Unlabelled:Cardiovascular 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.
Background:Cardiac rehabilitation (CR) is a class IA-recommended treatment and secondary preventive strategy for people with coronary artery disease. eHealth solutions show promise in providing accessible CR. However, few provide a comprehensive and multimodal CR program, and most programs are developed without input from end users. Objective:The aim of this study was to provide a detailed description of the multidisciplinary co-creation process used to develop a person-centered, multimodal digital CR program that integrated the perspectives of stakeholders and end users throughout, with the rationale to optimize the acceptance and feasibility of the CR program. Methods:We incorporated a user-centered design with a 3-phase, iterative, multidisciplinary stakeholder co-creation process. This included defining needs (phase 1), development and refinement (phase 2), and testing and optimization (phase 3). Overall, 26 internal stakeholders from the project team participated, including clinicians, researchers, software developers, and user representatives. In addition, 35 external stakeholders (20 clinicians and 15 people with cardiac-related problems) participated in a user-centered development process that included workshops, intervention content development, and usability testing. Data were collected through contextual inquiry, 4 workshops with stakeholders, 12 co-creation sessions with user representatives, 2 usability testing sessions using a think-aloud methodology, standardized usability surveys, and individual and focus group interviews. Intervention content was developed and finalized based on existing evidence, stakeholder input, and user testing. A smoke test and a minimum viable product test were conducted. Data were analyzed using rapid qualitative content analysis and descriptive statistics. Results:Analyses of data gathered across the 3 phases identified 5 themes important to stakeholders and end users: a program with (1) personalized and tailored solutions with reliable, trustworthy, and evidence-based content (ie, a face-to-face introduction session at the start of the program and a tailored registration page to individually register personal risk factors, goals, and medications); (2) options for feedback (ie, an asynchronous messaging service and tailored face-to-face video consultations with clinicians); (3) peer support (ie, exercise sessions in groups); (4) digital reminders (ie, notifications on a smartphone or tablet); and (5) motivational features (ie, interactive cardiac-specific learning modules on coronary artery disease and secondary prevention). Evidence-based content development resulted in 9 cardiac-specific interactive learning modules, a module for personal health data entry, live exercise sessions, tailored digital follow-up, and an asynchronous messaging service. Readability was assessed using the Gunning Fog Index, with levels ranging from 4 to 13. The overall System Usability Scale score was 86.2 (SD 12.2), indicating excellent usability. Conclusions:A multidisciplinary stakeholder co-creation process combining well-known evidence-based concepts with stakeholder input identified several areas for improvement and facilitated the development of the digital CR program. Contextually tailored, co-designed interventions may improve relevance and responsiveness to user needs. The feasibility and effectiveness of the program will be further evaluated in the eCardiacRehab trial.
Background:Patients hospitalized for heart failure (HF) face a high-risk early postdischarge period. Objective:We aimed to develop and internally validate the soluble ST2, age, and estimated glomerular filtration rate-heart failure (SAGE-HF) score, a simplified risk model incorporating soluble suppression of tumorigenicity 2 (sST2), age, and renal function, for predicting 30-day postdischarge major adverse cardiovascular events (MACE). Methods:This two-center prospective cohort study included 218 patients hospitalized with acute decompensated heart failure (ADHF) and/or heart failure with reduced ejection fraction (HFrEF) and 59 non-HF controls in Vietnam (July 2025-January 2026). Serum sST2 concentrations were measured at admission using enzyme-linked immunosorbent assay and compared between the ADHF/HFrEF cohort and controls, with adjustment for clinical covariates. Among the ADHF/HFrEF cohort, the primary endpoint was 30-day MACE, defined as a composite of all-cause death or HF rehospitalization. Candidate predictors were considered based on clinical relevance, biological plausibility, and exploratory univariable associations, and prespecified parsimonious logistic regression models were evaluated using discrimination (area under the receiver operating characteristic curve [AUC]), calibration, Brier score, and Akaike information criterion. Internal validation was performed using 1000 bootstrap resamples, and model robustness was assessed using Firth penalized logistic regression. The SAGE-HF score was derived from the final model. All data were analyzed using the R environment (version 4.5.5; R Foundation for Statistical Computing). Results:Among 277 participants, sST2 concentrations were significantly higher in patients with HF than in non-HF controls after adjustment. Among 218 patients, 47 (21.6%) experienced 30-day MACE. In multivariable analysis, age (odds ratio [OR] 1.04 per year, 95% CI 1.01-1.07; P=.02), sST2 (OR 1.06 per ng/mL, 95% CI 1.01-1.11; P=.01), and estimated glomerular filtration rate (OR 0.98 per unit, 95% CI 0.97-1.00; P=.02) were independently associated with MACE. The final model demonstrated acceptable discrimination (AUC 0.741, 95% CI 0.664-0.819) and good calibration, with stable performance after bootstrap validation (corrected AUC 0.725). Firth analysis yielded consistent results. The SAGE-HF score showed progressive risk stratification, with predicted 30-day MACE ranging from 7.0% to 60.1% and maintained acceptable discrimination and calibration. Conclusions:The SAGE-HF score, incorporating age, sST2, and estimated glomerular filtration rate, demonstrated acceptable performance for predicting 30-day MACE after HF hospitalization and may support early risk stratification, pending external validation.
