Deep Learning of Electrocardiograms in Sinus Rhythm From US Veterans to Predict Atrial Fibrillation.

Importance: Early detection of atrial fibrillation (AF) may help prevent adverse cardiovascular events such as stroke. Deep learning applied to electrocardiograms (ECGs) has been successfully used for early identification of several cardiovascular diseases. Objective: To determine whether deep learning models applied to outpatient ECGs in sinus rhythm can predict AF in a large and diverse patient population. Design, Setting, and Participants: This prognostic study was performed on ECGs acquired from January 1, 1987, to December 31, 2022, at 6 US Veterans Affairs (VA) hospital networks and 1 large non-VA academic medical center. Participants included all outpatients with 12-lead ECGs in sinus rhythm. Main Outcomes and Measures: A convolutional neural network using 12-lead ECGs from 2 US VA hospital networks was trained to predict the presence of AF within 31 days of sinus rhythm ECGs. The model was tested on ECGs held out from training at the 2 VA networks as well as 4 additional VA networks and 1 large non-VA academic medical center. Results: A total of 907 858 ECGs from patients across 6 VA sites were included in the analysis. These patients had a mean (SD) age of 62.4 (13.5) years, 6.4% were female, and 93.6% were male, with a mean (SD) CHA2DS2-VASc (congestive heart failure, hypertension, age, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age, sex category) score of 1.9 (1.6). A total of 0.2% were American Indian or Alaska Native, 2.7% were Asian, 10.7% were Black, 4.6% were Latinx, 0.7% were Native Hawaiian or Other Pacific Islander, 62.4% were White, 0.4% were of other race or ethnicity (which is not broken down into subcategories in the VA data set), and 18.4% were of unknown race or ethnicity. At the non-VA academic medical center (72 483 ECGs), the mean (SD) age was 59.5 (15.4) years and 52.5% were female, with a mean (SD) CHA2DS2-VASc score of 1.6 (1.4). A total of 0.1% were American Indian or Alaska Native, 7.9% were Asian, 9.4% were Black, 2.9% were Latinx, 0.03% were Native Hawaiian or Other Pacific Islander, 74.8% were White, 0.1% were of other race or ethnicity, and 4.7% were of unknown race or ethnicity. A deep learning model predicted the presence of AF within 31 days of a sinus rhythm ECG on held-out test ECGs at VA sites with an area under the receiver operating characteristic curve (AUROC) of 0.86 (95% CI, 0.85-0.86), accuracy of 0.78 (95% CI, 0.77-0.78), and F1 score of 0.30 (95% CI, 0.30-0.31). At the non-VA site, AUROC was 0.93 (95% CI, 0.93-0.94); accuracy, 0.87 (95% CI, 0.86-0.88); and F1 score, 0.46 (95% CI, 0.44-0.48). The model was well calibrated, with a Brier score of 0.02 across all sites. Among individuals deemed high risk by deep learning, the number needed to screen to detect a positive case of AF was 2.47 individuals for a testing sensitivity of 25% and 11.48 for 75%. Model performance was similar in patients who were Black, female, or younger than 65 years or who had CHA2DS2-VASc scores of 2 or greater. Conclusions and Relevance: Deep learning of outpatient sinus rhythm ECGs predicted AF within 31 days in populations with diverse demographics and comorbidities. Similar models could be used in future AF screening efforts to reduce adverse complications associated with this disease.

Ventricular arrhythmia detection for contemporary Biotronik and Abbott implantable cardioverter defibrillators with markedly prolonged detection in Biotronik devices.

BACKGROUND: Implantable cardioverter defibrillators (ICDs) are typically programed with both ventricular tachycardia (VT) and ventricular fibrillation (VF) treatment zones. Biotronik and Abbott ICDs do not increment the VT counter when the tachycardia accelerates to the VF zone, which could result in a prolonged delay in tachycardia detection. METHODS: Patients with Biotronik and Abbott ICDs receiving care at Veterans Affairs facilities in Northern California were identified. Patient information and device tracings for patients with any ICD therapies were examined to assess for possible delayed tachycardia detection. RESULTS: Among 52 patients with Biotronik ICDs, 8 (15%) experienced appropriate ICD therapy over a median follow-up of 29 months. Among 68 patients with Abbott ICDs, 26 (38%) experienced appropriate ICD therapy over a median follow-up of 83 months. Three of the patients with Biotronik ICDs who received appropriate therapy experienced a delay in VT/VF detection due to the tachycardia rate oscillating between the VT and VF treatment zones (longest 31.2 s on detection), compared with four of the patients with Abbott ICDs (longest 4.1 s on the detection and 8 s on redetect). One of the patients with a Biotronik ICD experienced recurrent syncope associated with delayed detection and another died on the day of delayed detection. One of the patients with an Abbott ICD experienced syncope. CONCLUSIONS: Because contemporary Biotronik and Abbott ICDs freeze the VT counters when tachycardia is in the VF zone, ICD therapies can be markedly delayed when the tachycardia oscillates between the VT and VF zone.

