Abstract BackgroundPublic awareness and understanding of stroke are crucial to reducing its overall burden. Massive open online courses (MOOCs) are free, self-directed online learning platforms that provide accessible education. ObjectiveThis study aimed to describe the process, outcomes, and evaluation of co-designing the Preventing Stroke MOOC. MethodsThe co-design process was conducted with various stakeholder groups from July 2020 to December 2024 in Australia over three phases: (1) predesign, (2) co-design, and (3) postdesign. In the predesign phase, we consulted with MOOC researchers and held a half-day in-person workshop with community members to prioritize the stroke prevention topics, structure, format, and features, followed by user testing of a single developed module. The larger study comprised 16 virtual focus groups (14 co-design and 2 user testing focus groups) with community members and health knowledge experts. Of the 14 co-design focus groups, 12 informed the MOOC design, including content, structure, and delivery. This was followed by drafting content for each module, expert review (eg, clinicians and researchers), and development of platform materials. In the postdesign phase, we conducted user testing to collect feedback and evaluated the co-design process to determine how participants’ input shaped the final MOOC design. ResultsIn the predesign phase, 5 MOOC experts and community members contributed to a pilot module comprising 9 pages on the prioritized topics, including a focus on stories from people with lived experience of stroke. A total of 8 people tested this module, providing positive feedback about content and delivery. The co-design phase engaged 22 participants across 12 focus groups relevant to the MOOC. Their suggestions contributed to the multimedia design of the 7 core modules, with personalization via 9 optional mini-modules. Feedback on technical issues and suggested improvements identified during postdesign user testing were resolved. Evaluation of the MOOC co-design process found that the majority of participants’ suggestions gathered during the co-design phase were included in the final MOOC. ConclusionsA structured, co-design process resulted in the Preventing Stroke MOOC reflecting participants’ suggestions and provided important insights on the educational information needed to inform the general population about stroke prevention.
BackgroundCardiac rehabilitation (CR) improves patient quality of life, morbidity, and mortality. Unfortunately, it is underused by patients. Digital health interventions offer a solution to increase participation in CR. However, patients’ interest in virtual CR, especially among those in the inpatient setting, has not been fully explored. The benefits of CR have been predominantly demonstrated in traditional, center-based CR programs. ObjectiveThe objective of this cross-sectional study with prospective enrollment was to explore inpatient interest in virtual CR among adult patients who were hospitalized with a CR-qualifying diagnosis. MethodsA Qualtrics survey comprising multiple-choice questions was administered to cardiac inpatients at the progressive cardiac care unit at the Johns Hopkins Hospital from January 2020 to March 2024. Sociodemographic and clinical characteristics were retrieved from the electronic medical records. The study included English-speaking patients aged 18 years or older with a diagnosis eligible for CR. Digital health use was compared between an earlier enrollment period (2020‐2022) and a later enrollment period (2023‐2024). ResultsA total of 150 patients were included (mean age 64, SD 13 years; n=57, 38% female, and n=85, 56.7% White). With respect to sociodemographic characteristics, 26% (n=39) of the patients had a high school education or less, 53.8% (70/130) were married, 26.2% (34/130) were employed full time, and 63.8% (83/130) had private insurance. In exploratory enrollment-period comparisons, texting health care providers was more commonly reported in 2023 to 2024 than in 2020 to 2022 (74/123, 60.2% vs 6/27, 22.2%; P<.001), as was using a smartphone or tablet to learn about illnesses (93/123, 75.6% vs 17/27, 63%; P=.04) and to discuss health with a physician or nurse practitioner (86/123, 69.9% vs 12/27, 44.4%; P=.01). Participants with more than a high school education were more likely to perceive smartphones as beneficial for leading a healthier lifestyle (50/104, 48.1% vs 9/37, 24.3%, P=.01) and learning about illnesses (90/105, 85.7% vs 20/37, 54.1%, P<.001) than were participants with a high school education or less. Participants across all sociodemographic factors expressed interest in virtual CR (overall 107/150, 71.3%), with non-White participants being more interested than White participants (55/65, 84.6% vs 52/85, 61.2%, P=.002). ConclusionsMost cardiac inpatients expressed interest in home-based or virtual CR to alleviate barriers to in-person CR participation. Future work should emphasize digital equity and user support to optimize the widespread adoption of virtual CR.
Background:Atrial fibrillation (AF) is a major contributor of thromboembolic stroke. However, early detection remains challenging and is severely hindered by structural limitations in health care infrastructure. While current clinical guidelines recognize a 30-second single-lead electrocardiogram (ECG) as a standard diagnostic tool, the actual deployment and real-world utility of mobile screening technologies across health care settings remain poorly understood. Objective:The primary objective of this study was to evaluate the real-world feasibility of a mobile single-lead ECG for the detection of AF in routine inpatient and outpatient settings. Feasibility was evaluated by the proportion of successfully completed recordings, patient acceptance, and the occurrence and characteristics of technical challenges during deployment. Methods:This multicenter, cross-sectional pragmatic implementation study enrolled patients across 3 university hospitals and 3 primary care practices. Screening was performed using an US Food and Drug Administration (FDA)-approved single-lead ECG device (AliveCor KardiaMobile). Patients with an automated device reading indicating "Possible AF" and no prior diagnosis of AF underwent a confirmatory 12-lead ECG. Patient awareness and acceptance were evaluated using an exploratory, descriptive questionnaire. Results:A total of 237 patients were enrolled (194 inpatients and 43 outpatients). The device algorithm provided an automatic interpretation in 95.3% (41/43) of the outpatients and 85.6% (166/194) of inpatients, resulting in an overall automated detection of AF in 21.5% (51/237) of the cohort. The algorithm identified newly suspected AF in 1.7% (4/237) of patients without a prior AF diagnosis. Of these, 1 case underwent same-day confirmatory 12-lead ECG and was confirmed as AF. Technical challenges occurred in 16.0% (38/237) of all recordings, with a numerical higher rate observed in the inpatient setting (34/194, 17.5% vs 4/43, 9.3% in outpatients; P=.25). Only 6.3% (15/237) of patients reported regular use of telemedicine applications. Conclusions:Mobile single-lead ECG screening is feasible in routine inpatient and outpatient settings and identified a substantial prevalence of AF, which increased with age. The high detection rate underscores the potential of wearable ECG devices to facilitate early detection of AF. However, the absence of systematic confirmatory ECG testing and limited patient engagement with telemedicine applications highlight challenges in real-world implementation.
