Predicting All-Cause Mortality Risk in Atrial Fibrillation Patients: A Novel LASSO-Cox Model Generated From a Prospective Dataset

Abstract

Background: Although mortality remains high in patients with atrial fibrillation (AF), there have been limited studies exploring machine learning (ML) models on mortality risk prediction in patients with AF.Objectives: This study sought to develop an ML model that captures important variables in order to predict all-cause mortality in AF patients.Methods: In this single center prospective study, an ML-based mortality prediction model was developed and validated using a dataset of 2,012 patients who experienced AF from November 2018 to February 2020 at the First Affiliated Hospital of Shantou University Medical College. The dataset was randomly divided into a training set (70%, n = 1,223) and a validation set (30%, n = 552). A total of 122 features were collected for variable selection. Least absolute shrinkage and selection operator (LASSO) and random forest (RF) algorithms were used for variable selection. Ten ML models were developed using variables selected by LASSO or RF. The best model was selected and compared with conventional risk scores. A nomogram and user-friendly online tool were developed to facilitate the mortality predictions and management recommendations.Results: Thirteen features were selected by the LASSO regression algorithm. The LASSO-Cox model achieved an area under the curve (AUC) of 0.842 in the training dataset, and 0.854 in the validation dataset. A nomogram based on eight independent features was developed for the prediction of survival at 30, 180, and 365 days following discharge. Both the time dependent receiver operating characteristic (ROC) and decision curve analysis (DCA) showed better performances of the nomogram compared to the CHA2DS2-VASc and HAS-BLED models.Conclusions: The LASSO-Cox mortality predictive model shows potential benefits in death risk evaluation for AF patients over the 365-day period following discharge. This novel ML approach may also provide physicians with personalized management recommendations.

An Online Pre-procedural Nomogram for the Prediction of Contrast-Associated Acute Kidney Injury in Patients Undergoing Coronary Angiography

“…LASSO and RF algorithms were employed to determine optimal predictors by R package "glmnet" and "randomForest, " respectively. Intersective predictors of both algorithms were eventually assembled into a binary logistic regression model (25). Nomogram was developed and visualized by R package "regplot."…”

“…LASSO and RF algorithms were employed to determine optimal predictors by R package “glmnet” and “randomForest,” respectively. Intersective predictors of both algorithms were eventually assembled into a binary logistic regression model ( 25 ). Nomogram was developed and visualized by R package “regplot.” Furthermore, the accuracy of the nomogram was estimated by calibration plots, and a likelihood ratio statistic (unreliability U index) was given to test the null hypothesis of calibration that intercept = 0, slope = 1 ( 26 ).…”

Machine learning in the detection and management of atrial fibrillation

“…The authors were able to show that the developed algorithm was able to detect AF episodes up to 4.5 min before onset, possibly allowing for the development of algorithm-guided interventions before the onset of an AF episode, such as the intake of antiarrhythmic medication. Additionally, DNNs were shown to improve the estimation of AF-associated risks with working groups being able to show an improvement of estimation of all-cause mortality [ 42 ] and neurological outcome after an AF-related stroke [ 43 ]. A recent DNN-based analysis of pooled data from nine double-blinded, randomized, placebo-controlled trials evaluating betablockers in heart failure was also able to detect a mortality benefit in young patients with reduced LVEF and AF [ 44 ].…”