Background:Continuous vital sign monitoring ensures early detection, prevents intensive care unit (ICU) admissions, and improves patient outcomes. Continuous heart rate (HR) monitoring methods often require direct skin contact, which can lead to patient discomfort. The rising popularity of ballistocardiography (BCG) offers a promising, noncontact solution for continuous vital sign monitoring with improved patient comfort. Objective:This study aims to develop and validate a novel HR measurement algorithm leveraging convolutional neural networks (CNNs) and BCG signals for accurate, noncontact, and continuous HR monitoring. By integrating time-domain peak detection with short-time Fourier transform and CNN models, the proposed approach seeks to enhance HR measurement accuracy across diverse health care settings. The study follows the Food and Drug Administration (FDA)'s Good Machine Learning Practice guidelines and evaluates the algorithm's robustness, generalizability, and clinical applicability through extensive testing on a diverse dataset, ensuring improved patient comfort and early detection of clinical deterioration. Methods:The proposed algorithm combines time-domain peak detection with short-time Fourier transform and CNNs to enhance HR measurement from BCG signals. The CNN model developed was trained on 129,976 data points from 373 participants (HR range: 36-230 bpm), including ICU patients, and was tuned on 75,970 data points from 192 participants (HR range: 46-169 bpm), with HR obtained from clinical-grade electrocardiography devices to improve generalizability. The algorithm was tested on 70,211 data points from 205 participants, including ICU patients, across 5 independent studies to demonstrate robust performance against diverse settings, demographics, and comorbidities. The methodology is in compliance with the FDA's Good Machine Learning Practice for Medical Device Development: Guiding Principles. Results:The algorithm achieved a mean absolute error of under 3 bpm and a detection rate exceeding 80%, underscoring its robustness. The Bland-Altman analysis indicates high accuracy with a minimal bias of 0.25 and limits of agreement within 8.59 bpm. Additionally, the Pearson correlation coefficient of 0.97 from the Deming regression further demonstrates strong alignment with reference HR measurements, reinforcing its precision and reliability for clinical applications. Conclusions:This CNN-based algorithm presents a robust solution for contactless HR monitoring, addressing the limitations of prior methods in noise management and adaptability. Its demonstrated accuracy, particularly in real-world, noisy clinical environments, highlights its potential for broad application in patient monitoring and improved comfort.
Background:In cardiovascular care, illness and recovery affect both patients and their families, particularly within home-based remote patient management (RPM). A recent scientific statement from the American Heart Association highlighted the importance of family involvement, identifying digital technologies as a key enabling opportunity. Despite this, research into the needs of families and the implications of RPM remains limited. Objective:This study explored the lived experiences and unmet needs of patients with cardiovascular disease (CVD) and their relatives within RPM-supported cardiac care, using perioperative care and myocardial infarction as representative trajectories. Based on the identified gaps, we proposed a set of features for remote patient and family management (RPFM) interventions to address these needs. Methods:This qualitative study was conducted at a Dutch university hospital with over a decade of experience in RPM across CVD pathways. A human-centered design approach was employed, including focus groups with 24 participants (13 patients with CVD and 11 relatives). Data analysis followed a framework analysis approach, combining inductive theme identification with deductive mapping onto existing frameworks. Care experiences and unmet needs were identified inductively and segmented along the Family Systems Illness Model's phases of illness. The needs were subsequently categorized deductively into the domains of the Supportive Care Framework. Based on these user-informed insights, RPFM features were generated through author ideation and internal team consensus. Finally, these experiences, needs, and features were synthesized into a visual journey map illustrating key care moments across 3 phases: preadmission, admission, and postadmission. Results:We identified 47 unmet needs across 6 domains: informational (n=13), psychoemotional (n=13), social (n=7), physical (n=7), practical (n=6), and spiritual (n=1). The most significant gaps, described as "black holes" in care and support, emerged during the preoperative waiting period and early home recovery, which were characterized by a lack of information and psychoemotional support. To address the identified unmet needs, we generated 33 RPFM intervention feature ideas. These RPFM features were mapped across care phases and grouped into 7 categories: dynamic pathway navigation (n=8), on-demand support and information (n=5), family well-being modules (n=5), medical translation and consultation support (n=4), safety and assurance monitoring (n=4), collaborative lifestyle management (n=4), and peer support platform (n=3). Conclusions:This study identified experience gaps in RPM-supported cardiac care. It showed that most unmet needs for care and support extend beyond hospital admission and discharge and that most health behaviors and recovery occur in the home context, within the patient's relational ecosystem with loved ones. The proposed RPFM features provide preliminary directions for exploration toward a transition from individually focused monitoring to inclusive, family-centered care and management.
