Automated interpretations of single-lead electrocardiograms predict incident atrial fibrillation: The VITAL-AF Trial

BACKGROUND

Single-lead electrocardiograms (1L ECG) are increasingly used for atrial fibrillation (AF) detection. Automated 1L ECG interpretation may possess prognostic value for future AF among cases where screening does not result in a short-term AF diagnosis.

OBJECTIVE

Investigate the association between automated 1L ECG interpretation and incident AF.

METHODS

VITAL-AF was a randomized controlled trial investigating the effectiveness of screening for AF using 1L ECGs. For the present study, participants were divided into four groups based on automated classification of 1L ECGs. Patients with prevalent AF were excluded. Associations between groups and incident AF were assessed using Cox proportional hazards models adjusted for risk factors. The start of follow-up was defined as 60 days after the latest 1L ECG (as some individuals had numerous screening 1L ECGs).

RESULTS

The study sample included: Never screened (n=16,306), Normal (n=10,914), Other (n=2,675), Possible AF (n=561). Possible AF had the highest AF incidence (5.91 per 100 person-years, 95% Confidence Interval [CI] 4.24-8.23). Possible AF was associated with greater hazard of incident AF compared to Normal (adjusted Hazard Ratio (2.48, 95% CI 1.66-3.71). Other was associated with greater hazard of incident AF when compared to Normal (1.41, 95% CI 1.04-1.90).

CONCLUSIONS

In patients undergoing AF screening with 1L ECGs without prevalent AF or AF within 60 days of screening, presumptive positive and indeterminate 1L ECG interpretations were associated with future AF. Abnormal 1L ECGs may identify individuals at higher risk for future AF.

Natural Language Processing for Adjudication of Heart Failure in a Multicenter Clinical Trial: A Secondary Analysis of a Randomized Clinical Trial

Importance

The gold standard for outcome adjudication in clinical trials is medical record review by a physician clinical events committee (CEC), which requires substantial time and expertise. Automated adjudication of medical records by natural language processing (NLP) may offer a more resource-efficient alternative but this approach has not been validated in a multicenter setting.

Objective

To externally validate the Community Care Cohort Project (C3PO) NLP model for heart failure (HF) hospitalization adjudication, which was previously developed and tested within one health care system, compared to gold-standard CEC adjudication in a multicenter clinical trial.

Design, Setting, and Participants

This was a retrospective analysis of the Influenza Vaccine to Effectively Stop Cardio Thoracic Events and Decompensated Heart Failure (INVESTED) trial, which compared 2 influenza vaccines in 5260 participants with cardiovascular disease at 157 sites in the US and Canada between September 2016 and January 2019. Analysis was performed from November 2022 to October 2023.

Exposures

Individual sites submitted medical records for each hospitalization. The central INVESTED CEC and the C3PO NLP model independently adjudicated whether the cause of hospitalization was HF using the prepared hospitalization dossier. The C3PO NLP model was fine-tuned (C3PO + INVESTED) and a de novo NLP model was trained using half the INVESTED hospitalizations.

Main Outcomes and Measures

Concordance between the C3PO NLP model HF adjudication and the gold-standard INVESTED CEC adjudication was measured by raw agreement, κ, sensitivity, and specificity. The fine-tuned and de novo INVESTED NLP models were evaluated in an internal validation cohort not used for training.

Results

Among 4060 hospitalizations in 1973 patients (mean [SD] age, 66.4 [13.2] years; 514 [27.4%] female and 1432 [72.6%] male]), 1074 hospitalizations (26%) were adjudicated as HF by the CEC. There was good agreement between the C3PO NLP and CEC HF adjudications (raw agreement, 87% [95% CI, 86-88]; κ, 0.69 [95% CI, 0.66-0.72]). C3PO NLP model sensitivity was 94% (95% CI, 92-95) and specificity was 84% (95% CI, 83-85). The fine-tuned C3PO and de novo NLP models demonstrated agreement of 93% (95% CI, 92-94) and κ of 0.82 (95% CI, 0.77-0.86) and 0.83 (95% CI, 0.79-0.87), respectively, vs the CEC. CEC reviewer interrater reproducibility was 94% (95% CI, 93-95; κ, 0.85 [95% CI, 0.80-0.89]).

Conclusions and Relevance

The C3PO NLP model developed within 1 health care system identified HF events with good agreement relative to the gold-standard CEC in an external multicenter clinical trial. Fine-tuning the model improved agreement and approximated human reproducibility. Further study is needed to determine whether NLP will improve the efficiency of future multicenter clinical trials by identifying clinical events at scale.

Deep learned representations of the resting 12-lead electrocardiogram to predict at peak exercise

AIMS

To leverage deep learning on the resting 12-lead electrocardiogram (ECG) to estimate peak oxygen consumption (V˙O2peak) without cardiopulmonary exercise testing (CPET).

METHODS AND RESULTS

V ˙ O 2 peak estimation models were developed in 1891 individuals undergoing CPET at Massachusetts General Hospital (age 45 ± 19 years, 38% female) and validated in a separate test set (MGH Test, n = 448) and external sample (BWH Test, n = 1076). Three penalized linear models were compared: (i) age, sex, and body mass index ('Basic'), (ii) Basic plus standard ECG measurements ('Basic + ECG Parameters'), and (iii) basic plus 320 deep learning-derived ECG variables instead of ECG measurements ('Deep ECG-V˙O2'). Associations between estimated V˙O2peak and incident disease were assessed using proportional hazards models within 84 718 primary care patients without CPET. Inference ECGs preceded CPET by 7 days (median, interquartile range 27-0 days). Among models, Deep ECG-V˙O2 was most accurate in MGH Test [r = 0.845, 95% confidence interval (CI) 0.817-0.870; mean absolute error (MAE) 5.84, 95% CI 5.39-6.29] and BWH Test (r = 0.552, 95% CI 0.509-0.592, MAE 6.49, 95% CI 6.21-6.67). Deep ECG-V˙O2 also outperformed the Wasserman, Jones, and FRIEND reference equations (P < 0.01 for comparisons of correlation). Performance was higher in BWH Test when individuals with heart failure (HF) were excluded (r = 0.628, 95% CI 0.567-0.682; MAE 5.97, 95% CI 5.57-6.37). Deep ECG-V˙O2 estimated V˙O2peak <14 mL/kg/min was associated with increased risks of incident atrial fibrillation [hazard ratio 1.36 (95% CI 1.21-1.54)], myocardial infarction [1.21 (1.02-1.45)], HF [1.67 (1.49-1.88)], and death [1.84 (1.68-2.03)].

CONCLUSION

Deep learning-enabled analysis of the resting 12-lead ECG can estimate exercise capacity (V˙O2peak) at scale to enable efficient cardiovascular risk stratification.

Characteristics and Attitudes of Wearable Device Users and Nonusers in a Large Health Care System

BACKGROUND

Consumer wearable devices with health and wellness features are increasingly common and may enhance disease detection and management. Yet studies informing relationships between wearable device use, attitudes toward device data, and comprehensive clinical profiles are lacking.

METHODS AND RESULTS

WATCH-IT (Wearable Activity Tracking for Comprehensive Healthcare-Integrated Technology) studied adults receiving longitudinal primary or ambulatory cardiovascular care in the Mass General Brigham health care system from January 2010 to July 2021. Participants completed a 20-question electronic survey about perceptions and use of consumer wearable devices, with responses linked to electronic health records. Multivariable logistic regression was used to identify factors associated with device use. Among 214 992 individuals receiving longitudinal primary or cardiovascular care with an active electronic portal, 11 121 responded (5.2%). Most respondents (55.8%) currently used a wearable device, and most nonusers (95.3%) would use a wearable if provided at no cost. Although most users (70.2%) had not shared device data with their doctor previously, most believed it would be very (20.4%) or moderately (34.4%) important to share device-related health information with providers. In multivariable models, older age (odds ratio [OR], 0.80 per 10-year increase [95% CI, 0.77-0.82]), male sex (OR, 0.87 [95% CI, 0.80-0.95]), and heart failure (OR, 0.75 [95% CI, 0.63-0.89]) were associated with lower odds of wearable device use, whereas higher median income (OR, 1.08 per 1-quartile increase [95% CI, 1.04-1.12]) and care in a cardiovascular medicine clinic (OR, 1.17 [95% CI, 1.05-1.30]) were associated with greater odds of device use.