Abstract Background Smartwatches have significantly expanded personal health monitoring capabilities; however, their application in comprehensive cardiac assessment remains limited by single-lead electrocardiogram (ECG) acquisition. The inability to directly obtain augmented limb leads represents a major limitation for the evaluation of ischemic and arrhythmogenic conditions. Objective This study aimed to present and preliminarily evaluate a methodology for generating augmented limb leads (augmented voltage right [aVR], augmented voltage left [aVL], and augmented voltage foot [aVF]) from sequential smartwatch recordings through signal synchronization and RR interval alignment, and to combine this approach with precordial smartwatch recordings to reconstruct a full 12-lead ECG in a proof-of-concept setting. Methods A signal-processing framework was developed to preserve ECG morphology across sequential smartwatch lead recordings. The method is based on RR interval–guided temporal resampling, enabling interlead synchronization and analytical reconstruction of augmented limb leads. A prospective proof-of-concept clinical validation was conducted in 20 cases with diverse cardiac pathologies, including 1 healthy control. Smartwatch-derived ECGs were compared with standard 12-lead ECGs. Evaluation included expert-based clinical interpretation by 85 health care professionals and objective quantitative analysis of key electrocardiographic parameters, including heart rate, PR interval, QRS duration, QT interval, QTc, RR interval, P-wave duration, and PP interval. Mean values, SDs, and intermethod differences were calculated to assess agreement and variability. Results Reconstruction of a 12-lead ECG from smartwatch-acquired signals was successfully achieved in all cases. Visual correlation scores between smartwatch-derived and conventional ECGs ranged from 78.1% to 91.3% across specialty-level means (mean 81.1, SD 22.7), indicating high perceived concordance; individual case-level scores ranged more widely, from 47.0% to 96.3%. Evaluator-perceived axis agreement reached 95.4% across all cases, while the investigators’ direct axis measurements showed complete concordance in all cases. Quantitative analysis demonstrated small mean differences between modalities, with particularly high agreement observed in structurally stable parameters such as heart rate (mean 77.2, SD 22.3 vs mean 75.9, SD 23.1 bpm; Δ=1.3 bpm) and PR interval (mean 176.2, SD 29.1 vs mean 180.8, SD 19.8 ms; Δ=−4.6 ms). Greater variability was observed in atrial activation parameters such as P-wave duration (Δ=−13.3 ms) and time-dependent parameters such as QT (Δ=−19.2 ms) and QTc (Δ=−12.7 ms). Conclusions This proof-of-concept study demonstrates the feasibility of reconstructing a full 12-lead ECG from smartwatch-derived signals using a synchronization-based approach. The results demonstrated generally small mean differences between conventional and smartwatch-derived ECG measurements, although larger interval differences were observed in some individual cases. Importantly, due to the sequential nature of signal acquisition, the proposed methodology does not aim to achieve direct waveform equivalence but rather clinically meaningful agreement in electrocardiographic parameters. These findings support the potential of smartwatches as accessible tools for advanced cardiac monitoring and broaden their applicability beyond single-lead screening.
AbstractCardiovascular biomarkers are increasingly extracted or interpreted using artificial intelligence (AI) applied to electrocardiograms, imaging, laboratory measurements, electronic health records, wearable devices, and longitudinal data. Predictive performance alone, however, does not establish that a measurement is valid, that a model is transportable and calibrated, or that acting on its output improves care. Existing resources provide essential but complementary foundations: the US Food and Drug Administration–National Institutes of Health (FDA-NIH) BEST (Biomarkers, EndpointS, and other Tools) resource standardizes biomarker terminology; the V3 and V3+ frameworks address verification, analytical validation, clinical validation, and usability of digitally measured signals; prediction-model and trustworthy-AI guidance addresses reporting, risk of bias, early clinical evaluation, and deployability; and regulatory qualification pathways evaluate evidence within a defined context of use. A remaining practical challenge is translating these complementary requirements into accountable clinical action within cardiovascular workflows. This viewpoint proposes a clinician-centered operational framework organized around validation, governance, and workflow integration. For each pillar, it identifies accountable actors, required steps, documented outputs, and escalation or stop rules. Validation establishes whether the input measurement and model are fit for the intended population and decision. Governance assigns institutional authorization, clinical ownership, version control, monitoring, and authority to restrict, pause, or withdraw the intervention. Workflow integration specifies who receives the output, what confirmatory action follows, how disagreement is handled, and how decisions are documented. A worked example of an AI-enabled electrocardiographic screening output for possible left ventricular systolic dysfunction illustrates the pathway from local evaluation to echocardiographic confirmation and lifecycle monitoring. The framework does not replace established validation or regulatory standards; it operationalizes them as a clinician-centered and institutionally accountable pathway from validated signal to governed, patient-facing action.