High-Throughput Assessment of Real-World Medication Effects on QT Interval Prolongation: Observational Study.

BACKGROUND: Drug-induced prolongation of the corrected QT interval (QTc) increases the risk for Torsades de Pointes (TdP) and sudden cardiac death. Medication effects on the QTc have been studied in controlled settings but may not be well evaluated in real-world settings where medication effects may be modulated by patient demographics and comorbidities as well as the usage of other concomitant medications. OBJECTIVE: We demonstrate a new, high-throughput method leveraging electronic health records (EHRs) and the Surescripts pharmacy database to monitor real-world QTc-prolonging medication and potential interacting effects from demographics and comorbidities. METHODS: We included all outpatient electrocardiograms (ECGs) from September 2008 to December 2019 at a large academic medical system, which were in sinus rhythm with a heart rate of 40-100 beats per minute, QRS duration of <120 milliseconds, and QTc of 300-700 milliseconds, determined using the Bazett formula. We used prescription information from the Surescripts pharmacy database and EHR medication lists to classify whether a patient was on a medication during an ECG. Negative control ECGs were obtained from patients not currently on the medication but who had been or would be on that medication within 1 year. We calculated the difference in mean QTc between ECGs of patients who are on and those who are off a medication and made comparisons to known medication TdP risks per the CredibleMeds.org database. Using linear regression analysis, we studied the interaction of patient-level demographics or comorbidities on medication-related QTc prolongation. RESULTS: We analyzed the effects of 272 medications on 310,335 ECGs from 159,397 individuals. Medications associated with the greatest QTc prolongation were dofetilide (mean QTc difference 21.52, 95% CI 10.58-32.70 milliseconds), mexiletine (mean QTc difference 18.56, 95% CI 7.70-29.27 milliseconds), amiodarone (mean QTc difference 14.96, 95% CI 13.52-16.33 milliseconds), rifaximin (mean QTc difference 14.50, 95% CI 12.12-17.13 milliseconds), and sotalol (mean QTc difference 10.73, 95% CI 7.09-14.37 milliseconds). Several top QT prolonging medications such as rifaximin, lactulose, cinacalcet, and lenalidomide were not previously known but have plausible mechanistic explanations. Significant interactions were observed between demographics or comorbidities and QTc prolongation with many medications, such as coronary disease and amiodarone. CONCLUSIONS: We demonstrate a new, high-throughput technique for monitoring real-world effects of QTc-prolonging medications from readily accessible clinical data. Using this approach, we confirmed known medications for QTc prolongation and identified potential new associations and demographic or comorbidity interactions that could supplement findings in curated databases. Our single-center results would benefit from additional verification in future multisite studies that incorporate larger numbers of patients and ECGs along with more precise medication adherence and comorbidity data.

Dysregulated carbohydrate and lipid metabolism and risk of atrial fibrillation in advanced old age.

OBJECTIVE: Obesity and dysmetabolism are major risk factors for atrial fibrillation (AF). Fasting and postload levels of glucose and non-esterified fatty acids (NEFAs) reflect different facets of metabolic regulation. We sought to study their respective contributions to AF risk concurrently. METHODS: We assessed levels of fasting and postload glucose and NEFA in the Cardiovascular Health Study to identify associations with AF incidence and, secondarily, with ECG parameters of AF risk available at baseline. Linear and Cox regressions were performed. RESULTS: The study included 1876 participants (age 77.7±4.4). During the median follow-up of 11.4 years, 717 cases of incident AF occurred. After adjustment for potential confounders, postload glucose showed an association with incident AF (HR per SD increment of postload glucose=1.11, 95% CI 1.02 to 1.21, p=0.017). Both glucose measures, but not NEFA, were positively associated with higher P wave terminal force in V1 (PTFV1); the association remained significant only for postload glucose when the two measures were entered together (ß per SD increment=138 µV·ms, 95% CI 15 to 260, p=0.028). Exploratory analyses showed significant interaction by sex for fasting NEFA (pinteraction=0.044) and postload glucose (pinteraction=0.015) relative to AF, with relationships stronger in women. For postload glucose, the association with incident AF was observed among women but not among men. CONCLUSIONS: Among older adults, postload glucose was positively associated with incident AF, with consistent findings for PTFV1. In exploratory analyses, the relationship with AF appeared specific to women. These findings require further study but suggest that interventions to address postprandial dysglycaemia late in life might reduce AF.