CONCLUSIONS

Among patients in primary and cardiovascular medicine clinics, consumer wearable device use is common, and most users perceive value in wearable health data.

Evaluation of in vitro anticancer potential of pharmacological ethanolic plant extracts Acacia modesta and Opuntia monocantha against liver cancer cells

Abstract Acacia modesta (AM) and Opuntia monocantha (OM) are distributed in Pakistan, Afghanistan and India. Both of these plants have different pharmacological properties. This study was designed to evaluate anticancer potential of Acacia modesta (AM) and Opuntia monocantha (OM). Liver cancer cell line HepG2 was used for assessment of anticancer activity. For the evaluation of anti-proliferative effects, cell viability and cell death in all groups of cells were evaluated via MTT, crystal violet and trypan blue assays. For the evaluation of apoptosis ELISA of p53 performed. Furthermore, LDH assay to find out the ability of malignant cells to metabolize pyruvate to lactate and antioxidant enzymes activity (GSH, CAT and SOD) at the end HPLC was performed to find active compound of AM and OM. Cytotoxicity (MTT), Viability assays (trypan blue, crystal viability, MUSE analysis) showed more dead, less live cells in plant treated groups with increase of concentration. Scratch assay for the anti-migratory effect of these plants showed treated groups have not ability to heal scratch/wound. ELISA of p53 for cellular apoptosis showed more release of p53 in treated groups. Antioxidant assay via glutathione (GSH), superoxide dismutase (SOD), catalase (CAT) showed less anti-oxidative potential in treated cancer groups. LDH assay showed more lactate dehydrogenase release in treated groups compared with untreated. HPLC analysis showed the presence of phytochemicals such as steroids, alkaloids, phenols, flavonoids, saponins, tannins, anthraquinone and amino acids in AM and OM plant extracts. Based on all these findings, it can be concluded that ethanolic extracts of Acacia modesta and Opuntia monocantha have promising anti-cancer potential.

Utility of Left and Right Ventricular Strain in Arrhythmogenic Right Ventricular Cardiomyopathy: A Prospective Multicenter Registry

BACKGROUND

Imaging evaluation of arrhythmogenic right ventricular cardiomyopathy (ARVC) remains challenging. Myocardial strain assessment by echocardiography is an increasingly utilized technique for detecting subclinical left ventricular (LV) and right ventricular (RV) dysfunction. We aimed to evaluate the diagnostic and prognostic utility of LV and RV strain in ARVC.

METHODS

Patients with suspected ARVC (n = 109) from a multicenter registry were clinically phenotyped using the 2010 ARVC Revised Task Force Criteria and underwent baseline strain echocardiography. Diagnostic performance of LV and RV strain was evaluated using the area under the receiver operating characteristic curve analysis against the 2010 ARVC Revised Task Force Criteria, and the prognostic value was assessed using the Kaplan-Meier analysis.

RESULTS

Mean age was 45.3±14.7 years, and 48% of patients were female. Estimation of RV strain was feasible in 99/109 (91%), and LV strain was feasible in 85/109 (78%) patients. ARVC prevalence by 2010 ARVC Revised Task Force Criteria is 91/109 (83%) and 83/99 (84%) in those with RV strain measurements. RV global longitudinal strain and RV free wall strain had diagnostic area under the receiver operating characteristic curve of 0.76 and 0.77, respectively (both P<0.001; difference NS). Abnormal RV global longitudinal strain phenotype (RV global longitudinal strain > -17.9%) and RV free wall strain phenotype (RV free wall strain > -21.2%) were identified in 41/69 (59%) and 56/69 (81%) of subjects, respectively, who were not identified by conventional echocardiographic criteria but still met the overall 2010 ARVC Revised Task Force Criteria for ARVC. LV global longitudinal strain did not add diagnostic value but was prognostic for composite end points of death, heart transplantation, or ventricular arrhythmia (log-rank P=0.04).

CONCLUSIONS

In a prospective, multicenter registry of ARVC, RV strain assessment added diagnostic value to current echocardiographic criteria by identifying patients who are missed by current echocardiographic criteria yet still fulfill the diagnosis of ARVC. LV strain, by contrast, did not add incremental diagnostic value but was prognostic for identification of high-risk patients.

Deep Learning-Enabled Assessment of Left Heart Structure and Function Predicts Cardiovascular Outcomes

BACKGROUND

Deep learning interpretation of echocardiographic images may facilitate automated assessment of cardiac structure and function.

OBJECTIVES

We developed a deep learning model to interpret echocardiograms and examined the association of deep learning-derived echocardiographic measures with incident outcomes.

METHODS

We trained and validated a 3-dimensional convolutional neural network model for echocardiographic view classification and quantification of left atrial dimension, left ventricular wall thickness, chamber diameter, and ejection fraction. The training sample comprised 64,028 echocardiograms (n = 27,135) from a retrospective multi-institutional ambulatory cardiology electronic health record sample. Validation was performed in a separate longitudinal primary care sample and an external health care system data set. Cox models evaluated the association of model-derived left heart measures with incident outcomes.

RESULTS

Deep learning discriminated echocardiographic views (area under the receiver operating curve >0.97 for parasternal long axis, apical 4-chamber, and apical 2-chamber views vs human expert annotation) and quantified standard left heart measures (R2 range = 0.53 to 0.91 vs study report values). Model performance was similar in 2 external validation samples. Model-derived left heart measures predicted incident heart failure, atrial fibrillation, myocardial infarction, and death. A 1-SD lower model-left ventricular ejection fraction was associated with 43% greater risk of heart failure (HR: 1.43; 95% CI: 1.23-1.66) and 17% greater risk of death (HR: 1.17; 95% CI: 1.06-1.30). Similar results were observed for other model-derived left heart measures.

CONCLUSIONS

Deep learning echocardiographic interpretation accurately quantified standard measures of left heart structure and function, which in turn were associated with future clinical outcomes. Deep learning may enable automated echocardiogram interpretation and disease prediction at scale.

MSGene: Derivation and validation of a multistate model for lifetime risk of coronary artery disease using genetic risk and the electronic health record

Currently, coronary artery disease (CAD) is the leading cause of death among adults worldwide. Accurate risk stratification can support optimal lifetime prevention. We designed a novel and general multistate model (MSGene) to estimate age-specific transitions across 10 cardiometabolic states, dependent on clinical covariates and a CAD polygenic risk score. MSGene supports decision making about CAD prevention related to any of these states. We analyzed longitudinal data from 480,638 UK Biobank participants and compared predicted lifetime risk with the 30-year Framingham risk score. MSGene improved discrimination (C-index 0.71 vs 0.66), age of high-risk detection (C-index 0.73 vs 0.52), and overall prediction (RMSE 1.1% vs 10.9%), with external validation. We also used MSGene to refine estimates of lifetime absolute risk reduction from statin initiation. Our findings underscore the potential public health value of our novel multistate model for accurate lifetime CAD risk estimation using clinical factors and increasingly available genetics.

Accuracy and variability of cardiologist interpretation of single lead electrocardiograms for atrial fibrillation: The VITAL-AF trial

BACKGROUND

Screening for atrial fibrillation (AF) using consumer-based devices capable of producing a single lead electrocardiogram (1L ECG) is increasing. There are limited data on the accuracy of physician interpretation of these tracings. The goal of this study is to assess the sensitivity, specificity, confidence, and variability of cardiologist interpretation of point-of-care 1L ECGs.