Background:Heart rate variability (HRV) is a noninvasive indicator of autonomic nervous system activity that is increasingly used for health and performance monitoring. Digital and mobile technologies are increasingly providing opportunities for remote HRV monitoring outside of laboratory-based settings. Objective:This study aimed to describe the landscape of mobile apps that measure, analyze, and provide feedback on HRV, with a focus on how HRV is measured, analyzed, interpreted, and communicated to users. A secondary aim was to assess the transparency of these apps, including the extent to which they disclose the evidence underpinning their HRV metrics and feedback. Methods:This study was an app store search and content analysis. Searches were conducted in the Google Play Store and Apple iTunes Store. Apps were eligible for inclusion if they had functionality to record, analyze, or provide feedback on HRV and were available in English. Data were extracted from app descriptions, screenshots, websites, and, where necessary, contact with developers. Data were extracted on app metadata (developer, release and update dates, and pricing), alongside information about HRV measurement, analysis, and feedback. This included the type of sensor used; HRV measurement characteristics (sensor placement, recording duration, and body position or standardization procedures); methods to calculate and interpret HRV (ie, metrics derived and how they were interpreted for users); and additional app functionality such as reminders, the ability to log self-reported stressors, and the type of feedback or guidance provided based on HRV. We used previously published criteria for assessing the quality of information on the internet, which included authorship, scientific attribution, currency of updates, and data privacy. Results:Of 746 apps identified, 206 met eligibility criteria. Of these, 132 were primary measurement apps, 59 were aggregators, and 15 were hybrid. Photoplethysmogram was the most common sensing modality (n=117, 56.8%), followed by multiple sensors (n=60, 29.1%). Full data extraction across app metadata and HRV measurement and analysis data was only achievable for 93 (45.1%) apps, representing a transparent subset with sufficient available information for content analysis. The most commonly reported HRV metrics were root mean square of successive differences (n=51) and SD of normal-to-normal intervals (n=48), while frequency-domain power (n=22) and low frequency to high frequency ratios (n=15) were less common. Most apps presented data as personalized trends or individualized ranges (76/93, 81.7%), emphasizing user-specific context rather than isolated values. Although 86% (80/93) offered contextual guidance (eg, readiness or recovery scores), many relied on proprietary algorithms that were not transparently described, limiting independent assessment of how these scores were derived and validated. Conclusions:Consumer HRV apps are widely available but vary considerably in how data are collected, processed, and contextualized. While many offer personalized trends and guidance, methodological transparency is often limited, particularly regarding the proprietary algorithms underlying the feedback scores.
Background:Feasible and potentially scalable strategies are needed to address the growing cardiovascular disease (CVD) risk among people living with HIV. Bidirectional automated texting (BAT) programs that remind and encourage adherence to evidence-based CVD-reducing interventions represent a potentially scalable strategy, but data on their feasibility are lacking. Objective:The goal of the study was to determine whether participant sociodemographic factors and technological constraints influenced engagement with a BAT CVD prevention program by people living with HIV. We conducted an observational cohort analysis embedded within a stepped-wedge cluster randomized trial. Methods:The BAT program was designed to address the "Million Hearts" ABCS (aspirin therapy, blood pressure control, cholesterol management, and smoking cessation) of cardiovascular health. The parent study was a stepped-wedge randomized trial that rolled out in 3 wedges across 8 practice sites that provided care to people living with HIV. Participants received and could engage with the text messages weekly using their own phones during the study period. We used a zero-inflated negative binomial model to identify factors associated with participants sending text messages during the study. Results:Of the 471 participants, 94% owned a smartphone capable of text messaging, and 70% reported monthly incomes less than US $1500. Overall, 60.3% (n=284) engaged with the BAT program at least once. Regarding texting behavior, participants aged ≥65 years were more likely to send a text than those aged <50 years (P=.047), although age did not influence the number of texts sent. White participants showed lower texting intensity than Black participants (incidence rate ratio 0.69; P=.04). Conclusions:Overall, 60.3% (n=284) of the participants in the study engaged with BAT at least once. The BAT intervention for ABCS appears to be a feasible intervention for people living with HIV. Only a few factors were associated with sending a text or with the number of text messages sent.
Abstract BackgroundAtrial fibrillation (AF), the most prevalent cardiac arrhythmia, affects 2% to 4% of the global adult population and is associated with an increased risk of stroke. Early diagnosis of AF and atrial flutter (AFL) is crucial due to their association with stroke risk and the challenge posed by their often asymptomatic and episodic nature. Traditional electrocardiogram (ECG) interpretation requires substantial expert input and can be challenging, especially with poor-quality ECGs. ObjectiveThis study aimed to evaluate the performance of a deep neural network (DNN) model in detecting AF/AFL from a large, heterogeneous set of long-term ambulatory ECG recordings, including clinical data collected over 6 months at a university hospital, and assess its effectiveness in a setting reflecting the diversity and complexity of real-world clinical data. MethodsThe research combined public datasets totaling 10,248 patients, ECG data from our previous studies (648 patients), and authentic long-term ECG recordings from 4346 patients at Kuopio University Hospital for development of the DNN model. Its clinical accuracy and generalizability were assessed using a separate test dataset consisting of 1039 pseudonymized long-term ECG recordings from 1010 patients, all thoroughly reviewed and annotated by experts. ResultsThe DNN model demonstrated high effectiveness, achieving 96.4% sensitivity and more than 99.99% specificity for time-level AF and AFL detection. At the recording level, it identified AF and AFL with 100% sensitivity and 98.77% specificity, producing false positives in only 1.2% (11/897) of recordings, of which 81.8% (9/11) had other non–AF/AFL arrhythmias. The model maintained high performance across diverse patient characteristics, including varying ages, comorbidities, coexisting arrhythmias, and poor-quality ECG recordings. ConclusionsThe results demonstrate that the proposed DNN model may support automated screening for AF and AFL in long-term ambulatory ECG recordings and may reduce manual review workload in clinical practice.