Fasting and Post-Load Glucose and Non-Esterified Fatty Acids and Risk of Heart Failure and Its Subtypes in Older Adults.

BACKGROUND: Glucose and non-esterified fatty acids (NEFA) are myocardial fuels whose fasting and post-prandial levels are under different homeostatic regulation. The relationships of fasting and post-load glucose and NEFA with incident heart failure (HF) remain incompletely defined. METHODS: Serum glucose and NEFA were measured during fasting and 2 hours post-oral glucose tolerance test, performed in Cardiovascular Health Study participants not receiving hypoglycemic medication. Participants with prevalent HF or lacking relevant data were excluded. Outcomes were incident HF (primary), and HF with preserved (HFpEF) and reduced (HFrEF) ejection fraction (secondary). RESULTS: Among 2 238 participants (age 78 ± 4) with a median follow-up of 9.9 years, there were 737 HF events. After adjustment for demographic and lifestyle factors, both fasting (hazard ratio [HR] = 1.11 per SD [95% confidence interval {CI} = 1.01-1.23], p = .040) and post-load (HR = 1.14 per SD [1.05-1.24], p = 0.002) glucose were significantly associated with incident HF. No association was seen for fasting or post-load NEFA. Upon mutual adjustment, only post-load glucose (HR = 1.11 [1.003-1.22], p = .044), but not fasting glucose (HR = 1.06 [0.94-1.20], p = .340), remained associated with HF. Further adjustment for cardiovascular disease and other risk factors in the causal pathway did not affect the association for post-load glucose, but eliminated that for fasting glucose. Associations for fasting and post-load glucose appeared stronger with higher adiposity and were observed specifically for HFrEF but not HFpEF. CONCLUSIONS: Fasting and post-load glucose, but not NEFA, were associated with incident HF. The association was especially robust for post-load glucose, suggesting that pathways involved in post-prandial dysglycemia could offer new targets for HF prevention late in life.

Effects of Supervised Exercise Training on Atrial Fibrillation: A META-ANALYSIS OF RANDOMIZED CONTROLLED TRIALS.

PURPOSE: Atrial fibrillation (AF) is associated with the comorbidities of a sedentary lifestyle. Endurance athletes also show an increased incidence of AF. The role of exercise in the treatment of AF is unknown so this study aimed to examine the effects of supervised exercise on AF. METHODS: A meta-analysis of randomized controlled trials (RCTs) examining supervised exercise training in participants with AF was performed. The primary outcome was AF recurrence and burden. Secondary outcomes included AF symptoms, quality of life, and cardiorespiratory fitness (CRF). RESULTS: Thirteen RCTs, involving 1155 participants, were included. Paroxysmal AF was present in 34% and persistent AF in 64%. The types of exercise were diverse and included cardiac rehabilitation (64%), aerobic training (7%), Qi Gong (4%), interval training (11%), and yoga (15%). Exercise training reduced AF recurrence (relative risk = 0.77: 95% CI, 0.60-0.99), improved quality of life in 5 of the 10 components of the Short Form 36 survey, and improved CRF (standardized mean difference [SMD] = 0.56: 95% CI, 0.27-0.85). The AF burden was reduced only in studies that included continuous ambulatory monitoring (SMD =-0.49: 95% CI, -0.96 to -0.01) but not when all studies were included (SMD =-0.12: 95% CI, -0.61 to 0.38). There was no difference in adverse events between exercise and control. CONCLUSIONS: Supervised exercise training is safe, reduces AF recurrence, and improves quality of life and CRF in participants with AF. Further large RCTs with ambulatory monitoring and robust exercise regimens are needed to assess the effects of exercise training on AF burden and AF symptoms.

Electrical abnormalities with St. Jude/Abbott pacing leads: A systematic review and meta-analysis.