METHODS

Fifteen cardiologists reviewed point-of-care handheld 1L ECGs collected from patients aged 65 years or older enrolled in the VITAL-AF clinical trial [NCT035115057] who underwent cardiac rhythm assessments with a 1L ECG using an AliveCor KardiaMobile device. Random sampling of 1L ECGs for cardiologist review was stratified by the AliveCor algorithm interpretation. A 12L ECG performed on the same day for clinical purposes was used as the gold standard. Cardiologists each reviewed a common sample of 200 1L ECG tracings and completed a survey associated with each tracing. Cardiologists were blinded to both the AliveCor algorithm and same day 12L ECG interpretation. For each tracing, study cardiologists were asked to assess the rhythm (sinus rhythm, AF, unclassifiable), report their assessment of the quality of the tracing, and rate their confidence in rhythm interpretation. The outcomes included the sensitivity, specificity, variability, and confidence in physician interpretation. Variables associated with each measure were identified using multivariable regression.

RESULTS

The average sensitivity for AF was 77.4% (range 50%-90.6%, standard deviation [SD]=11.4%) and the average specificity was 73.0% (range 41.3%-94.6%, SD = 15.4%). The mean variability was 30.8% (range 0%-76.2%, SD = 23.2%). The average reviewer confidence of 1L ECG rhythm assessment was 3.6 out of 5 (range 2.5-4.2, SD = 0.6). Patient and tracing factors associated with sensitivity, specificity, variability, and confidence were identified and included age, body mass index, and presence of artifact.

CONCLUSION

Cardiologist interpretation of point-of-care handheld 1L ECGs has modest diagnostic sensitivity and specificity with substantial variability for AF classification despite high confidence. Variability in cardiologist interpretation of 1L ECGs highlights the importance of confirmatory testing for diagnosing AF.

Clinical perspectives on the adoption of the artificial intelligence-enabled electrocardiogram

The 12‑lead electrocardiogram (ECG) is a common and inexpensive diagnostic modality available at scale. The ECG reflects electrical activity throughout the cardiac cycle and is increasingly recognized to contain rich signal relevant across the spectrum of human conditions. Recent work has demonstrated that artificial intelligence (AI)-based algorithms may be able to extract latent information from within the 12‑lead ECG to classify the presence of disease and even predict the development of future disease. Despite recent development of many AI-based ECG algorithms, comparably few are used in routine clinical practice. Therefore, there is a critical unmet need to identify and mitigate potential barriers to the real-world clinical implementation of AI algorithms. We propose that the adoption of the AI-enabled ECG may be increased by future efforts focused on three key principles: a) maximizing credibility, b) optimizing practicality, and c) establishing clinical utility. In this mini-review, we discuss recent notable work focused on these principles and provide suggestions for future directions. AI-enabled ECG analysis possesses substantial potential to transform current methods to prevent, diagnose, and treat human disease, but a greater emphasis on their real-world application is required to bring that potential to reality.

AORTA Gene: Polygenic prediction improves detection of thoracic aortic aneurysm

Background

Thoracic aortic disease is an important cause of morbidity and mortality in the US, and aortic diameter is a heritable contributor to risk. Could a polygenic prediction of ascending aortic diameter improve detection of aortic aneurysm?

Methods

Deep learning was used to measure ascending thoracic aortic diameter in 49,939 UK Biobank participants. A genome-wide association study (GWAS) was conducted in 39,524 participants and leveraged to build a 1.1 million-variant polygenic score with PRScs-auto. Aortic diameter prediction models were built with the polygenic score ("AORTA Gene") and without it. The models were tested in a held-out set of 4,962 UK Biobank participants and externally validated in 5,469 participants from Mass General Brigham Biobank (MGB), 1,298 from the Framingham Heart Study (FHS), and 610 participants from All of Us.

Results

In each test set, the AORTA Gene model explained more of the variance in thoracic aortic diameter compared to clinical factors alone: 39.9% (95% CI 37.8-42.0%) vs 29.2% (95% CI 27.1-31.4%) in UK Biobank, 36.5% (95% CI 34.4-38.5%) vs 32.5% (95% CI 30.4-34.5%) in MGB, 41.8% (95% CI 37.7-45.9%) vs 33.0% (95% CI 28.9-37.2%) in FHS, and 34.9% (95% CI 28.8-41.0%) vs 28.9% (95% CI 22.9-35.0%) in All of Us. AORTA Gene had a greater AUROC for identifying diameter ≥4cm in each test set: 0.834 vs 0.765 (P=7.3E-10) in UK Biobank, 0.808 vs 0.767 in MGB (P=4.5E-12), 0.856 vs 0.818 in FHS (P=8.5E-05), and 0.827 vs 0.791 (P=7.8E-03) in All of Us.

Conclusions

Genetic information improved estimation of thoracic aortic diameter when added to clinical risk factors. Larger and more diverse cohorts will be needed to develop more powerful and equitable scores.

Natural Language Processing for Adjudication of Heart Failure Hospitalizations in a Multi-Center Clinical Trial

Background

The gold standard for outcome adjudication in clinical trials is chart review by a physician clinical events committee (CEC), which requires substantial time and expertise. Automated adjudication by natural language processing (NLP) may offer a more resource-efficient alternative. We previously showed that the Community Care Cohort Project (C3PO) NLP model adjudicates heart failure (HF) hospitalizations accurately within one healthcare system.

Methods

This study externally validated the C3PO NLP model against CEC adjudication in the INVESTED trial. INVESTED compared influenza vaccination formulations in 5260 patients with cardiovascular disease at 157 North American sites. A central CEC adjudicated the cause of hospitalizations from medical records. We applied the C3PO NLP model to medical records from 4060 INVESTED hospitalizations and evaluated agreement between the NLP and final consensus CEC HF adjudications. We then fine-tuned the C3PO NLP model (C3PO+INVESTED) and trained a de novo model using half the INVESTED hospitalizations, and evaluated these models in the other half. NLP performance was benchmarked to CEC reviewer inter-rater reproducibility.

Results

1074 hospitalizations (26%) were adjudicated as HF by the CEC. There was high agreement between the C3PO NLP and CEC HF adjudications (agreement 87%, kappa statistic 0.69). C3PO NLP model sensitivity was 94% and specificity was 84%. The fine-tuned C3PO and de novo NLP models demonstrated agreement of 93% and kappa of 0.82 and 0.83, respectively. CEC reviewer inter-rater reproducibility was 94% (kappa 0.85).

Conclusion

Our NLP model developed within a single healthcare system accurately identified HF events relative to the gold-standard CEC in an external multi-center clinical trial. Fine-tuning the model improved agreement and approximated human reproducibility. NLP may improve the efficiency of future multi-center clinical trials by accurately identifying clinical events at scale.

Characteristics and Attitudes of Wearable Device Users and Non-Users in a Large Healthcare System

Introduction

Consumer wearable devices with health and wellness features are increasingly common and may enhance prevention and management of cardiovascular disease. However, the characteristics and attitudes of wearable device users versus non-users are poorly understood.

Methods

Wearable Activity Tracking for Comprehensive Healthcare-Integrated Technology (WATCH-IT) was a prospective study of adults aged ≥18 years receiving longitudinal primary or ambulatory cardiovascular care at one of eleven hospitals within the Mass General Brigham multi-institutional healthcare system between January 2010-July 2021. We invited patients, including wearable users and non-users, to participate via an electronic patient portal. Participants were asked to complete a 20-question survey regarding perceptions and use of consumer wearable devices. Responses were linked to electronic health record data. Multivariable logistic regression was used to identify factors associated with device use.

Results

Among 280,834 individuals receiving longitudinal primary or cardiovascular care, 65,842 did not have an active electronic portal or opted out of research contact. Of the 214,992 individuals sent a survey link, 11,121 responded (5.2%), comprising the WATCH-IT patient sample. Most respondents (55.8%) reported current use of a wearable device, and most non-users (95.3%) reported they would use a wearable device if provided at no cost. Although most users (70.2%) had not shared device data with their doctor previously, the majority believed it would be very (20.4%) or moderately (34.4%) important to share device-related health information with providers. In multivariable models, older age (odds ratio [OR] 0.80 per 10-year increase, 95% CI 0.77-0.82), male sex (0.87, 95% CI 0.80-0.95), and heart failure (0.75, 95% CI 0.63-0.89) were associated with lower odds of wearable device use, whereas higher median zip code income (1.08 per 1-quartile increase, 95% CI 1.04-1.12) and care in a cardiovascular medicine clinic (1.17, 95% CI 1.05-1.30) were associated with greater odds of device use. Nearly all respondents (98%) stated they would share device data with researchers studying health outcomes.

