as “a naturalistic decision-making process that influences actions that maintain physiologic stability, facilitate the perception of symptoms, and direct the management of those symptoms”.^6(p226) This process requires that patients adhere to what is often a complex pharmacological treatment plan, monitor signs and symptoms, track vital signs, and follow a specific diet and exercise regimen,⁷ all of which play an important role in uptitrating and achieving optimal GDMT.

Prior research to promote GDMT optimization has focused primarily on creating resources for clinicians.⁸ For example, the OPTIMIZE-HF trial utilized evidence-based practice algorithms, critical pathways, standardized orders, discharge checklists, pocket cards, and chart stickers to assist clinicians and hospitals in improving HF management.⁹ To a lesser extent the OPTIMIZE-HF trial encouraged patient activation, by providing patients with educational materials.⁹ There is a need for interventions that also promote patient activation, engagement, and shared decision-making to optimize GDMT. For instance, the Electronically Delivered Patient-Activation for Intensification of Medications for Chronic Heart Failure with Reduced Ejection Fraction (EPIC HF) trial demonstrated that a 3-minute video and 1-page checklist delivered to patients prior to cardiology clinic visits led to improved medical optimization through greater patient activation.¹⁰ Thus, interventions that combine clinician and patient-facing tools have the potential to improve GDMT initiation and optimization during and following hospitalization to reduce HF morbidity and mortality.

Digital health interventions (DHIs) that incorporate smartphone and/or smartwatch devices have the potential to reach patients where they are and empower them to take an active role in their HF self-care. DHIs are defined by the Food and Drug Administration as computing platforms, connectivity, software, and sensors for health care and related uses.¹¹ DHIs include categories such as mobile health, health information technology, wearable devices, telehealth and telemedicine, and personalized medicine.¹¹ Among patients with cardiovascular disease ([CVD], including individuals with HF) or CVD risk factors, 73% owned a smartphone and 48% used a health/wellness app on their mobile device.¹² In another study among individuals with CVD, 18% used wearable devices to monitor their activity and health.¹³ Patient-generated data from DHIs has the potential to help clinicians make informed decisions and uptitrate HF medications appropriately based on labs, vital signs, and symptoms.

Human-Centered Design Approach to Intervention Development

Human-centered design (HCD) is a creative design methodology for problem-solving that starts with people and ends with innovative solutions tailored to meet their needs.¹⁴ It has been widely adopted in engineering and social sciences, but only more recently in healthcare.^15–18 Two recent studies engaged patients with HF in HCD to develop digital health tools for HF self-care.^19,20 However, there is further opportunity to engage additional stakeholders, such as clinicians and caregivers, in the HCD process as HF self-care and GDMT optimization requires a team-based approach.

Purpose

Johns Hopkins University, University of Michigan, and Stanford University were among the American Heart Association Health Technologies and Innovation Strategically Focused Research Network Centers selected to lead a collaborative research study to improve HF management. We aimed to co-design a digital toolkit for HF self-care post-hospitalization with a focus on promoting GDMT optimization. A HCD approach was utilized among diverse clinicians, patients, and caregivers to inform the digital toolkit and support scalability and impact on patient care to improve outcomes and quality of life. The digital toolkit developed from this HCD work will later be evaluated in a multi-center, clinical trial.

Methods

Study Setting, Participants, and Recruitment

In this study, we included three relevant stakeholder groups (clinicians, patients, and caregivers) to inform the development of the digital toolkit for HF management and meet end-user needs. Clinicians (e.g. advanced practice providers, pharmacists, and physicians) at Johns Hopkins Hospital, Johns Hopkins Bayview Medical Center, University of Michigan Health, and Stanford were eligible to participate in the study if they had greater than three years of experience providing care to patients with CVD. The study team at each site used a structured recruitment email to recruit potentially eligible clinicians. Patients were included in the study if they were 18 years of age or older, proficient in English, and were admitted to the hospital for HFrEF (EF≤40%), within 30 days of enrollment. Caregivers of the enrolled patients were given the opportunity to participate in the study if they were 18 years of age or older, proficient in English, and had a history of caring for the patient participating in the study. Patients and caregivers were excluded if they were unable to provide informed consent or did not have access to a computer, tablet device, or smartphone. Patients on inotropes, with or preparing to receive a left ventricular assist device, or awaiting a heart transplant were not included given the severity of their condition. Patients and caregivers were recruited from Johns Hopkins Hospital and the Johns Hopkins Heart Failure Bridge Clinic. The Heart Failure Bridge Clinic helps patients manage their HF by providing a smooth transition from hospital to home as well as by offering support during HF exacerbations. All participants were consented using DocuSign. This research has been approved by the Johns Hopkins University Institutional Review Board (IRB00303148).