BACKGROUND: Although there is a paucity of contemporary data on pacemaker lead survival rates, small studies suggest that some leads may have higher malfunction rates than do others. OBJECTIVE: The purpose of this study was to determine the malfunction rates of current pacemaker leads. METHODS: A meta-analysis including studies that examined the non-implant-related lead malfunction rates of current commercially available active fixation pacemaker leads was performed. An electronic search of MEDLINE/PubMed, Scopus, and Embase was performed. DerSimonian and Laird random effects models were used. RESULTS: Eight studies with a total of 14,579 leads were included. Abbott accounted for 10,838 (74%), Medtronic 2510 (17%), Boston Scientific 849 (6%), and MicroPort 382 (3%) leads. The weighted mean follow-up period was 3.6 years. Lead abnormalities occurred in 5.0% of all leads, 6.1% of Abbott leads, 1.1% of Medtronic, 1.4% of Boston Scientific, and 5.5% of MicroPort. The most common lead abnormality was lead noise with normal impedance. Abbott leads were associated with an increased risk of abnormalities (relative risk [RR] 7.81; 95% confidence interval [CI] 3.21-19.04), reprogramming (RR 7.95; 95% CI 3.55-17.82), and lead revision or extraction (RR 8.91; 95% CI 3.36-23.60). Abbott leads connected to an Abbott generator had the highest abnormality rate (8.0%) followed by Abbott leads connected to a non-Abbott generator (4.7%) and non-Abbott leads connected to an Abbott generator (0.4%). CONCLUSIONS: Abbott leads are associated with an increased risk of abnormalities compared with leads of other manufacturers, primarily manifesting as lead noise with normal impedance, and are associated with an increased risk of lead reprogramming and lead revision or extraction.

Increased Rate of Pump Thrombosis and Cardioembolic Events Following Ventricular Tachycardia Ablation in Patients Supported With Left Ventricular Assist Devices.

Ventricular arrhythmias are common following left ventricular assist device implantation (LVAD), and the effects of ventricular tachycardia (VT) ablation on thrombosis and embolic events are unknown. We aimed to assess LVAD thrombosis, stroke, and embolic event rates after VT ablation. Left ventricular assist device implantation patients from two academic centers who underwent endocardial VT ablation between 2009 and 2016 were compared to a control group with VT who were not ablated and followed for one year. The primary composite outcome was confirmed or suspected LVAD thrombosis, stroke, or other embolic event. Survival analysis was conducted with Kaplan-Meier curves, log-rank tests, and Cox regression. Forty-three LVAD patients underwent VT ablation, and 73 LVAD patients had VT but were not ablated. Patients who were ablated were more likely have VT prior to LVAD (p = 0.04), monomorphic VT (p < 0.01), and to be on antiarrhythmics (p < 0.01). Fifty-eight percent of the patients in the ablation group experienced the primary composite outcome (11% had confirmed device thrombosis [DT], 41% suspected DT, 39% had a stroke or embolic event) compared to 30% in the control group (12% with confirmed DT, 11% with suspected DT, 14% with stroke or embolic event) (p = 0.002). In multivariable regression, ablation was an independent predictor of the primary composite outcome (hazard ratios, 2.24; 95% confidence interval, 1.09-4.61; p = 0.03). Patients with LVADs referred for endocardial VT ablation had elevated rates of DT and embolic events.

Electrophysiologic approach to diagnosis and ablation of patients with permanent junctional reciprocating tachycardia associated with complex anatomy and/or physiology.

INTRODUCTION: Permanent junctional reciprocating tachycardia (PJRT) is a rare supraventricular tachycardia (SVT), typically involving a single decremental posteroseptal accessory pathway (AP). METHODS: Four patients with long RP SVT underwent electrophysiology (EP) study and ablation. The cases were reviewed. RESULTS: Case 1 recurred despite 3 prior ablations at the site of earliest retrograde atrial activation during orthodromic reciprocating tachycardia (ORT). Mapping during a repeat EP study demonstrated a prepotential in the coronary sinus (CS). Ablation over the earliest atrial activation in the CS resulted in dissociation of the potential from the atrium during sinus rhythm. The potential was traced back to the CS os and ablated. Case 2 underwent successful ablation at 6 o'clock on the mitral annulus (MA). ORT recurred and successful ablation was performed at 1 o'clock on the MA. Case 3 had tachycardia with variation in both V-A and A-H intervals which precluded the use of usual maneuvers so we used simultaneous atrial and ventricular pacing and introduced a premature atrial contraction with a closely coupled premature ventricular contraction. Case 4 had had two prior atrial fibrillation ablations with continued SVT over a decremental atrioventricular bypass tract that was successfully ablated at 5 o'clock on the tricuspid annulus. A second SVT consistent with a concealed nodoventricular pathway was successfully ablated at the right inferior extension of the AV nodal slow pathway. CONCLUSION: We describe challenging cases of PJRT by virtue of complex anatomy, diagnostic features, and multiple arrhythmia mechanisms.

Posterior Wall Isolation for Atrial Fibrillation: Effects on Echocardiographic Parameters of cardiac function.