Conclusions

Within an electronically assembled cohort of patients in primary and cardiovascular medicine clinics with linkage to detailed health records, wearable device use is common. Most users perceive value in wearable data. Our platform may enable future study of the relationships between wearable technology and resource utilization, clinical outcomes, and health disparities.

Effect of clinic-based single-lead electrocardiogram rhythm assessment on oral anticoagulation prescriptions in patients with previously diagnosed atrial fibrillation

Background

Despite benefits of oral anticoagulation (OAC), many individuals with diagnosed atrial fibrillation (AF) do not receive OAC.

Objective

The purpose of this study was to assess whether cardiac rhythm assessment for AF impacted use of OAC in patients with previously diagnosed AF.

Methods

VITAL-AF was a cluster randomized controlled trial conducted in 16 primary care practices assessing the efficacy of AF rhythm assessment with single-lead electrocardiogram in routine care. Patients 65 years and older were offered rhythm assessment at visits. In this secondary analysis, we evaluated rhythm assessment uptake and compared initiation and discontinuation of OAC in patients with previously diagnosed AF from intervention and control arms over 1 year.

Results

The study included 4593 patients with previously diagnosed AF (2250 intervention; 2343 control). In the intervention arm, 2022 (89.9%) completed rhythm assessment (median 2 visits with rhythm assessment) and 40.1% had ≥1 "Possible AF" result. Initiation of OAC was similar in the intervention (17.7%) and control (19.1%) arms but was influenced by the rhythm assessment result: higher with a "Possible AF" (26.1%; adjusted odds ratio [aOR] 1.62; 95% confidence interval [CI] 1.04-2.51), and lower with a "Normal" result (9.9%; aOR 0.45; 95% CI 0.29-0.71) compared to control. OAC discontinuation was similar in the intervention (6.3%) and control (7.2%) arms, with lower discontinuation with a "Possible AF" result (3.8%; aOR 0.51; 95% CI 0.32-0.81).

Conclusions

Including patients with previously diagnosed AF in a point-of-care rhythm assessment strategy did not increase overall OAC use compared to the control arm. However, the rhythm assessment result influenced both initiation and discontinuation of OAC.

Genetic Susceptibility to Atrial Fibrillation Identified via Deep Learning of 12-Lead Electrocardiograms

BACKGROUND

Artificial intelligence (AI) models applied to 12-lead ECG waveforms can predict atrial fibrillation (AF), a heritable and morbid arrhythmia. However, the factors forming the basis of risk predictions from AI models are usually not well understood. We hypothesized that there might be a genetic basis for an AI algorithm for predicting the 5-year risk of new-onset AF using 12-lead ECGs (ECG-AI)-based risk estimates.

METHODS

We applied a validated ECG-AI model for predicting incident AF to ECGs from 39 986 UK Biobank participants without AF. We then performed a genome-wide association study (GWAS) of the predicted AF risk and compared it with an AF GWAS and a GWAS of risk estimates from a clinical variable model.

RESULTS

In the ECG-AI GWAS, we identified 3 signals (P<5×10-8) at established AF susceptibility loci marked by the sarcomeric gene TTN and sodium channel genes SCN5A and SCN10A. We also identified 2 novel loci near the genes VGLL2 and EXT1. In contrast, the clinical variable model prediction GWAS indicated a different genetic profile. In genetic correlation analysis, the prediction from the ECG-AI model was estimated to have a higher correlation with AF than that from the clinical variable model.

CONCLUSIONS

Predicted AF risk from an ECG-AI model is influenced by genetic variation implicating sarcomeric, ion channel and body height pathways. ECG-AI models may identify individuals at risk for disease via specific biological pathways.

Pacing-Induced Cardiomyopathy

Right ventricular (RV) pacing-induced cardiomyopathy (PICM) is typically defined as left ventricular systolic dysfunction resulting from electrical and mechanical dyssynchrony caused by RV pacing. RV PICM is common, occurring in 10-20% of individuals exposed to frequent RV pacing. Multiple risk factors for PICM have been identified, including male sex, wider native and paced QRS durations, and higher RV pacing percentage, but the ability to predict which individuals will develop PICM remains modest. Biventricular and conduction system pacing, which better preserve electrical and mechanical synchrony, typically prevent the development of PICM and reverse left ventricular systolic dysfunction after PICM has occurred.

Accelerometer-Derived "Weekend Warrior" Physical Activity and Incident Cardiovascular Disease

Importance

Guidelines recommend 150 minutes or more of moderate to vigorous physical activity (MVPA) per week for overall health benefit, but the relative effects of concentrated vs more evenly distributed activity are unclear.

Objective

To examine associations between an accelerometer-derived "weekend warrior" pattern (ie, most MVPA achieved over 1-2 days) vs MVPA spread more evenly with risk of incident cardiovascular events.

Design, Setting, and Participants

Retrospective analysis of UK Biobank cohort study participants providing a full week of accelerometer-based physical activity data between June 8, 2013, and December 30, 2015.

Exposures

Three MVPA patterns were compared: active weekend warrior (active WW, ≥150 minutes with ≥50% of total MVPA achieved in 1-2 days), active regular (≥150 minutes and not meeting active WW status), and inactive (<150 minutes). The same patterns were assessed using the sample median threshold of 230.4 minutes or more of MVPA per week.

Main Outcomes and Measures

Associations between activity pattern and incident atrial fibrillation, myocardial infarction, heart failure, and stroke were assessed using Cox proportional hazards regression, adjusted for age, sex, racial and ethnic background, tobacco use, alcohol intake, Townsend Deprivation Index, employment status, self-reported health, and diet quality.

Results

A total of 89 573 individuals (mean [SD] age, 62 [7.8] years; 56% women) who underwent accelerometry were included. When stratified at the threshold of 150 minutes or more of MVPA per week, a total of 37 872 were in the active WW group (42.2%), 21 473 were in the active regular group (24.0%), and 30 228 were in the inactive group (33.7%). In multivariable-adjusted models, both activity patterns were associated with similarly lower risks of incident atrial fibrillation (active WW: hazard ratio [HR], 0.78 [95% CI, 0.74-0.83]; active regular: 0.81 [95% CI, 0.74-0.88; inactive: HR, 1.00 [95% CI, 0.94-1.07]), myocardial infarction (active WW: 0.73 [95% CI, 0.67-0.80]; active regular: 0.65 [95% CI, 0.57-0.74]; and inactive: 1.00 [95% CI, 0.91-1.10]), heart failure (active WW: 0.62 [95% CI, 0.56-0.68]; active regular: 0.64 [95% CI, 0.56-0.73]; and inactive: 1.00 [95% CI, 0.92-1.09]), and stroke (active WW: 0.79 [95% CI, 0.71-0.88]; active regular: 0.83 [95% CI, 0.72-0.97]; and inactive: 1.00 [95% CI, 0.90-1.11]). Findings were consistent at the median threshold of 230.4 minutes or more of MVPA per week, although associations with stroke were no longer significant (active WW: 0.89 [95% CI, 0.79-1.02]; active regular: 0.87 [95% CI, 0.74-1.02]; and inactive: 1.00 [95% CI, 0.90-1.11]).

Conclusions and Relevance

Physical activity concentrated within 1 to 2 days was associated with similarly lower risk of cardiovascular outcomes to more evenly distributed activity.

Clinical and Genetic Atrial Fibrillation Risk and Discrimination of Cardioembolic From Noncardioembolic Stroke

BACKGROUND

Stroke is a leading cause of death and disability worldwide. Atrial fibrillation (AF) is a common cause of stroke but may not be detectable at the time of stroke. We hypothesized that an AF polygenic risk score (PRS) can discriminate between cardioembolic stroke and noncardioembolic strokes.