Unlabelled:Synthetic data offer significant potential for cardiology research by enabling data sharing, preserving privacy, and supporting machine learning model development. By generating artificial patient records that reflect real-world distributions, synthetic data can accelerate clinical research, improve model performance for rare cardiovascular conditions, and facilitate transnational collaborations that would otherwise be restricted by data-sharing barriers. Despite these advantages, the increasing use of synthetic data raises important ethical, regulatory, and methodological concerns that remain insufficiently addressed. Key challenges include assessing the validity and generalizability of synthetic datasets, understanding their limitations in representing complex and heterogeneous patient populations, and preventing the amplification of existing biases in cardiovascular care. Current regulatory frameworks, including the General Data Protection Regulation (GDPR) and Health Insurance Portability and Accountability Act (HIPAA), do not fully address emerging risks such as reidentification and data leakage, and there is no harmonized guidance to govern the use of synthetic data as stand-alone evidence for medical device evaluation or therapeutic research. In this viewpoint, we argue that responsible integration of synthetic data in cardiology requires, first, clear differentiation between synthetic data as a privacy-preserving distributional substitute and synthetic data as a counterfactual simulation tool, and, second, fit-for-purpose governance frameworks that pair rigorous utility and fidelity testing with explicit, adversary-aware privacy evaluation before synthetic cohorts are accepted as evidence in research or product evaluation. A prerequisite for that governance is conceptual clarity about what synthetic data are being used for. Synthetic data in health care serve 2 fundamentally distinct roles that carry entirely different validity requirements, failure modes, and regulatory implications, yet they are routinely conflated. The first role is as a privacy-preserving distributional substitute: the goal is statistical fidelity to the real data distribution, so that analyses of the synthetic dataset yield results equivalent to those of the original. The second role is as a tool for counterfactual simulation: the goal is to generate data that could not have been observed, such as rare conditions, hypothetical interventions, or extrapolations to new populations. These 2 roles are methodologically distinct. A dataset that accurately reflects real-world distributions may be inadequate for extrapolating findings to underrepresented subgroups. Conversely, a simulator optimized for novel scenario generation may systematically diverge from real-world distributions. This distinction informs every subsequent discussion of validity, bias, and regulation in this viewpoint and our proposed 4 concrete actions for the cardiology research community, including mandatory 3-layer (fidelity, utility, and privacy) validation, systematic subgroup reporting, explicit intended-use scoping, and domain-specific acceptability thresholds for synthetic data-based evidence.
Background:Atrial fibrillation (AF) is the most common sustained heart rhythm disorder and is a challenging chronic disease to manage. Patients' daily self-care decisions are associated with improved AF outcomes, quality of life, and decreased hospital use and cost. However, many patients find these real-world or naturalistic decisions difficult, often because of their inherent complexity and ambiguity, coupled with the uncertainty of AF. Intervention research using technology to support AF self-care has largely emphasized making decisions with clinicians. Patients with AF are increasingly using consumer technology; yet, little is known about the use of technology by patients with AF in independent self-care decision-making. Addressing this gap will facilitate developing interventions that better leverage technology to enhance patients' naturalistic decision-making. Objective:This study aimed to explore the experiences of older adult patients in using technology to support self-care decision-making. Methods:Following an interpretive descriptive qualitative approach, older adult patients with AF were recruited from 3 specialty heart function clinics in a Western Canadian province to participate in 1 of 6 facilitated virtual focus groups for 1.5 hours. Patients were asked about their self-care decision-making since AF diagnosis, their AF-specific technology use and its use in making self-care decisions, their technology motivations, benefits, constraints, and other possibilities for use. Inductive thematic analysis was used to code the transcribed data, moving from open coding to clustering of common codes into categories, looking for patterns of meaning between and across categories to iteratively arrive at main themes and subthemes. Results:Thirty patients (n=15, 50% women) with AF (mean age, 73, SD 5.7 years; range 63-85 years) participated in the focus groups. Participants' experiences of using technology to make daily self-care decisions were highly variable but centered on its personalized use to meet their individualized needs, preferences, and life context. The personalizing process of technology use in decision-making was characterized by three themes: (1) beginning technology use in their own times and ways, during their AF trajectory-pre-, at the time, at some point AF post diagnosis, and could be either self-initiated and/or provider recommended or influenced; (2) developing patterns of AF self-care decision-making using technology, including establishing their personal baseline, keeping out of the danger zone, watchful waiting, and seeking decision-making support; and (3) finding the place for technology in normalizing daily life, either settling on or limiting its use to normalize life. Conclusions:Findings expand understandings of naturalistic decision-making by elucidating the personalized process of technology use in AF self-care.
Background:Home-based cardiac rehabilitation (CR) using digital health technologies (ie, cardiac telerehabilitation [CTR]) has emerged as a practical alternative to conventional center-based CR, particularly during and after the COVID-19 pandemic. However, maintaining sustained participation in CR remains challenging. Gamification holds the potential to enhance motivation and adherence in CR, but its role in CTR for patients with acute coronary syndrome (ACS) remains under-studied. Objective:This feasibility study evaluated the feasibility, acceptability, and safety of a combination of a gamification-enabled smartphone app and a smartwatch supporting CTR after ACS. We focused specifically on participation motivation, adherence to prescribed exercise intensity, and short-term physiological outcomes. Methods:This single-arm, pre-post intervention study was conducted at 2 Japanese institutions. Sixteen patients diagnosed with ST-elevated myocardial infarction or non-ST-elevated myocardial infarction and undergoing percutaneous coronary intervention were enrolled after discharge. Each patient received a smartphone and smartwatch connected to the Shin-po Kei app, through which participants earned points when exercising within their target anaerobic threshold heart rate (HR) range (±10 bpm). The 1-month intervention prescribed walking for 30 minutes or longer 3 to 5 days or more per week, at an intensity corresponding to a Borg scale score of 11 to 13. The primary end point was the number of app log-in days divided by the number of intervention days (app use rate). Secondary end points included the target HR exercise adherence rate (defined as the proportion of patients who exercised for 30 minutes within the target HR range on 3 or more days per week), changes in peak oxygen consumption, and changes in perceived exertion using a visual analog scale. Results:The mean patient age was 62 (SD 13) years, and 94% (15/16) were male. The mean intervention duration was 29 (SD 6) days, and the overall app use rate was 59.9%. The target HR exercise adherence rate was 50%. Peak oxygen consumption increased from 18.4 (SD 4.1) to 21.1 (SD 5.4) mL/kg/min, and the mean visual analog scale score for exercise-related hassle reduced from 57 (SD 33) to 32 (SD 26). No adverse events related to app use were identified, and 1 hospitalization unrelated to the intervention occurred during the study period. Conclusions:This feasibility study suggests that short-term engagement with a smartwatch-linked gamified smartphone app is achievable in CTR after ACS. However, adherence to predefined gamified exercise standards was lower than app use rates, suggesting that refinement of reward rules and target HR thresholds may be necessary.