Human-Centered Design Sessions

We used the iDesign framework, a virtually inclusive DHI design approach to promoting health equity, to guide the HCD sessions in this study.²¹ The iDesign framework includes seven steps with varying sequence depending on the type or maturity of the project: (1) defining challenges and empathizing, (2) ideation, (3) virtual onboarding, (4) virtual prototyping, (5) ranking prototypes based on feasibility, viability, and desirability, (6) engineering work, and (7) testing (Figure 1).²¹ The co-design activities in this study targeted three of the core phases: defining challenges and empathizing, ideation, and virtual prototyping. The virtual onboarding phase was not conducted for this study given that the intervention is in the early design phase. The study team is conducting the remaining steps (steps 5–7) related to (5) ranking the prototypes based on feasibility, viability, and desirability, (6) engineering work, and (7) testing. The results of the study team ranking are not shared due to a lack of generalizability given that feasibility, viability, and desirability would likely vary by research team depending on funding and development timelines. However, other researchers should consider how the generated ideas might inform their own interventions.

Figure 1. Steps of the Virtual Inclusive Digital Health Intervention Design to Promote Health Equity (iDesign) Framework²¹.

This figure was published (and can be reproduced) under the terms of Creative Commons Atrribution 4.0 license.

All of the HCD sessions were conducted virtually through the Zoom platform and activities were facilitated using Google Jamboard^™. The clinicians participated in two, 1-hour sessions. The patients and caregivers participated in three, 1.5-hour sessions separate from the clinicians. The clinician and patient/caregiver sessions were conducted separately so they could freely discuss the challenges they face in caring for patients with heart failure or living with and managing heart failure, respectively. Furthermore, we wanted to limit the sample size per session to allow sufficient opportunity for all participants to engage. The sessions were guided using a semi-structured interview guide. The activities, across the various HCD phases, and the guiding questions are outlined in Table 1.

Table 1.

Activities Guiding the Human-Centered Design Sessions

Session	Activities	Duration	Guiding Questions/Probes
Clinicians
Defining Challenges and Empathizing	Journey mapping Finding themes Creating “How Might We (HMW)” questions	1 hour	What challenges do you face in optimizing guideline-directed medical therapy among your patients during and after a hospitalization for acute decompensated heart failure? What challenges have you observed your patients facing in managing their heart failure care and adhering to guideline-directed medical therapy following discharge from the hospital? What challenges and opportunities do you foresee in using patient generated data, displayed on a clinician dashboard, to inform your clinical practice and titrate medications among patients with heart failure following hospitalization?
Ideation	Brainstorming around HMW questions	1 hour	HMW educate patients on their heart failure disease process, management, and medications in a way that doesn’t overwhelm clinicians? HMW best collect and display patient data to facilitate medication optimization? HMW as clinicians promote medication optimization and patient adherence? HMW optimize the workflow and process of caring for patients with heart failure, including follow-up for uptitration of medications?
Prototyping	Prototyping around HMW questions
Patients and Caregivers
Defining Challenges and Empathizing	Journey mapping Finding themes	1.5 hours	Take a moment to think back to your experiences, good and bad, during your recent hospitalization for heart failure, whether you were diagnosed with heart failure for the first time or you were admitted to the hospital for exacerbation of your heart failure symptoms. Also think back on your experiences in managing your heart failure condition following discharge from the hospital. We want to hear either your personal experience as a patient or your experience as a caregiver.
Ideation	Creating HMW questions Brainstorming around HMW questions	1.5 hours	HMW better engage caregivers (family, friends, significant others, peers with heart failure) to facilitate and support heart failure care? HMW better enable patients to track their heart failure care activities (e.g. fluid restrictions, low sodium diet), symptoms, medication management, and progress (e.g. improvement in functionality)? HMW help patients better manage their medications, especially if they are new or being changed frequently, to promote adherence (e.g. understanding what medications do, taking medications as prescribed, refilling prescriptions on time)?
Prototyping	Prototyping around HMW questions	1.5 hours