METHODS

We evaluated AF and stroke risk in 26 145 individuals of European descent from the Stroke Genetics Network case-control study. AF genetic risk was estimated using 3 recently developed PRS methods (LDpred-funct-inf, sBayesR, and PRS-CS) and 2 previously validated PRSs. We performed logistic regression of each AF PRS on AF status and separately cardioembolic stroke, adjusting for clinical risk score (CRS), imputation group, and principal components. We calculated model discrimination of AF and cardioembolic stroke using the concordance statistic (c-statistic) and compared c-statistics using 2000-iteration bootstrapping. We also assessed reclassification of cardioembolic stroke with the addition of PRS to either CRS or a modified CHA₂DS₂-VASc score alone.

RESULTS

Each AF PRS was significantly associated with AF and with cardioembolic stroke after adjustment for CRS. Addition of each AF PRS significantly improved discrimination as compared with CRS alone (P<0.01). When combined with the CRS, both PRS-CS and LDpred scores discriminated both AF and cardioembolic stroke (c-statistic 0.84 for AF; 0.74 for cardioembolic stroke) better than 3 other PRS scores (P<0.01). Using PRS-CS PRS and CRS in combination resulted in more appropriate reclassification of stroke events as compared with CRS alone (event reclassification [net reclassification indices]⁺=14% [95% CI, 10%-18%]; nonevent reclassification [net reclassification indices]^-=17% [95% CI, 15%-0.19%]) or the modified CHA₂DS₂-VASc score (net reclassification indices⁺=11% [95% CI, 7%-15%]; net reclassification indices^-=14% [95% CI, 12%-16%]) alone.

CONCLUSIONS

Addition of polygenic risk of AF to clinical risk factors modestly improves the discrimination of cardioembolic from noncardioembolic strokes, as well as reclassification of stroke subtype. Polygenic risk of AF may be a useful biomarker for identifying strokes caused by AF.

Corrigendum: How often is occult atrial fibrillation in cryptogenic stroke causal vs. incidental? A meta-analysis

[This corrects the article DOI: 10.3389/fneur.2023.1103664.].

Cross-modal autoencoder framework learns holistic representations of cardiovascular state

A fundamental challenge in diagnostics is integrating multiple modalities to develop a joint characterization of physiological state. Using the heart as a model system, we develop a cross-modal autoencoder framework for integrating distinct data modalities and constructing a holistic representation of cardiovascular state. In particular, we use our framework to construct such cross-modal representations from cardiac magnetic resonance images (MRIs), containing structural information, and electrocardiograms (ECGs), containing myoelectric information. We leverage the learned cross-modal representation to (1) improve phenotype prediction from a single, accessible phenotype such as ECGs; (2) enable imputation of hard-to-acquire cardiac MRIs from easy-to-acquire ECGs; and (3) develop a framework for performing genome-wide association studies in an unsupervised manner. Our results systematically integrate distinct diagnostic modalities into a common representation that better characterizes physiologic state.

Genetics of myocardial interstitial fibrosis in the human heart and association with disease

Myocardial interstitial fibrosis is associated with cardiovascular disease and adverse prognosis. Here, to investigate the biological pathways that underlie fibrosis in the human heart, we developed a machine learning model to measure native myocardial T1 time, a marker of myocardial fibrosis, in 41,505 UK Biobank participants who underwent cardiac magnetic resonance imaging. Greater T1 time was associated with diabetes mellitus, renal disease, aortic stenosis, cardiomyopathy, heart failure, atrial fibrillation, conduction disease and rheumatoid arthritis. Genome-wide association analysis identified 11 independent loci associated with T1 time. The identified loci implicated genes involved in glucose transport (SLC2A12), iron homeostasis (HFE, TMPRSS6), tissue repair (ADAMTSL1, VEGFC), oxidative stress (SOD2), cardiac hypertrophy (MYH7B) and calcium signaling (CAMK2D). Using a transforming growth factor β1-mediated cardiac fibroblast activation assay, we found that 9 of the 11 loci consisted of genes that exhibited temporal changes in expression or open chromatin conformation supporting their biological relevance to myofibroblast cell state acquisition. By harnessing machine learning to perform large-scale quantification of myocardial interstitial fibrosis using cardiac imaging, we validate associations between cardiac fibrosis and disease, and identify new biologically relevant pathways underlying fibrosis.

The Use of Artificial Intelligence to Predict the Development of Atrial Fibrillation

PURPOSE OF REVIEW

Atrial fibrillation (AF) is a major public health problem associated with preventable morbidity. Artificial intelligence (AI) is emerging as potential tool to prioritize individuals at increased risk for AF for preventive interventions. This review summarizes recent advances in the use of AI models to estimate AF risk.

RECENT FINDINGS

Several AI-enabled models have been recently developed which can discriminate AF risk with reasonable accuracy. AI models utilizing the electrocardiogram waveform appear to extract predictive information which is additive beyond traditional clinical risk factors. By identifying individuals at higher risk for AF, AI-based models may improve the efficiency of preventive efforts (e.g., screening, risk factor modification) intended to reduce risk of AF and associated morbidity.

Clinical and genetic associations of deep learning-derived cardiac magnetic resonance-based left ventricular mass

Left ventricular mass is a risk marker for cardiovascular events, and may indicate an underlying cardiomyopathy. Cardiac magnetic resonance is the gold-standard for left ventricular mass estimation, but is challenging to obtain at scale. Here, we use deep learning to enable genome-wide association study of cardiac magnetic resonance-derived left ventricular mass indexed to body surface area within 43,230 UK Biobank participants. We identify 12 genome-wide associations (1 known at TTN and 11 novel for left ventricular mass), implicating genes previously associated with cardiac contractility and cardiomyopathy. Cardiac magnetic resonance-derived indexed left ventricular mass is associated with incident dilated and hypertrophic cardiomyopathies, and implantable cardioverter-defibrillator implant. An indexed left ventricular mass polygenic risk score ≥90th percentile is also associated with incident implantable cardioverter-defibrillator implant in separate UK Biobank (hazard ratio 1.22, 95% CI 1.05-1.44) and Mass General Brigham (hazard ratio 1.75, 95% CI 1.12-2.74) samples. Here, we perform a genome-wide association study of cardiac magnetic resonance-derived indexed left ventricular mass to identify 11 novel variants and demonstrate that cardiac magnetic resonance-derived and genetically predicted indexed left ventricular mass are associated with incident cardiomyopathy.

How often is occult atrial fibrillation in cryptogenic stroke causal vs. incidental? A meta-analysis

Introduction

Long-term cardiac monitoring studies have unveiled low-burden, occult atrial fibrillation (AF) in some patients with otherwise cryptogenic stroke (CS), but occult AF is also found in some individuals without a stroke history and in patients with stroke of a known cause (KS). Clinical management would be aided by estimates of how often occult AF in a patient with CS is causal vs. incidental.

Methods

Through a systematic search, we identified all case–control and cohort studies applying identical long-term monitoring techniques to both patients with CS and KS. We performed a random-effects meta-analysis across these studies to determine the best estimate of the differential frequency of occult AF in CS and KS among all patients and across age subgroups. We then applied Bayes' theorem to determine the probability that occult AF is causal or incidental.

Results

The systematic search identified three case–control and cohort studies enrolling 560 patients (315 CS, 245 KS). Methods of long-term monitoring were implantable loop recorder in 31.0%, extended external monitoring in 67.9%, and both in 1.2%. Crude cumulative rates of AF detection were CS 47/315 (14.9%) vs. KS 23/246 (9.3%). In the formal meta-analysis, the summary odds ratio for occult AF in CS vs. KS in all patients was 1.80 (95% CI, 1.05–3.07), p = 0.03. With the application of Bayes' theorem, the corresponding probabilities indicated that, when present, occult AF in patients with CS is causal in 38.2% (95% CI, 0–63.6%) of patients. Analyses stratified by age suggested that detected occult AF in patients with CS was causal in 62.3% (95 CI, 0–87.1%) of patients under the age of 65 years and 28.5% (95 CI, 0–63.7%) of patients aged 65 years and older but estimates had limited precision.