Abstract BackgroundMost studies assessing digital interventions for people with heart failure (HF) focus on clinical outcomes, and few include patient perspectives. Understanding patient experiences of the use of a digital HF platform along with community health worker (CHW) care as part of a digitally enabled CHW intervention can inform management of HF at home and improve the postdischarge phase of care. ObjectiveThis study aimed to identify patient perceptions related to the use of a digitally enabled CHW intervention. MethodsThis qualitative study included interviews with adults (aged ≥18 years) with HF who were assigned to the intervention arm of a pilot randomized controlled trial from September 2022 to June 2023. For 30 days after hospital discharge, intervention participants were paired with a CHW and instructed to use a digital platform that tracked biometrics (eg, heart rate, oxygenation, blood pressure, body weight, steps taken, and symptoms) and offered educational videos. In-depth interviews were conducted after the 30-day intervention was complete (between 31 and 45 days after hospital discharge). Key interview domains included barriers and facilitators to the intervention, use of remote monitoring in HF, and the role of CHWs in HF home care. ResultsInterviews with participants (N=19; mean age 62.1, SD 15.1 years) yielded five key themes: (1) the combined intervention was well received, and CHWs made the use of the digital platform more approachable; (2) the digital platform enhanced HF knowledge and confidence in self-care; (3) digital platform use was easy to integrate into daily routines; (4) in addition to assisting with navigation of unmet social needs (eg, transportation, insurance benefits, and food access), CHWs provided emotional support and increased motivation for clinical care plan adherence and platform use; and (5) connectivity issues and other technical challenges occurred with digital platform use. ConclusionsThe digital platform was easily integrated into patients’ daily routines. CHWs played a key role in making the platform more approachable for participant use. Further research is needed to better understand the impact of this intervention in larger HF populations over more extended time intervals.
Background:Social robots (SRs) are innovative tools in health care, offering both medical and psychological support for patients with heart failure (HF). For successful implementation, patient acceptability of SRs is crucial. Living in urban areas and having a lower comorbidity burden have been linked to higher acceptability; however, the role of psychological factors remains underexplored. Objective:This study aimed to examine the associations between negative (eg, depression and anxiety) and positive (eg, optimism) psychological factors and personality traits (eg, openness and extraversion) with SR acceptability in patients with HF. Methods:Patients with HF watched brief videos about SRs and completed validated measures of depressive symptoms (Patient Health Questionnaire-9), anxiety symptoms (Generalized Anxiety Disorder-7), positive psychological well-being (Brief Inventory of Thriving), and personality traits (Ten-Item Personality Inventory). Medical information was extracted from patients' records. SR acceptability was assessed using the Unified Theory of Acceptance and Use of Technology (UTAUT). Pearson correlations and multiple linear regression, adjusted for age, sex, smart technology experience, urbanicity, and comorbidities, were conducted. Results:Of the 101 patients (women: n=36, 35.6%, mean age 68, SD 10 y), 23% (23/101) scored in the clinical range for depression, and 17% (17/101) scored in the clinical range for anxiety. Well-being scores were moderate, and conscientiousness and agreeableness were the most common. UTAUT behavioral intention was moderate; 69% (67/97) of participants were likely to use an SR if available. Well-being scores correlated positively with SR acceptability in 4 of 5 UTAUT subscales, whereas no significant bivariate associations were observed for psychological distress or personality traits. In the multiple regression models, higher Brief Inventory of Thriving scores were associated with increased SR acceptability, including UTAUT facilitating conditions (B=0.17; P=.01) and behavioral intention (B=0.17; P=.04), independent of depressive and anxiety symptoms. Conclusions:Psychological well-being is associated with determinants of SR acceptability in patients with HF, while psychological distress and personality traits are not associated with these determinants. These patient-level factors ought to be examined more closely before SR implementation.