Data Collection and Analysis

Study data were collected and managed using REDCap electronic data capture tools hosted at Johns Hopkins University.^22,23 REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies. All participants completed a brief demographic survey sent via email from REDCap. The Zoom HCD sessions were recorded. The audio recordings from the clinician and patient/caregiver “Defining Challenges and Empathizing” sessions were transcribed verbatim. Two authors (EMS and IMA) independently used an inductive approach to coding the qualitative data, allowing the data to determine the themes, and met to resolve any discrepancies through open discussion. To resolve discrepancies, each coder provided their rationale for using a certain code and through discussion came to a consensus for every disagreement. This process helped to further refine the themes. A third author was not needed to resolve discrepancies. F4Analyze was used to code the data. Thematic analysis was used to generate the themes based on the coding framework. The output from the “Ideation” and “Virtual Prototyping” sessions were recorded on Google Jamboard^™ during the HCD sessions.

Results

We enrolled 14 clinicians, 10 patients, and 2 caregivers in the study. Of those enrolled, 10 clinicians, 3 patients, and 1 caregiver participated in the HCD sessions. Of the clinicians, 70% (7/10) attended both HCD sessions and 30% (3/10) attended one session.

The clinicians had a median age of 37 years (IQR: 35–41) and 12 years (IQR: 10–14) of experience caring for patients with HF; 80% (8/10) were women; 50% (5/10) were Asian and 50% (5/10) were White; and 50% (5/10) were advanced practice providers. For physicians, the median number of years of experience caring for patients with HF included those as a resident, fellow, and attending. For advanced practice providers, the median number of years of experience caring for patients with HF included those as a nurse and nurse practitioner.

The patients and caregiver had a median age of 57 years (IQR: 53–60), 25% (1/4) were women, 50% (2/4) were Black; and 75% (3/4) were married. Among the patients, 67% (2/3) had an Associate degree (academic, occupational, technical, or vocational program) and 33% (1/3) had a Doctoral or Professional School degree; 67% (2/3) were currently employed; and 100% (3/3) had health insurance.

Defining Challenges and Empathizing Sessions

Clinicians.

The clinicians discussed challenges faced in optimizing GDMT among their patients during and after hospitalization for HF. Five themes emerged from the data, including: (1) logistical issues, (2) barriers to patient buy-in, (3) hospitalizations and post-discharge challenges, (4) clinician inertia, and (5) using remotely provided patient data. Supplemental Digital Content 1 (table) provides defining quotes for each of these themes.

The most common logistical issues that clinicians reported as barriers to optimizing GDMT were limited clinic bandwidth, resulting in the inability to schedule multiple visits with patients in a short period of time, and administrative support for completing insurance prior authorizations for medications.

The clinicians also discussed multiple barriers to patient buy-in to GDMT, including patients’: perceived lack of improvement in their condition resulting in them not wanting to increase their medication doses, lack of objective targets for treatment to help them understand how they are benefiting from medication uptitration, knowledge deficit of the disease process in general, feeling overwhelmed by the number of medications and sometimes need to retrial medications, trying to balance medication side effects and quality of life, and socioeconomic reasons.

The high number of hospitalizations in this population was also discussed as a challenge for clinicians to medication optimization due to frequent medication changes in the hospital leading to patient confusion, downtitration of HF medications due to lack of knowledge by other clinicians or competing treatment priorities, and clinical contraindications.