Conclusion

Current evidence is preliminary, but it indicates that in cryptogenic stroke when occult AF is found, it is causal in about 38.2% of patients. These findings suggest that anticoagulation therapy may be beneficial to prevent recurrent stroke in a substantial proportion of patients with CS found to have occult AF.

Artificial Intelligence Enabled Classification of Hypertrophic Heart Diseases Using Electrocardiograms

Background

Differentiating among cardiac diseases associated with left ventricular hypertrophy (LVH) informs diagnosis and clinical care.

Objective

To evaluate if artificial intelligence-enabled analysis of the 12-lead electrocardiogram (ECG) facilitates automated detection and classification of LVH.

Methods

We used a pretrained convolutional neural network to derive numerical representations of 12-lead ECG waveforms from patients in a multi-institutional healthcare system who had cardiac diseases associated with LVH (n = 50,709), including cardiac amyloidosis (n = 304), hypertrophic cardiomyopathy (n = 1056), hypertension (n = 20,802), aortic stenosis (n = 446), and other causes (n = 4766). We then regressed LVH etiologies relative to no LVH on age, sex, and the numerical 12-lead representations using logistic regression ("LVH-Net"). To assess deep learning model performance on single-lead data analogous to mobile ECGs, we also developed 2 single-lead deep learning models by training models on lead I ("LVH-Net Lead I") or lead II ("LVH-Net Lead II") from the 12-lead ECG. We compared the performance of the LVH-Net models to alternative models fit on (1) age, sex, and standard ECG measures, and (2) clinical ECG-based rules for diagnosing LVH.

Results

The areas under the receiver operator characteristic curve of LVH-Net by specific LVH etiology were cardiac amyloidosis 0.95 [95% CI, 0.93-0.97], hypertrophic cardiomyopathy 0.92 [95% CI, 0.90-0.94], aortic stenosis LVH 0.90 [95% CI, 0.88-0.92], hypertensive LVH 0.76 [95% CI, 0.76-0.77], and other LVH 0.69 [95% CI 0.68-0.71]. The single-lead models also discriminated LVH etiologies well.

Conclusion

An artificial intelligence-enabled ECG model is favorable for detection and classification of LVH and outperforms clinical ECG-based rules.

Machine Learning to Understand Genetic and Clinical Factors Associated With the Pulse Waveform Dicrotic Notch

BACKGROUND

Absence of a dicrotic notch on finger photoplethysmography is an easily ascertainable and inexpensive trait that has been associated with age and prevalent cardiovascular disease. However, the trait exists along a continuum, and little is known about its genetic underpinnings or prognostic value for incident cardiovascular disease.

METHODS

In 169 787 participants in the UK Biobank, we identified absent dicrotic notch on photoplethysmography and created a novel continuous trait reflecting notch smoothness using machine learning. Next, we determined the heritability, genetic basis, polygenic risk, and clinical relations for the binary absent notch trait and the newly derived continuous notch smoothness trait.

RESULTS

Heritability of the continuous notch smoothness trait was 7.5%, compared with 5.6% for the binary absent notch trait. A genome-wide association study of notch smoothness identified 15 significant loci, implicating genes including NT5C2 (P=1.2×10-26), IGFBP3 (P=4.8×10-18), and PHACTR1 (P=1.4×10-13), compared with 6 loci for the binary absent notch trait. Notch smoothness stratified risk of incident myocardial infarction or coronary artery disease, stroke, heart failure, and aortic stenosis. A polygenic risk score for notch smoothness was associated with incident cardiovascular disease and all-cause death in UK Biobank participants without available photoplethysmography data.

CONCLUSIONS

We found that a machine learning derived continuous trait reflecting dicrotic notch smoothness on photoplethysmography was heritable and associated with genes involved in vascular stiffness. Greater notch smoothness was associated with greater risk of incident cardiovascular disease. Raw digital phenotyping may identify individuals at risk for disease via specific genetic pathways.

Prevalence and Incidence of Atrial Fibrillation Among Older Primary Care Patients

This cohort study is a secondary analysis of the VITAL-AF trial and assesses the prevalence and incidence of atrial fibrillation among the trial’s control participants.

Development of a Prediction Model for Ascending Aortic Diameter Among Asymptomatic Individuals

IMPORTANCE

Ascending thoracic aortic disease is an important cause of sudden death in the US, yet most aortic aneurysms are identified incidentally.

OBJECTIVE

To develop and validate a clinical score to estimate ascending aortic diameter.

DESIGN, SETTING, AND PARTICIPANTS

Using an ongoing magnetic resonance imaging substudy of the UK Biobank cohort study, which had enrolled participants from 2006 through 2010, score derivation was performed in 30 018 participants and internal validation in an additional 6681. External validation was performed in 1367 participants from the Framingham Heart Study (FHS) offspring cohort who had undergone computed tomography from 2002 through 2005, and in 50 768 individuals who had undergone transthoracic echocardiography in the Community Care Cohort Project, a retrospective hospital-based cohort of longitudinal primary care patients in the Mass General Brigham (MGB) network between 2001-2018.

EXPOSURES

Demographic and clinical variables (11 covariates that would not independently prompt thoracic imaging).

MAIN OUTCOMES AND MEASURES

Ascending aortic diameter was modeled with hierarchical group least absolute shrinkage and selection operator (LASSO) regression. Correlation between estimated and measured diameter and performance for identifying diameter 4.0 cm or greater were assessed.

RESULTS

The 30 018-participant training cohort (52% women), were a median age of 65.1 years (IQR, 58.6-70.6 years). The mean (SD) ascending aortic diameter was 3.04 (0.31) cm for women and 3.32 (0.34) cm for men. A score to estimate ascending aortic diameter explained 28.2% of the variance in aortic diameter in the UK Biobank validation cohort (95% CI, 26.4%-30.0%), 30.8% in the FHS cohort (95% CI, 26.8%-34.9%), and 32.6% in the MGB cohort (95% CI, 31.9%-33.2%). For detecting individuals with an ascending aortic diameter of 4 cm or greater, the score had an area under the receiver operator characteristic curve of 0.770 (95% CI, 0.737-0.803) in the UK Biobank, 0.813 (95% CI, 0.772-0.854) in the FHS, and 0.766 (95% CI, 0.757-0.774) in the MGB cohorts, although the model significantly overestimated or underestimated aortic diameter in external validation. Using a fixed-score threshold of 3.537, 9.7 people in UK Biobank, 1.8 in the FHS, and 4.6 in the MGB cohorts would need imaging to confirm 1 individual with an ascending aortic diameter of 4 cm or greater. The sensitivity at that threshold was 8.9% in the UK Biobank, 11.3% in the FHS, and 18.8% in the MGB cohorts, with specificities of 98.1%, 99.2%, and 96.2%, respectively.

CONCLUSIONS AND RELEVANCE

A prediction model based on common clinically available data was derived and validated to predict ascending aortic diameter. Further research is needed to optimize the prediction model and to determine whether its use is associated with improved outcomes.

One Clinician Is All You Need-Cardiac Magnetic Resonance Imaging Measurement Extraction: Deep Learning Algorithm Development

BACKGROUND

Cardiac magnetic resonance imaging (CMR) is a powerful diagnostic modality that provides detailed quantitative assessment of cardiac anatomy and function. Automated extraction of CMR measurements from clinical reports that are typically stored as unstructured text in electronic health record systems would facilitate their use in research. Existing machine learning approaches either rely on large quantities of expert annotation or require the development of engineered rules that are time-consuming and are specific to the setting in which they were developed.

OBJECTIVE

We hypothesize that the use of pretrained transformer-based language models may enable label-efficient numerical extraction from clinical text without the need for heuristics or large quantities of expert annotations. Here, we fine-tuned pretrained transformer-based language models on a small quantity of CMR annotations to extract 21 CMR measurements. We assessed the effect of clinical pretraining to reduce labeling needs and explored alternative representations of numerical inputs to improve performance.