BACKGROUND: Posterior wall isolation (PWI) in addition to circumferential pulmonary vein isolation (CPVA) requires more ablation of left atrial tissue. We compared the effect of PWI versus non-PWI methods (CPVA + linear lesions) on echocardiographic parameters of left atrial and left ventricular function. METHODS: We selected patients who had pre and post ablation echocardiogram at our institution. Parameters assessed were: Left ventricular outflow tract velocity time integral (VTI), left ventricular ejection fraction (LVEF), atrial Doppler velocity across mitral valve (A), E/E', and deceleration time. RESULTS: Of the 72 patients studied, 32.5% had PWA in addition to CPVA. The mean duration between echocardiograms was 650 + 542 days. PWA group had an average postoperative VTI 0.21 + 0.05 vs 0.21 + 0.05 in the non-PWA (p=0.61) group. Average improvement compared to pre ablation parameters: VTI was 0.03 + 0.06 vs 0.008 + 0.05 (p=0.17), postoperative A was 0.49 + 0.19 vs 0.57 + 0.19 (p=0.16), postoperative LVEF was 57.5 + 9.9% vs 57.8 + 10.8 % (p=0.89), with average change in LVEF 1.5 ± 7.8 vs 0.86 ± 9.7 (p=0.78) in PWA and non-PWA groups respectively. There was no significant difference in change in deceleration time or E/E' when comparing the two groups. CONCLUSIONS: PWA did not adversely affect echocardiographic parameters of left atrial function or left ventricular systolic or diastolic function when compared to other types of ablation.

Patient Perception of the Remote Versus Clinic Visits for Interrogation of Implantable Cardioverter Defibrillators.

Implantable Cardioverter Defibrillators (ICDs) are used in the management of sudden cardiac arrest. Compared with clinic visits, remote interrogation of these devices has shown clinical benefit and lower cost. We hypothesize that demographic and socioeconomic factors influence patient satisfaction with remote monitoring and therefore the choice of a pathway for follow-up. Questionnaires were mailed to 85 patients (mean age 63 ± 13.5 years, 73% male), with ICDs implanted for primary prevention of sudden cardiac arrest. Information regarding education, social support, employment, and income was collected. To compare clinic and remote monitoring, patients were given questionnaires to assess which parameters they consider important: convenience, accuracy, human contact, scheduling, and cost. Of the 34 responders, patients rated clinic visit to be as accurate with better opportunity to ask questions and better human contact, but there was no difference in perception of convenience, scheduling, or cost between the 2 groups. Significant number of patients dropped from the labor market after ICD implantation; however labor status, education, or income did not influence the preference of clinic appointment. Survey respondents preferred clinic to remote interrogation because they believe clinic appointments allow better interaction. Educating patients about the benefits of remote interrogation and improved communication will enhance utilization of this sophisticated technology for superior patient care.

Atrial arrhythmias in patients with arrhythmogenic right ventricular cardiomyopathy: Prevalence, echocardiographic predictors, and treatment.

INTRODUCTION: The clinical role of atrial arrhythmias (AA) in arrhythmogenic right ventricular cardiomyopathy (ARVC) and the echocardiographic variables that predict them are not well defined. We describe the prevalence, types, echocardiographic predictors, and management of AA in patients with ARVC. METHODS: We retrospectively evaluated medical records of 117 patients with definite ARVC (2010 Task Force Criteria) from two tertiary care centers. We identified those patients with sustained AA (>30 seconds), including atrial fibrillation (AF), atrial flutter (AFL), and atrial tachycardia (AT). We collected demographic, genetic, and clinical data. The median follow-up was 3.4 years (interquartile range = 2.0-5.7). RESULTS: Total 26 patients (22%) had one or more types of AA: AF (n = 19), AFL (n = 9), and AT (n = 8). We performed genetic testing on 84 patients with ARVC (71.8%). Two patients with AA (8%) had peripheral emboli, and one patient (4%) suffered inappropriate implantable cardioverter-defibrillator shock. We performed catheter ablation of AA in eight patients (31%), with no procedural complications. Right atrial area and left atrial volume index were independently associated with increased odds of AA; odds ratio (OR), 1.1 (95% confidence interval [CI]:1.02-1.16) (P = .01) and OR, 1.1 (95% CI:1.03-1.15) (P = .003), respectively. An increase in tricuspid annular plane peak systolic excursion was independently associated with reduced odds; OR, 0.3 (95% CI: 0.1-0.94) (P = .003). CONCLUSIONS: Atrial arrhythmias (AA) are common in patients with arrhythmogenic right ventricular cardiomyopathy (ARVC). Inappropriate shocks and systemic emboli may be associated with AA. Atrial size and right ventricular dysfunction may help identify patients with ARVC at increased odds of AA.