Clinician inertia was also brought up by way of clinician guilt, and how sometimes clinicians feel guilty changing patients’ medications so frequently, especially since the medications are often costly. The clinicians also talked about challenges in using remotely provided patient data to make decisions and optimize therapy due to a lack of confidence in the completeness and accuracy of the data as well as the uncompensated time to review the data and make medication changes.

Patients and Caregiver.

The patients and caregiver discussed challenges, motivators, and facilitators to managing HF following hospital discharge. Six challenge-related themes arose from the data, including: (1) vague HF symptoms, (2) the transition from hospital to home, (3) changing one’s diet, (4) incorporating exercise, (5) managing medications, and (6) management across multiple chronic conditions. Four motivator-related themes emerged from the data, including: (1) being present for family, (2) taking ownership of the problem, (3) regaining one’s independence and abilities, and (4) tracking objective markers of clinical improvement. Lastly, three facilitator-related themes arose from the data, including: (1) having social support, (2) having religion or faith, and (3) receiving HF education. Supplemental Digital Content 2 (table) provides defining quotes for each of these themes.

The patient and caregiver group discussed how atypical or vague HF symptoms were a challenge to management and knowing when to seek care. They also talked about how the post-discharge period, the transition from hospital to home, can be a particularly scary time due to the many changes one has to incorporate into their life. The difficult changes that were mentioned included changing one’s diet, incorporating exercise, and learning to manage medications. Changing one’s diet, including limiting fluids and tracking sodium, was mentioned to be particularly difficult. In regards to medications, these individuals did not struggle with remembering to take their medications but they did discuss how it was easy to forget refills. Individuals with multiple chronic conditions talked about how diet and medication management was all the more challenging for them given the varying dietary restrictions and number of medications.

Despite the challenges that were brought up, the patients and caregiver also talked about motivators for HF management post-discharge. Being present for one’s family as well as taking ownership of the problem were highlighted as motivating factors. Importantly, regaining one’s independence and abilities was discussed in detail and seen as critical for having good quality of life. Lastly, being able to track objective markers of clinical improvement was seen as highly motivating for continuing self-care behaviors.

Having both family and peer social support was seen as a facilitator to HF management. The individuals spoke about how they couldn’t have adapted to having HF without the support of their loved ones. Interestingly, the individuals all felt they lacked support from other individuals with HF due to limited opportunity to engage with peers. Having faith or religion was also brought up as a facilitator to HF management. Finally, receiving the proper education from clinicians was seen as an important contributor to engaging in HF management.

Ideation and Virtual Prototyping Sessions

To address the challenges identified above, during the ideation and virtual prototyping sessions, the clinicians brainstormed around different intervention components that could be helpful in (1) educating patients on their HF disease process, management, and medications, (2) collecting and displaying patient data to facilitate medication optimization, (3) promoting medication optimization and patient adherence, and (4) optimizing the workflow and process of caring for patients with HF. The patients and caregivers also brainstormed around intervention components that could (1) help engage caregivers and leverage social support to facilitate HF care, (2) enable patients to track their HF care activities, and (3) help patients better manage their medications.

Table 2 includes the list of core intervention recommendations that were suggested by clinicians, patients, and the caregiver including the need to develop (1) educational materials on the HF disease process, medications, and medication optimization process; (2) platforms for peer support; (3) digital tool(s) for patients to track their data; and (4) a dashboard for clinicians to review patient recorded data.

Table 2.

List of Core Intervention Recommendations from Clinicians, Patients, and Caregivers during Ideation and Prototyping