METHODS

Our study sample comprised 99,252 patients that received longitudinal cardiology care in a multi-institutional health care system. There were 12,720 available CMR reports from 9280 patients. We adapted PRAnCER (Platform Enabling Rapid Annotation for Clinical Entity Recognition), an annotation tool for clinical text, to collect annotations from a study clinician on 370 reports. We experimented with 5 different representations of numerical quantities and several model weight initializations. We evaluated extraction performance using macroaveraged F1-scores across the measurements of interest. We applied the best-performing model to extract measurements from the remaining CMR reports in the study sample and evaluated established associations between selected extracted measures with clinical outcomes to demonstrate validity.

RESULTS

All combinations of weight initializations and numerical representations obtained excellent performance on the gold-standard test set, suggesting that transformer models fine-tuned on a small set of annotations can effectively extract numerical quantities. Our results further indicate that custom numerical representations did not appear to have a significant impact on extraction performance. The best-performing model achieved a macroaveraged F1-score of 0.957 across the evaluated CMR measurements (range 0.92 for the lowest-performing measure of left atrial anterior-posterior dimension to 1.0 for the highest-performing measures of left ventricular end systolic volume index and left ventricular end systolic diameter). Application of the best-performing model to the study cohort yielded 136,407 measurements from all available reports in the study sample. We observed expected associations between extracted left ventricular mass index, left ventricular ejection fraction, and right ventricular ejection fraction with clinical outcomes like atrial fibrillation, heart failure, and mortality.

CONCLUSIONS

This study demonstrated that a domain-agnostic pretrained transformer model is able to effectively extract quantitative clinical measurements from diagnostic reports with a relatively small number of gold-standard annotations. The proposed workflow may serve as a roadmap for other quantitative entity extraction.

Wearable accelerometer-derived physical activity and incident disease

Physical activity is regarded as favorable to health but effects across the spectrum of human disease are poorly quantified. In contrast to self-reported measures, wearable accelerometers can provide more precise and reproducible activity quantification. Using wrist-worn accelerometry data from the UK Biobank prospective cohort study, we test associations between moderate-to-vigorous physical activity (MVPA) – both total MVPA minutes and whether MVPA is above a guideline-based threshold of ≥150 min/week—and incidence of 697 diseases using Cox proportional hazards models adjusted for age, sex, body mass index, smoking, Townsend Deprivation Index, educational attainment, diet quality, alcohol use, blood pressure, anti-hypertensive use. We correct for multiplicity at a false discovery rate of 1%. We perform analogous testing using self-reported MVPA. Among 96,244 adults wearing accelerometers for one week (age 62 ± 8 years), MVPA is associated with 373 (54%) tested diseases over a median 6.3 years of follow-up. Greater MVPA is overwhelmingly associated with lower disease risk (98% of associations) with hazard ratios (HRs) ranging 0.70–0.98 per 150 min increase in weekly MVPA, and associations spanning all 16 disease categories tested. Overall, associations with lower disease risk are enriched for cardiac (16%), digestive (14%), endocrine/metabolic (10%), and respiratory conditions (8%) (chi-square p < 0.01). Similar patterns are observed using the guideline-based threshold of ≥150 MVPA min/week. Some of the strongest associations with guideline-adherent activity include lower risks of incident heart failure (HR 0.65, 95% CI 0.55–0.77), type 2 diabetes (HR 0.64, 95% CI 0.58–0.71), cholelithiasis (HR 0.61, 95% CI 0.54–0.70), and chronic bronchitis (HR 0.42, 95% CI 0.33–0.54). When assessed within 456,374 individuals providing self-reported MVPA, effect sizes for guideline-adherent activity are substantially smaller (e.g., heart failure HR 0.84, 95% CI 0.80–0.88). Greater wearable device-based physical activity is robustly associated with lower disease incidence. Future studies are warranted to identify potential mechanisms linking physical activity and disease, and assess whether optimization of measured activity can reduce disease risk.

Endophenotype effect sizes support variant pathogenicity in monogenic disease susceptibility genes

Accurate and efficient classification of variant pathogenicity is critical for research and clinical care. Using data from three large studies, we demonstrate that population-based associations between rare variants and quantitative endophenotypes for three monogenic diseases (low-density-lipoprotein cholesterol for familial hypercholesterolemia, electrocardiographic QTc interval for long QT syndrome, and glycosylated hemoglobin for maturity-onset diabetes of the young) provide evidence for variant pathogenicity. Effect sizes are associated with pathogenic ClinVar assertions (P < 0.001 for each trait) and discriminate pathogenic from non-pathogenic variants (area under the curve 0.82-0.84 across endophenotypes). An effect size threshold of ≥ 0.5 times the endophenotype standard deviation nominates up to 35% of rare variants of uncertain significance or not in ClinVar in disease susceptibility genes with pathogenic potential. We propose that variant associations with quantitative endophenotypes for monogenic diseases can provide evidence supporting pathogenicity.

Cost-effectiveness of Screening for Atrial Fibrillation Using Wearable Devices

Question

Is population-based atrial fibrillation (AF) screening using wearable devices cost-effective?

Findings

In this economic evaluation of 30 million simulated individuals with an age, sex, and comorbidity profile matching the US population aged 65 years or older, AF screening using wearable devices was cost-effective, with the overall preferred strategy identified as wearable photoplethysmography, followed conditionally by wearable electrocardiography with patch monitor confirmation (incremental cost-effectiveness ratio, $57 894 per quality-adjusted life-year). The cost-effectiveness of screening was consistent across multiple scenarios, including strata of sex, screening at earlier ages, and with variation in the association of anticoagulation with risk of stroke associated with screening-detected AF.

Meaning

This study suggests that contemporary AF screening using wearable devices may be cost-effective.

Importance

Undiagnosed atrial fibrillation (AF) is an important cause of stroke. Screening for AF using wrist-worn wearable devices may prevent strokes, but their cost-effectiveness is unknown.

Objective

To evaluate the cost-effectiveness of contemporary AF screening strategies, particularly wrist-worn wearable devices.

Design, Setting, and Participants

This economic evaluation used a microsimulation decision-analytic model and was conducted from September 8, 2020, to May 23, 2022, comprising 30 million simulated individuals with an age, sex, and comorbidity profile matching the US population aged 65 years or older.

Interventions

Eight AF screening strategies, with 6 using wrist-worn wearable devices (watch or band photoplethysmography, with or without watch or band electrocardiography) and 2 using traditional modalities (ie, pulse palpation and 12-lead electrocardiogram) vs no screening.

Main Outcomes and Measures

The primary outcome was the incremental cost-effectiveness ratio, defined as US dollars per quality-adjusted life-year (QALY). Secondary measures included rates of stroke and major bleeding.

Results

In the base case analysis of this model, the mean (SD) age was 72.5 (7.5) years, and 50% of the individuals were women. All 6 screening strategies using wrist-worn wearable devices were estimated to be more effective than no screening (range of QALYs gained vs no screening, 226-957 per 100 000 individuals) and were associated with greater relative benefit than screening using traditional modalities (range of QALYs gained vs no screening, −116 to 93 per 100 000 individuals). Compared with no screening, screening using wrist-worn wearable devices was associated with a reduction in stroke incidence by 20 to 23 per 100 000 person-years but an increase in major bleeding by 20 to 44 per 100 000 person-years. The overall preferred strategy was wearable photoplethysmography, followed conditionally by wearable electrocardiography with patch monitor confirmation, which had an incremental cost-effectiveness ratio of $57 894 per QALY, meeting the acceptability threshold of $100 000 per QALY. The cost-effectiveness of screening was consistent across multiple scenarios, including strata of sex, screening at earlier ages (eg, ≥50 years), and with variation in the association of anticoagulation with risk of stroke in the setting of screening-detected AF.

Conclusions and Relevance

This economic evaluation of AF screening using a microsimulation decision-analytic model suggests that screening using wearable devices is cost-effective compared with either no screening or AF screening using traditional methods.

