Chen Yun-Yu, Lin Yenn-Jiang, Chien Kuo-Liong, Chao Tze-Fan, Lo Li-Wei, Chang Shih-Lin, Chung Fa-Po, Lin Chin-Yu, Chang Ting-Yung, Kuo Ling, Hsieh Yu-Cheng, Li Cheng-Hung, Chen Shih-Ann
Heart Rhythm Center and Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan.
Institute of Epidemiology and Preventive Medicine College of Public Health, National Taiwan University, Taipei, Taiwan.
Int J Cardiol Heart Vasc. 2021 Apr 28;34:100787. doi: 10.1016/j.ijcha.2021.100787. eCollection 2021 Jun.
The stroke risk scoring system for atrial fibrillation (AF) patients can vary considerably based on patients' status while receiving ablation. This study aimed to demonstrate a novel scoring system for stroke risk stratification based on the status of catheter ablation.
First, 787 patients with AF undergoing ablation were matched according to age, sex, and underlying diseases with the same number of patients not undergoing ablation using the propensity-score (PS)-matched cohort. Multivariate Cox model-derived coefficients were used to construct a simple point-based clinical model using the PS-matched cohort. Thereafter, the novel model (AF-CA-Stroke score) was validated in a nationwide AF cohort.
The AF-CA-Stroke score was calculated based on age (point = 5), ablation status (point = 4), prior history of stroke (point = 4), chronic kidney disease (point = 2), diabetes mellitus (point = 1), and congestive heart failure (point = 1). Risk function to predict the 1-, 5-, 10-year absolute stroke risks was reported. The estimated area under the receive operating characteristic curve of the AF-CA-Stroke score in the PS-matched cohort was 0.845 (95% confidence interval: 0.824-0.865) to predict long-term stroke. A validation study showed that discrimination abilities in the AF-CA-Stroke scores were significantly higher than those in the CHADS/CHADSVASc scores. The best cut-off value of the AF-CA-Stroke score to predict future strokes was ≥ 5.
This novel model-based point scoring system effectively identifies stroke risk using clinical factors and AF ablation status of patients with AF. Various age stratifications and AF ablation should be considered in AF management.
心房颤动（AF）患者的卒中风险评分系统可能会因患者接受消融治疗时的状态而有很大差异。本研究旨在基于导管消融状态建立一种用于卒中风险分层的新型评分系统。
首先，采用倾向评分（PS）匹配队列，将787例接受消融治疗的AF患者按照年龄、性别和基础疾病与相同数量未接受消融治疗的患者进行匹配。使用PS匹配队列，通过多变量Cox模型得出的系数构建一个基于简单点数的临床模型。此后，在全国性AF队列中对该新型模型（AF-CA-卒中评分）进行验证。
AF-CA-卒中评分基于年龄（点数=5）、消融状态（点数=4）、既往卒中史（点数=4）、慢性肾脏病（点数=2）、糖尿病（点数=1）和充血性心力衰竭（点数=1）进行计算。报告了预测1年、5年、10年绝对卒中风险的风险函数。在PS匹配队列中，AF-CA-卒中评分的受试者工作特征曲线下面积估计值为0.845（95%置信区间：0.824-0.865），用于预测长期卒中。一项验证研究表明，AF-CA-卒中评分的鉴别能力显著高于CHADS/CHADSVASc评分。预测未来卒中的AF-CA-卒中评分的最佳截断值为≥5。
这种基于新型模型的点数评分系统利用AF患者的临床因素和AF消融状态有效地识别卒中风险。在AF管理中应考虑不同的年龄分层和AF消融情况。