Core Intervention Recommendation	Recommendation Description
Educational Materials on HF Disease Process, Medications, and Medication Optimization Process	Modalities Videos (made by clinicians and other patients with HF) and reading materials available to patients and caregivers in the hospital and post-discharge Scenario based education on hospital iPads HF family class in the hospital to learn about key HF topics Key Topics Nutrition (e.g. how to read food labels, determine sodium content) Medications (e.g. how they work, consequences of not taking as prescribed, side effects, costs) Medication optimization process Delivery Break up information Set expectations up front for medication optimization process
Platforms for Peer Support	Virtual support groups (Zoom, chat rooms) and recordings of other patients to build comradery, ask questions, and learn about others’ experiences with HF
Digital Tool(s) for Patients to Track Their Data	Leverage home health nurses to support patient data entry post-discharge Data entry into smartphone application or flowsheet via patient electronic medical record portal Ability to enter key data including vital signs, labs, and symptoms Alerts for abnormal parameters and to re-check data Helpful notes that help interpret key values (e.g. elevated potassium levels) Reminders to check for signs or symptoms of exacerbation Auto population of medication changes directly into patient digital tool Reminders to take medications and a refill countdown A medication titration schedule to help patients plan for dose increases Graphs correlating medication changes with changes in vital signs, labs, and symptoms Tool with functionality to look up sodium content in meals at restaurants Incorporate goal setting to regain independence and usual abilities
Clinician Dashboard to Review Patient Recorded Data	Ability to view and provide feedback on vital signs, labs, and subjective data Alerts at different thresholds that would require clinician follow-up and/or medication change Ability to filter patients by risk of readmission and mortality (i.e. need closer follow-up) Graphs and averages to visualize trends Ability to provide positive reinforcement to encourage patient data entry

HF: heart failure

Discussion

In this study, three American Heart Association, Health Technologies and Innovation, Strategically Focused Research Network Centers leading a HF collaborative project sought to leverage HCD methodology and engage clinicians, patients and caregivers in the design of a digital toolkit for HF self-care post-hospitalization. The planned focus of these sessions was on defining challenges to GDMT optimization. While much of the conversation focused on medication optimization, participants also highlighted additional challenges to self-care following hospitalization. The findings from this HCD work will inform the development of a digital toolkit as well as the design of a multi-center trial.

Two prior studies have also used a HCD approach to inform the development of a digital toolkit for HF self-care.^19,20 This research builds upon prior work by engaging clinicians and caregivers, in addition to patients,^19,20 in defining challenges to GDMT optimization and brainstorming solutions for integration into the digital toolkit. By including diverse stakeholders, we will be more likely to develop a solution that meets various end-user needs.

There were some commonalities in the key themes that emerged from both the clinician and patient/caregiver sessions. One of the barriers to patient buy-in to GDMT that the clinicians discussed was the lack of objective targets for treatment to help patients understand how they are benefiting from the medications. The clinicians gave the example of how there are clear goals for hypertension management²⁴ but less so for HF management. The patients also discussed how being able to track objective markers of clinical improvement was highly motivating for engaging in self-care. Therefore, finding ways to display to patients the quality of their medical therapies compared to best practice recommendations or visual displays that help patients track objective and subjective markers of clinical improvement may be of value. In the future, remote pulmonary artery pressure monitoring could be explored as a way of providing additional objective targets for clinicians and patients.²⁵ A meta-analysis of three randomized clinical trials (MONITOR-HF, CHAMPION, and GUIDE-HF) demonstrated that among patients with remote pulmonary artery pressure monitoring, total HF hospitalizations were reduced by 30%, independent of EF.²⁵ Across all three trials, there was also a higher number of cumulative medication changes in the pulmonary artery pressure monitoring group, especially for diuretics.²⁵ Pulmonary artery pressure monitoring is an invasive procedure, which may limit acceptability and scalability. However, remote monitoring and clear objective markers of clinical improvement or deterioration may trigger interactions between clinicians and patients to optimize medical therapy.²⁵

Furthermore, the clinicians, patients, and caregiver all identified patients’ limited knowledge of the disease and medication optimization process as a barrier to GDMT optimization, indicating there is an opportunity to develop more engaging educational materials for both patients and caregivers. The patients included in the HCD sessions were not very familiar with the medication uptitration process and could not clearly identify if their clinician had worked with them to increase their medication doses. Other work has also found that many patients are not familiar with GDMT for HF and question the safety and effectiveness of therapy.²⁶