Natural Language Processing to Improve Prediction of Incident Atrial Fibrillation Using Electronic Health Records

Background Models predicting atrial fibrillation (AF) risk, such as Cohorts for Heart and Aging Research in Genomic Epidemiology AF (CHARGE-AF), have not performed as well in electronic health records. Natural language processing (NLP) may improve models by using narrative electronic health record text. Methods and Results From a primary care network, we included patients aged ≥65 years with visits between 2003 and 2013 in development (n=32 960) and internal validation cohorts (n=13 992). An external validation cohort from a separate network from 2015 to 2020 included 39 051 patients. Model features were defined using electronic health record codified data and narrative data with NLP. We developed 2 models to predict 5-year AF incidence using (1) codified+NLP data and (2) codified data only and evaluated model performance. The analysis included 2839 incident AF cases in the development cohort and 1057 and 2226 cases in internal and external validation cohorts, respectively. The C-statistic was greater (P<0.001) in codified+NLP model (0.744 [95% CI, 0.735-0.753]) compared with codified-only (0.730 [95% CI, 0.720-0.739]) in the development cohort. In internal validation, the C-statistic of codified+NLP was modestly higher (0.735 [95% CI, 0.720-0.749]) compared with codified-only (0.729 [95% CI, 0.715-0.744]; P=0.06) and CHARGE-AF (0.717 [95% CI, 0.703-0.731]; P=0.002). Codified+NLP and codified-only were well calibrated, whereas CHARGE-AF underestimated AF risk. In external validation, the C-statistic of codified+NLP (0.750 [95% CI, 0.740-0.760]) remained higher (P<0.001) than codified-only (0.738 [95% CI, 0.727-0.748]) and CHARGE-AF (0.735 [95% CI, 0.725-0.746]). Conclusions Estimation of 5-year risk of AF can be modestly improved using NLP to incorporate narrative electronic health record data.

Deep learning on resting electrocardiogram to identify impaired heart rate recovery

Background and Objective

Postexercise heart rate recovery (HRR) is an important indicator of cardiac autonomic function and abnormal HRR is associated with adverse outcomes. We hypothesized that deep learning on resting electrocardiogram (ECG) tracings may identify individuals with impaired HRR.

Methods

We trained a deep learning model (convolutional neural network) to infer HRR based on resting ECG waveforms (HRR_pred) among UK Biobank participants who had undergone exercise testing. We examined the association of HRR_pred with incident cardiovascular disease using Cox models, and investigated the genetic architecture of HRR_pred in genome-wide association analysis.

Results

Among 56,793 individuals (mean age 57 years, 51% women), the HRR_pred model was moderately correlated with actual HRR (r = 0.48, 95% confidence interval [CI] 0.47-0.48). Over a median follow-up of 10 years, we observed 2060 incident diabetes mellitus (DM) events, 862 heart failure events, and 2065 deaths. Higher HRR_pred was associated with lower risk of DM (hazard ratio [HR] 0.79 per 1 standard deviation change, 95% CI 0.76-0.83), heart failure (HR 0.89, 95% CI 0.83-0.95), and death (HR 0.83, 95% CI 0.79-0.86). After accounting for resting heart rate, the association of HRR_pred with incident DM and all-cause mortality were similar. Genetic determinants of HRR_pred included known heart rate, cardiac conduction system, cardiomyopathy, and metabolic trait loci.

Conclusion

Deep learning-derived estimates of HRR using resting ECG independently associated with future clinical outcomes, including new-onset DM and all-cause mortality. Inferring postexercise heart rate response from a resting ECG may have potential clinical implications and impact on preventive strategies warrants future study.

Prediction performance and fairness heterogeneity in cardiovascular risk models

Prediction models are commonly used to estimate risk for cardiovascular diseases, to inform diagnosis and management. However, performance may vary substantially across relevant subgroups of the population. Here we investigated heterogeneity of accuracy and fairness metrics across a variety of subgroups for risk prediction of two common diseases: atrial fibrillation (AF) and atherosclerotic cardiovascular disease (ASCVD). We calculated the Cohorts for Heart and Aging in Genomic Epidemiology Atrial Fibrillation (CHARGE-AF) score for AF and the Pooled Cohort Equations (PCE) score for ASCVD in three large datasets: Explorys Life Sciences Dataset (Explorys, n = 21,809,334), Mass General Brigham (MGB, n = 520,868), and the UK Biobank (UKBB, n = 502,521). Our results demonstrate important performance heterogeneity across subpopulations defined by age, sex, and presence of preexisting disease, with fairly consistent patterns across both scores. For example, using CHARGE-AF, discrimination declined with increasing age, with a concordance index of 0.72 [95% CI 0.72-0.73] for the youngest (45-54 years) subgroup to 0.57 [0.56-0.58] for the oldest (85-90 years) subgroup in Explorys. Even though sex is not included in CHARGE-AF, the statistical parity difference (i.e., likelihood of being classified as high risk) was considerable between males and females within the 65-74 years subgroup with a value of - 0.33 [95% CI - 0.33 to - 0.33]. We also observed weak discrimination (i.e., < 0.7) and suboptimal calibration (i.e., calibration slope outside of 0.7-1.3) in large subsets of the population; for example, all individuals aged 75 years or older in Explorys (17.4%). Our findings highlight the need to characterize and quantify the behavior of clinical risk models within specific subpopulations so they can be used appropriately to facilitate more accurate, consistent, and equitable assessment of disease risk.

LMNA Variants and Risk of Adult-Onset Cardiac Disease

BACKGROUND

Genetic variants in LMNA may cause cardiac disease, but population-level contributions of variants to cardiac disease burden are not well-characterized.

OBJECTIVES

We sought to determine the frequency and contribution of rare LMNA variants to cardiomyopathy and arrhythmia risk among ambulatory adults.

METHODS

We included 185,990 UK Biobank participants with whole-exome sequencing. We annotated rare loss-of-function and missense LMNA variants for functional effect using 30 in silico prediction tools. We assigned a predicted functional effect weight to each variant and calculated a score for each carrier. We tested associations between the LMNA score and arrhythmia (atrial fibrillation, bradyarrhythmia, ventricular arrhythmia) or cardiomyopathy outcomes (dilated cardiomyopathy and heart failure). We also examined associations for variants located upstream vs downstream of the nuclear localization signal.

RESULTS

Overall, 1,167 (0.63%) participants carried an LMNA variant and 15,079 (8.11%) had an arrhythmia or cardiomyopathy event during a median follow-up of 10.9 years. The LMNA score was associated with arrhythmia or cardiomyopathy (OR: 2.21; P < 0.001) and the association was more significant when restricted to variants upstream of the nuclear localization signal (OR: 5.05; P < 0.001). The incidence rate of arrhythmia or cardiomyopathy was 8.43 per 1,000 person-years (95% CI: 6.73-10.12 per 1,000 person-years) among LMNA variant carriers and 6.38 per 1,000 person-years (95% CI: 6.27-6.50 per 1,000 person-years) among noncarriers. Only 3 (1.2%) of the variants were reported as pathogenic in ClinVar.

CONCLUSIONS

Middle-aged adult carriers of rare missense or loss-of-function LMNA variants are at increased risk for arrhythmia and cardiomyopathy.

Genetic analysis of right heart structure and function in 40,000 people

Congenital heart diseases often involve maldevelopment of the evolutionarily recent right heart chamber. To gain insight into right heart structure and function, we fine-tuned deep learning models to recognize the right atrium, right ventricle and pulmonary artery, measuring right heart structures in 40,000 individuals from the UK Biobank with magnetic resonance imaging. Genome-wide association studies identified 130 distinct loci associated with at least one right heart measurement, of which 72 were not associated with left heart structures. Loci were found near genes previously linked with congenital heart disease, including NKX2-5, TBX5/TBX3, WNT9B and GATA4. A genome-wide polygenic predictor of right ventricular ejection fraction was associated with incident dilated cardiomyopathy (hazard ratio, 1.33 per standard deviation; P = 7.1 × 10-13) and remained significant after accounting for a left ventricular polygenic score. Harnessing deep learning to perform large-scale cardiac phenotyping, our results yield insights into the genetic determinants of right heart structure and function.