Background:Mobile health (mHealth) interventions are increasing in popularity for the management of heart failure and coronary artery disease. The use of these interventions is dependent on rates of smartphone ownership. It is estimated that approximately 90% of the Australian adult population owns a smartphone; however, international studies suggest that smartphone ownership is significantly lower in patient populations, ranging from 34% to 91%. Smartphone ownership in patients with cardiovascular disease has not previously been examined. Objective:This study aimed to examine and compare pre-COVID-19 and post-COVID-19 pandemic smartphone ownership rates of inpatients admitted with coronary artery disease or heart failure. Methods:Data from prescreening logs of 2 multicenter randomized controlled trials, TeleClinical Care (TCC)-Pilot and TCC-Cardiac, were reviewed. TCC-Pilot recruited patients between February 2019 and March 2020. This formed the pre-COVID-19 cohort, with 377 patients screened who lived in Sydney, had a qualifying hospital admission, and had information regarding their phone ownership status. TCC-Cardiac recruited patients from July 2021 to February 2023, with 718 patients meeting the criteria and forming the post-COVID-19 cohort. Supplemental patient demographic and medical history data were collected from the electronic medical record. Results:In the pre-COVID-19 cohort (N=377), 194 (51.5%) patients owned smartphones, 79 (21%) owned phones that were incompatible with the mHealth intervention, and the remaining 104 (27.6%) did not own a mobile phone. Smartphone owners were predominantly male (P<.001) and more often had private health insurance (P=.002). In the post-COVID-19 cohort (N=718), 366 (51%) patients owned smartphones, 106 (14.8%) owned incompatible phones, and the remaining 246 (34.3%) did not own any mobile phone. In both cohorts, younger patients were more likely to own smartphones (P<.001). Multiple comorbidities were associated with not owning a phone. Conclusions:Smartphone ownership accounted for just over 50% of the patients in this population. It was less common among older adults, patients with comorbidities, and those with markers of lower socioeconomic status. This needs to be considered when delivering mHealth interventions.
Background:Accurate identification of clinical symptoms and signs (S&S) is essential for the early detection of high-burden cardiorespiratory conditions, including lung cancer, chronic obstructive pulmonary disease, and heart failure. Although symptom data play a central role in diagnostic reasoning and predictive modeling, most S&S information remains embedded in unstructured electronic health record notes, limiting their use in automated phenotyping, surveillance, and clinical decision support. Traditional natural language processing systems struggle with domain variability and contextual nuance in clinical text. Recent advances in large language models (LLMs) offer a promising alternative, yet challenges remain in hallucinations, overinference, and safe deployment. This study evaluated whether locally deployed open-source models could reliably extract cardiorespiratory S&S and map them to ICD-10-CM (International Classification of Diseases, Tenth Revision, Clinical Modification) codes using optimized prompting strategies. Objective:This study aims to assess the accuracy of open-source LLMs in extracting explicitly stated cardiorespiratory S&S from clinical notes and mapping them to ICD-10-CM codes (R00-R09) and to compare performance across 4 prompt-engineering strategies, including a multimodule LLM framework. Methods:A total of 593 clinical notes from the MTSamples database were manually reviewed, with 93 notes used for prompt development and comparison using Llama 3.3-70B, and 500 notes used as testing data for the final best prompt setting using both Llama 3.3-70B and gpt-oss-120B. Four prompting conditions were evaluated: (1) instruction-only, (2) ICD-10-CM definition-based prompts, (3) assumption-free prompts, and (4) a multimodule LLM framework with postprocessing. Performance was measured using precision, recall, and F1-score for both S&S extraction and ICD-10-CM code generation. Results:Across all prompt strategies, model performance improved as more structure and constraints were added. Instruction-only prompting demonstrated high recall but poor precision. Incorporating ICD-10-CM definitions improved coding accuracy, and assumption-free prompting further balanced precision and recall. The multimodule approach with postprocessing achieved the highest performance during prompt development. On the independent test corpus, entity-level microaveraged evaluation showed that gpt-oss-120B outperformed Llama 3.3-70B in both tasks. For S&S extraction, Llama 3.3-70B achieved a precision of 0.63, a recall of 0.86, and an F1-score of 0.73, whereas gpt-oss-120B achieved a precision of 0.89, a recall of 0.87, and an F1-score of 0.88. For ICD-10-CM code mapping, Llama 3.3-70B achieved a precision of 0.59, a recall of 0.83, and an F1-score of 0.69, whereas gpt-oss-120B achieved a precision of 0.90, a recall of 0.84, and an F1-score of 0.87. Conclusions:Locally deployed LLMs, when paired with optimized prompting and multimodule orchestration, can accurately extract cardiorespiratory S&S and generate ICD-10-CM codes from unstructured clinical notes. This approach increases the level of data safety by enabling on-premises processing without external data transmission and demonstrates strong potential for scalable, domain-adaptive symptom extraction pipelines in biomedical informatics. Future work should expand datasets and evaluate generalizability across clinical domains.
