基于心脏超声构建心房颤动射频消融术后复发的预测模型

Predictive Model of Atrial Fibrillation Recurrence After Radiofrequency Ablation Based on Cardiac Ultrasand

目的基于心脏超声构建心房颤动射频消融术后复发的预测模型,评价其临床效果。方法选取2021年7月至2023年7月郑州大学第一附属医院收治的306例接受心房颤动射频消融术患者为研究对象,根据术后6个月内复发情况分为复发组、未复发组。比较两组临床资料及心脏超声定量参数,分析术后复发的影响因素,构建术后复发的预测模型,并进行评价与验证。结果306例患者术后6个月内复发89例,复发率为29.08%;复发组高血压、阻塞性睡眠呼吸暂停低通气综合征(OSAHS)、慢性心力衰竭患者占比高于未复发组,病程长于未复发组,左房前后径、右房横径、肺动脉收缩压、经二尖瓣左室舒张早期血流速度/组织多普勒成像舒张早期二尖瓣环运动速度(E/e’)高于未复发组(P<0.05)。高血压、OSAHS、慢性心力衰竭、病程、左房前后径、E/e’均是术后复发的相关影响因素(P<0.05);构建预测模型,该模型的C-index为0.899,评价与验证结果显示该模型具有较高预测价值及明显临床正向净收益。结论左房前后径、慢性心力衰竭、病程、E/e’、高血压、OSAHS是心房颤动射频消融术后复发的独立影响因素,基于上述因素构建随机森林预测模型具有较好预测效能及临床正向净收益,可为临床早期预测术后复发情况提供参考,并对后续临床决策具有一定指导意义。

Objective To construct a prediction model of atrial fibrillation recurrence after radiofrequency ablation based on cardiac ultrasand,and to evaluate its clinical effectiveness.Methods A total of 306 patients undergoing radiofrequency ablation of atrial fibrillation in the First Affiliated Hospital of Zhengzhou University from July 2021 to July 2023 were selected as the study objects,and were divided into relapse group and non-recurrence group according to the recurrence situation within 6 months after surgery.The clinical data and quantitative parameters of cardiac ultrasound in the two groups were compared,and the factors affecting postoperative recurrence were analyzed,and the prediction model of postoperative recurrence was constructed,and evaluated and verified.Results A total of 89 of 306 patients recurred within 6 months after operation,the recurrence rate was 29.08%.The proportion of patients with hypertension,obstructive sleep apnea hypopnea syndrome(OSAHS)and chronic heart failure in the recurrent group was higher than that in the non-recurrent group,the course of disease was longer than that in the non-recurrent group,and left anteroposterior atrial diameter,right atrial transversal diameter,pulmonary artery systolic blood pressure,early mitral annulus motion velocity(E/e’)were higher than the non-recurrence group(P<0.05).Logistic regression analysis showed that hypertension,OSAHS,chronic heart failure,disease duration,left anterior and posterior atrial diameter,E/e’were related factors for postoperative recurrence(P<0.05).A prediction model was constructed,which had a C-index of 0.899,and the results of evaluation and validation showed that the model had high predictive value and obvious clinical positive net benefit.Conclusion Left anterior and posterior atrial diameter,chronic heart failure,disease duration,E/e’,hypertension and OSAHS are independent influencing factors for recurrence of atrial fibrillation after radiofrequency ablation.The random forest prediction model based on the above factors has good predictive efficacy and positive clinical net benefit,and can provide reference for early clinical prediction of postoperative recurrence.And it has certain guiding significance for the follow-up clinical decision.