The clinicians cited socioeconomic reasons, such as medication costs and limited transportation to appointments, as patient barriers to buy-in to GDMT. Interestingly, the patients and caregiver did not cite this as a challenge; however, the participants were well-educated, employed, and had health insurance. Prior research has found that lower socioeconomic status (education, employment, insurance type, transportation, and federal poverty level) is associated with worse HF self-care.²⁷

In this study, the clinicians mentioned that balancing medication side effects and quality of life was often a challenge for patients and a barrier to GDMT optimization. The patients also discussed how important it was for them to regain their independence and quality of life, but as a motivator to GDMT uptake. Other research has noted that patients with HF and family members often struggle with poor quality of life and that patients have a difficult time with the loss of independence and need to rely on others.^27,28 The patients also discussed the importance of both family and peer social support; similar to other research, they indicated how critical family support was but how social support networks were often lacking.^27,28 The patients and caregiver felt that virtual support groups or recordings of other patients should be leveraged to build comradery, provide an opportunity to ask questions, and learn about others’ experiences with HF. The patients and caregiver also spent much of the conversation discussing the challenges in managing dietary restrictions, both for HF but also in tandem with those of other chronic conditions. This challenge has been noted in prior work as well.²⁹

Including clinicians in these HCD sessions resulted in additional findings not well documented in the literature. In particular, the high number of hospitalizations in the HF population was noted to result in additional challenges to GDMT optimization. For example, during a hospitalization, patients are often cared for by clinicians with less specialized knowledge in HF care. This deficit in clinician knowledge may result in HF medications invariably being downtitrated during their hospital stay, leading to HF specialists having to restart the medication optimization process post-discharge.

The clinicians also discussed the challenges associated with limited clinic bandwidth when frequent follow-up is required for medication optimization. DHIs present an opportunity to engage patients in remote monitoring and medication optimization. However, steps should be taken in DHI design to address concerns regarding using virtually provided patient data to make clinical decisions.

This study has some limitations that should be considered when interpreting the findings. First, the sample size for the patient and caregiver group was modest. Initially, ten patients and two caregivers were enrolled in the study but eight were lost to follow-up (n=4) or chose to withdraw from the study for various reasons (n=4). There was no prevailing reason for why participants decided to withdraw from the study; however, it is possible that the severity of their condition and the challenging post-discharge period were contributing factors. The small sample size did allow ample opportunity for all participants to voice their thoughts and engage in in-depth conversation. We believe there was thematic saturation given the similar concerns shared by patients and clinicians. Furthermore, the clinicians had a median of 12 years of experience caring for diverse patients with HF. Additionally, the patient and caregiver participants needed to have access to the internet to participate in the virtual HCD sessions. Second, the clinician participants were generally younger and identified solely as White and Asian. Enrolling more senior clinicians and Black individuals may have resulted in additional ideas. Lastly, the clinicians consisted solely of advanced practice providers and physicians associated with a HF clinic. Including pharmacists or clinicians that care for patients with HF but are not associated with a HF clinic (e.g., general cardiologists or internists) could have provided additional valuable insight into barriers to GDMT as many patients are not cared for by HF specialists. However, half of the clinicians were advanced practice providers (i.e. nurse practitioners) who provide much of the care for these patients.

This study also has a number of strengths that should be noted. First, we used a HCD approach to define the problem and generate potential solutions for integration into the digital toolkit that will later be evaluated in a clinical trial. Second, we included multiple stakeholders (clinicians, patients, and caregivers) to develop a solution that would meet diverse end-user needs. Third, we focused the discussion on the post-discharge period and specifically enrolled patients with a recent hospitalization for HF. This approach brought forth interesting ideas from both groups about challenges, motivators, and facilitators to GDMT optimization during this critical and difficult time.

Conclusions

In this study, we used HCD methodology to engage clinicians, patients, and caregivers in the design of a digital toolkit to promote GDMT optimization. The findings expand prior work by focusing on challenges to GDMT optimization during the post-discharge period. These results will inform the development of a digital toolkit, including a patient-facing smartphone application and clinician dashboard. A multi-center clinical trial will be conducted to evaluate the impact of the intervention on GDMT optimization among patients with HFrEF.