Background: Regular physical activity is critical for preventing secondary stroke following a stroke or transient ischemic attack (TIA). Although mobile health (mHealth) interventions have shown promise for promoting short-term increases in physical activity, evidence on their long-term effects and the mechanisms that support sustained behavior change remains limited. In particular, little is known about how people poststroke or TIA integrate the skills, knowledge, and habits gained through mHealth interventions into their daily lives once structured intervention support ends. Objective: This study aimed to explore the perceived barriers and facilitators of maintaining physical activity among individuals poststroke or TIA after completing an mHealth intervention. Methods: A qualitative approach was used, involving a strategic sample of 12 participants recruited after they had completed a 6-month mHealth intervention for people poststroke or TIA. The intervention included supervised physical therapy with mHealth support for physical exercises and behavior change (eg, counseling, goal-setting, and self-monitoring), followed by a 6-month postintervention period with access to self-managed mHealth support. To enable richness and depth in participants' accounts, the dataset consisted of 2 semistructured interviews conducted 3 months and 6 months after completing the intervention, along with participant-generated photographs. Between the interviews, participants took photos to reflect their experiences of maintaining physical activity after the intervention. These images served as prompts for dialog and reflection during the second interview. Data were analyzed using reflexive thematic analysis. Results: We generated 3 themes: building experience and knowledge to maintain physical activity, staying physically active in a complex life situation, and the meaning of context for maintaining physical activity after the intervention. Barriers and facilitators were conceptually integrated into a comprehensive understanding of maintaining physical activity, symbolized by the metaphor of a soap bubble, which requires persistence and a supportive environment to stay afloat. The complexity of participants' health and life situations created barriers to maintaining physical activity. To overcome these barriers, developing personal strategies within a supportive social and physical context was crucial. This development was facilitated by the mHealth intervention, which enabled knowledge acquisition of the principles of physical activity after stroke or TIA, along with increased awareness of its health benefits. Conclusions: mHealth interventions for people poststroke and TIA can serve as a catalyst for physical activity engagement and an enhancer of the knowledge and experience necessary to maintain physical activity after the intervention. Despite health-related and contextual barriers, participants may use personalized strategies, supported by their social and environmental context, to navigate these challenges. These insights highlight opportunities for future mHealth interventions to strengthen the interaction between individuals and their environment and empower tailored strategies for behavior change and long-term physical activity maintenance. Trial Registration: ClinicalTrials.gov NCT05111951; https://clinicaltrials.gov/study/NCT05111951 International Registered Report Identifier (IRRID): RR2-10.1186/s12883-023-03163-0
Background:Both poor sleep health and hypertensive disorders of pregnancy (HDP) are independent risk factors for cardiovascular disease. Whether poor postpartum sleep contributes to the relationship between HDP and future cardiovascular disease is unknown. This pilot study evaluated the feasibility and acceptability of studying sleep health using a wearable device (Oura ring) among mothers of young children. Objective:We evaluated indices of sleep health both objectively with the Oura ring and subjectively via questionnaires and qualitative interviews among mothers with and without a history of HDP. We also aimed to compare cardiovascular health (CVH) among mothers with vs without a prior history of HDP. Methods:Women who were 3 to 7 years after childbirth completed baseline questionnaires (the Pittsburgh Sleep Quality Index [PSQI], Mediterranean Eating Pattern for Americans tool, and 7-item International Physical Activity Questionnaire) and wore the Oura ring continuously for 2 weeks to monitor sleep. Optimal sleep health was defined as a sleep duration of ≥7 hours, a PSQI score of ≤5, sleep timing with a sleep midpoint between 2 AM and 4 AM, a sleep efficiency of >85%, and a sleep onset variability of <60 minutes. CVH was assessed using the Life's Essential 8 score, with 8 factors assessed via questionnaires (diet, physical activity, and nicotine exposure) and objective measurements (BMI, blood pressure, blood lipids, blood glucose, and sleep duration). Semistructured interviews were conducted. Results:In total, among 49 women, 28 (57%) with prior HDP and 21 (43%) with prior normotensive pregnancy were included, with an average of 4.9 (SD 1.2) years after delivery. Average sleep quality was suboptimal in both groups (mean PSQI score 7.0, SD 3.5 in the HDP group vs mean PSQI score 5.9, SD 2.4 in the control group; P=.22). Average sleep duration was suboptimal (6.7, SD 0.8 hours), with no difference between groups. Approximately half (n=23, 47%) had abnormal sleep timing, which was more common among those with a prior normotensive pregnancy. Sleep onset variability was high (mean 1.2, SD 0.5 hours), with no significant differences by HDP status. The mean CVH score fell within the moderate range (70.7, SD 12.8), with no differences between groups. The components of the CVH score that were lowest (ie, worst) among the entire cohort were diet (mean 37.3, SD 25.6) and BMI (mean 50.8, SD 35.4 kg/m2). Common barriers to sleep included parenting, work, and household responsibilities. The study met our criteria for the feasibility and acceptability of using the Oura ring to study sleep in this population. Conclusions:Among postpartum women, sleep health was suboptimal regardless of HDP history. Interventions to improve sleep and CVH should target all mothers during the first decade after childbirth.
Background:Acute kidney injury critically impacts outcomes in cardiogenic shock secondary to acute myocardial infarction (CS-AMI). Acute kidney injury is one of the strongest independent predictors of in-hospital mortality in CS-AMI. Despite evidence that early renal replacement therapy (RRT) initiation improves survival, comprehensive prediction models for RRT in this population remain lacking. Objective:This study aimed to develop and internally validate a Least Absolute Shrinkage and Selection Operator (LASSO) regression-based prediction model and clinical nomogram for in-hospital RRT in patients with CS-AMI. Methods:This multicenter retrospective cohort study included 1431 patients with CS-AMI from the Gulf Cardiogenic Shock (Gulf-CS) registry across 13 centers in 6 Gulf countries (2020-2022). LASSO logistic regression was applied to a training set (1071/1431, 80%) to select baseline predictors of RRT; performance was evaluated on a held-out testing set (268/1431, 20%). Internal validation included 10-fold cross-validation and bootstrapping (1000 iterations). Cluster-robust SEs accounted for center effects. The model was compared to a parsimonious model (age+creatinine clearance), and a clinical nomogram was developed. Results:Of 1431 patients, 190 (13.3%) required RRT. Patients requiring RRT were significantly older (mean 64.17, SD 12.14 y vs mean 59.75, SD 11.77 y; P<.001), with higher prevalences of diabetes mellitus (72.1% vs 61.9%; P=.008), peripheral arterial disease (11.6% vs 3.7%; P<.001), and prior cerebrovascular accident (11.1% vs 5.7%; P=.005). The RRT group had lower creatinine clearance (46 vs 72 mL/min; P<.001), higher baseline lactate (2.7 vs 2.1 mmol/L; P<.001), and more advanced Society for Cardiovascular Angiography and Interventions (SCAI) shock stages (stages D and E: 90.5% vs 64.9%; P<.001). LASSO selected 15 baseline predictors. The model achieved an area under the receiver operating characteristic curve (AUC) of 0.714 on the testing set, significantly outperforming the parsimonious model (AUC: 0.624; P<.001). Bootstrap-corrected AUC was 0.745 (95% CI 0.730-0.756). In-hospital mortality was markedly higher in the RRT group (75.8% vs 38.8%; P<.001), with longer hospital stay (10 vs 6 d; P<.001), more major bleeding (16.8% vs 7.3%; P<.001), and cerebrovascular accidents (11.1% vs 4.9%; P=.001). Conclusions:We have developed and internally validated a robust 15-variable nomogram (Gulf-CS-Nomogram) that accurately predicts the need for RRT in patients with CS-AMI using baseline data intended for use after coronary angiography. This tool may facilitate early nephrology consultation and timely RRT initiation to improve outcomes.
