2022 年第 11 期 第 17 卷

个体化预测心房颤动患者射频消融术后早期复发的列线图模型的建立与验证

Establishment and validation of nomogram model for individualized prediction of early recurrence of atrial fibrillation after radiofrequency ablation

作者：李静1周丹2阳国兴3

英文作者：Li Jing1 Zhou Dan2 Yang Guoxing3

单位：1四川省医学科学院·四川省人民医院心内科，成都610072；2四川省德阳市人民医院内分泌科，德阳618000；3四川省医学科学院·四川省人民医院运营管理部，成都610072

英文单位：1Department of Cardiology Sichuan Academy of Medical Sciences·Sichuan Provincal People′s Hospital Chengdu 610072 China; 2Department of Endocrinology Deyang People′s Hospital Sichuan Province Deyang 618000 China; 3Department of Operation Sichuan Academy of Medical Sciences·Sichuan Provincal People′s Hospital Chengdu 610072 China

关键词：心房颤动；射频消融术；早期复发；列线图

英文关键词：Atrialfibrillation;Radiofrequencyablation;Earlyrecurrence;Nomogram

• 摘要：

• 目的 建立个体化预测心房颤动患者射频消融术后早期复发的列线图模型并进行验证。方法 选取2019年1月至2021年3月在四川省人民医院行射频消融术的心房颤动患者230例。根据患者术后3个月心房颤动复发情况分为复发组（71例）和未复发组（159例）。比较2组临床资料，采用二元Logistic回归方法分析患者术后早期复发的独立危险因素，采用R语言软件建立心房颤动患者射频消融术后早期复发的列线图模型并进行验证。结果 复发组患者体重指数、左心房内径和B型脑钠肽（BNP）水平及合并高脂血症、阻塞性睡眠呼吸暂停综合征及持续性心房颤动比例均高于/大于未复发组［(28±3)kg/m2比(22±3)kg/m2、(44±4)mm比(38±4)mm、(531±151)ng/L比(442±151)ng/L、47.9%(34/71)比30.8%(49/159)、43.7%(31/71)比23.9%(38/159)、36.6%(26/71)比17.0%(27/159)］，差异均有统计学意义（均P＜0.05）。Logistic回归分析结果显示，体重指数、BNP、左心房内径增加及合并阻塞性睡眠呼吸暂停综合征和持续性心房颤动是心房颤动患者射频消融术后早期复发的独立危险因素（比值比=1.723、1.677、1.005、3.395、4.177，95%置信区间：1.421～2.090、1.353～2.079、1.002～1.009、1.054～10.934、1.297～13.454，均P＜0.05）。基于以上独立危险因素建立列线图模型，初始列线图的一致性指数为0.938，通过1 000次Bootstrap自举法对模型进行验证后，一致性指数为0.952，且校正曲线显示该模型对预测心房颤动患者射频消融术后早期复发的概率与实际发生概率具有较高的一致性。结论 体重指数、BNP、左心房内径增加及合并阻塞性睡眠呼吸暂停综合征和持续性心房颤动与心房颤动患者射频消融术后早期复发存在密切关系，基于此建立的列线图模型具有较高的预测效能，可为该类患者术后复发的评估提供参考。

• Objective To establish and validate a nomograph model for individualized prediction of early recurrence of atrial fibrillation after radiofrequency ablation. Methods A total of 230 patients with atrial fibrillation were selected, who underwent radiofrequency ablation in Sichuan Provincial People′s Hospital from January 2019 to March 2021. According to the recurrence of atrial fibrillation 3 months after operation, the patients were divided into recurrence group (71 cases) and non recurrence group (159 cases). The clinical data of two groups were compared. Bivariate Logistic regression method was used to analyze the independent risk factors of early postoperative recurrence in patients, and R language software was used to establish a nomogram model for early recurrence of atrial fibrillation after radiofrequency ablation and to validate. Results The body mass index, left atrial diameter, brain natriuretic peptide (BNP) level and the proportions of hyperlipidemia, obstructive sleep apnea syndrome and persistent atrial fibrillation in the recurrence group were higher/greater than those in the non recurrence group［(28±3)kg/m2 vs (22±3)kg/m2, (44±4)mm vs (38±4)mm, (531±151)ng/L vs (442±151)ng/L, 47.9%(34/71) vs 30.8%(49/159), 43.7%(31/71) vs 23.9%(38/159), 36.6%(26/71) vs 17.0%(27/159)］(all P＜0.05). Logistic regression analysis showed that increased body mass index, BNP, left atrial diameter, combined with obstructive sleep apnea syndrome and persistent atrial fibrillation were independent risk factors for early recurrence of atrial fibrillation after radiofrequency ablation（odds ratio=1.723, 1.677, 1.005, 3.395, 4.177, 95% confidence interval: 1.421-2.090, 1.353-2.079, 1.002-1.009, 1.054-10.934, 1.297-13.454, all P＜0.05）. Based on the above independent risk factors, the nomogram model was established. The concordance index of the initial nomogram was 0.938. After 1 000 times of Bootstrap method to validate the model, the concordance index was 0.952, and the calibration curve showed that the model probability has a high consistency to actual probability in predicting the probability of early recurrence of atrial fibrillation after radiofrequency ablation. Conclusions The increase of body mass index, BNP, left atrial diameter and combined with obstructive sleep apnea syndrome and persistent atrial fibrillation are closely related to the early recurrence of atrial fibrillation in patients after radiofrequency ablation. The nomogram model based on this has high predictive efficacy. It can provide reference for the evaluation of postoperative recurrence of the patients.