房颤患者死亡风险预测模型研究进展

Advances in Mortality Risk Prediction Modeling for Atrial Fibrillation

摘要: 心房颤动(atrial fibrillation, AF)是最常见的持续性心律失常，其患者的心血管死亡和全因死亡风险均显著升高，疾病负担日益加重。准确识别高危人群并进行死亡风险分层，对改善患者预后具有重要意义。目前国内外已建立了多种用于预测房颤患者死亡风险的临床模型，主要包括基于传统临床变量的风险评分(如CHA₂DS₂-VASc评分及其改良版本)、整合生物标志物的风险评分以及基于大型前瞻性真实世界研究的预测模型。近年来，随着机器学习与深度学习方法的不断进展，新型预测模型不断涌现。本文旨在介绍不同房颤死亡风险预测模型，以期为临床医护人员提供参考。

Abstract: Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, which is associated with a markedly elevated risk of cardiovascular and all-cause mortality, thereby imposing the growing burden of society. Precise risk stratification and early identification of high-risk individuals are critical for optimizing prognostic management and improving clinical outcomes. Currently, various predictive models for mortality risk in AF patients have been established, including risk scores based on clinical variables (such as the CHA₂DS₂-VASc score and its modified versions), risk scores incorporating biomarkers, and predictive models derived from large-scale prospective real-world studies. In recent years, with the advancement of machine learning and deep learning, novel predictive models have emerged. In this paper, we will introduce different risk prediction models for mortality in AF patients, providing a practical reference for clinical decision-making.