Artificial intelligence predicts clinically relevant atrial high-rate episodes in patients with cardiac implantable electronic devices.

To assess the utility of machine learning (ML) algorithms in predicting clinically relevant atrial high-rate episodes (AHREs), which can be recorded by a pacemaker. We aimed to develop ML-based models to predict clinically relevant AHREs based on the clinical parameters of patients with implanted pacemakers in comparison to logistic regression (LR). We included 721 patients without known atrial fibrillation or atrial flutter from a prospective multicenter (11 tertiary hospitals) registry comprising all geographical regions of Korea from September 2017 to July 2020. Predictive models of clinically relevant AHREs were developed using the random forest (RF) algorithm, support vector machine (SVM) algorithm, and extreme gradient boosting (XGB) algorithm. Model prediction training was conducted by seven hospitals, and model performance was evaluated using data from four hospitals. During a median follow-up of 18 months, clinically relevant AHREs were noted in 104 patients (14.4%). The three ML-based models improved the discrimination of the AHREs (area under the receiver operating characteristic curve: RF: 0.742, SVM: 0.675, and XGB: 0.745 vs. LR: 0.669). The XGB model had a greater resolution in the Brier score (RF: 0.008, SVM: 0.008, and XGB: 0.021 vs. LR: 0.013) than the other models. The use of the ML-based models in patient classification was associated with improved prediction of clinically relevant AHREs after pacemaker implantation.

Mortality among ischemic and nonischemic heart failure patients with a primary implantable cardioverter-defibrillator.

BACKGROUND: The efficacy of implantable cardioverter defibrillators (ICDs) for primary prevention is controversial in patients with nonischemic heart failure (HF). We evaluated the mortality and predictors of mortality in patients with prophylactic ICD implantation for ischemic and nonischemic HF. METHODS: From 2008 to 2017, 1097 patients (667, nonischemic HF and 430, ischemic HF) who underwent prophylactic ICD implantation, were identified from the Korean National Health Insurance Service database. We used propensity score overlap weighting to correct the differences between two groups. RESULTS: Those with ischemic HF were older (67.0 ± 10.1 vs 61.8 ± 14.2 years), more often male (71.4% vs 63.7%), and had more comorbidities than patients with nonischemic HF. During a median follow-up of 37.3 months (interquartile range [IQR], 14.2-53.8 months), all-cause mortality was higher in unweighted patients with ischemic HF than in those with nonischemic HF (10.9 vs 6.4 per 100 person-years; hazard ratio [HR], 1.74; 95% confidence interval [CI], 1.38-2.20; P < .001). However, after weighting, the annual all-cause mortality rate was similar in both groups (9.5 vs 8.8 per 100 person-years), with no significant difference in the risk of all-cause mortality (HR, 1.08; 95% CI, 0.68-1.71; P = .755). Older age and chronic kidney disease were independent predictors of all-cause mortality in both groups. There was no significant difference in cardiac and noncardiac mortality between the weighted nonischemic and ischemic HF groups. CONCLUSIONS: The all-cause, cardiac, and noncardiac mortality rates were similar between patients with nonischemic and ischemic HF who underwent prophylactic ICD implantation.