BACKGROUND:Most studies assessing digital interventions for people with heart failure (HF) focus on clinical outcomes and few include patient perspectives. Understanding patient experience with the use of a digital HF platform along with community health worker (CHW) care as part of a digitally-enabled CHW intervention, can inform management of HF at home and improve the post-discharge phase of care. OBJECTIVE:To identify patient perceptions related to the use of a digitally-enabled community health worker intervention. METHODS:This qualitative study includes interviews with adults (age ≥18) with HF who were assigned to the intervention arm of a pilot randomized controlled trial September 2022 through June 2023. For 30 days after hospital discharge, intervention participants were paired with a CHW and instructed to use a digital platform that tracked biometrics (e.g., heart rate, oxygenation, blood pressure, body weight, steps taken, symptoms) and offered educational videos. In-depth interviews were performed after the 30-day intervention was complete (between 31 to 45 days after hospital discharge). Key interview domains included: barriers and facilitators to the intervention, use of remote monitoring in HF, and the role of CHWs in HF home care. RESULTS:Interviews with participants (N=19; mean age 62.1, SD 15.1 years) yielded five key themes: (1) The combined intervention was well-received and CHWs made the use of the digital platform more approachable; (2) The digital platform enhanced HF knowledge and confidence in self-care; (3) Digital platform use was easy to integrate into daily routines; (4) In addition to assisting with navigation of unmet social needs (e.g., transportation, insurance benefits, food access), CHWs provided emotional support and increased motivation for clinical care plan adherence and platform use; (5) Connectivity issues and other technical challenges occurred with digital platform use. CONCLUSIONS:The digital platform was easily integrated into patient daily routines. CHWs played a key role in making the platform more approachable for participant use. Further research is needed to better understand the impact of this intervention in larger HF populations over more extended time intervals. INTERNATIONAL REGISTERED REPORT:RR2-10.2196/55687.
Background:Telehealth has shown promise in enhancing care transitions and physical health outcomes in patients with cardiovascular disease. However, limited studies have explored its effect on functional status, psychological health, and rehospitalization, specifically in older patients undergoing coronary artery bypass grafting (CABG). Objective:This study aimed to evaluate the effectiveness of a telehealth intervention in improving functional status, reducing anxiety and depression, and decreasing rehospitalization rates compared with usual care among older patients undergoing CABG. Methods:The study was a 2-arm parallel randomized controlled trial. This was conducted in 2 phases. Phase 1 was conducted in the cardiac surgical units at a university hospital in Bangkok, Thailand. Phase 2 involved following up with the participant at home 30 and 90 days after discharge. A total of 84 older adults undergoing CABG were randomly assigned to either the control group (n=42), which received usual care (discharge planning), or the intervention group (n=42), which received a telehealth intervention based on the transitional care model in addition to usual care. The telehealth intervention included home monitoring via the "Zip Heart" app and scheduled video consultations. The primary outcome was functional status, measured using the Thai version of the Enforced Social Dependency Scale. Secondary outcomes included anxiety and depression, assessed using the Thai Hospital Anxiety and Depression Scale, and rates of rehospitalization. Data were collected at baseline, 30, and 90 days after discharge. Analyses were conducted using an intention-to-treat approach, with missing outcome data handled using multiple imputation. Two-way repeated-measures ANOVA was used to evaluate group, time, and group-by-time interaction effects. Results:A total of 84 participants were randomized and included in the intention-to-treat analysis (intervention group, n=42; control group, n=42). At baseline, there were no statistically significant differences between the two groups. Significant group-by-time interactions were observed for functional status scores (F2,164=32.09, ηp²=.28; P<.001), anxiety (F2, 164=20.22, ηp²=.2; P<.001), and depression (F2,164=16.81, ηp²=.17; P<.001). The intervention group demonstrated significantly greater improvements in functional status and greater reductions in anxiety and depression at both 30 and 90 days after discharge compared to the control group (all P<.001). Additionally, rehospitalization rates were significantly lower in the intervention group at 30 days (Z=2.77; P=.006) and between 31 and 90 days post discharge (Z=2.31; P=.02). Conclusions:The Telehealth intervention is effective in improving functional and psychological outcomes and reducing rehospitalization rates among older patients undergoing CABG. Integrating telehealth into usual care can support recovery and enhance continuity of care.
