摘要
目的 构建冠心病(CHD)伴房颤患者经皮冠脉介入术(PCI)术后1 年双抗治疗出血风险的列线图预测模型.方法 回顾性选取我院2019 年6 月至2022 年4 月收治的428 例CHD合并房颤患者为研究对象,患者入院后均采取PCI术治疗,术后均采用双联抗血小板治疗并随访1 年,统计出血事件发生率.结果 随访过程中 60 例患者发生出血事件,发生率为14.02%(60/428).经组间比较、相关性分析及LASSO回归进行变量筛选,筛选出年龄、血肌酐(Scr)、C-反应蛋白(CRP)、CRUSADE评分、HAS-BLED评分、凝血酶原时间(PT)、手术时间、D-二聚体、最大振幅(MA值)、凝血指数(CI值)10 个连续性变量和既往出血史分类变量;多因素Logistic回归分析显示,既往出血史、Scr、CRUSADE评分、HAS-BLED评分、D-二聚体、MA值是影响患者PCI术后有无出血的独立危险因素(P＜0.05).上述危险因素预测患者出血风险的ROC曲线下面积分别为0.734、0.742、0.772、0.843、0.793、0.782,同时构建的预测患者出血风险的列线图模型C-in-dex为0.968(95%CI 0.941～0.996),重复抽样内部验证法对列线图模型进行验证并绘制校准曲线,Hosmer-Lemeshow拟合优度检验值为1.764,P =0.984,DCA决策曲线显示列线图模型在预测患者出血风险的净收益率较高.结论 既往出血史、Scr、CRUSADE评分、HAS-BLED评分、D-二聚体、MA值是影响CHD合并房颤患者PCI术后出血的独立危险因素,基于上述变量构建的列线图模型能够有效预测出血事件的发生.
Abstract
Objective To construct a nomogram prediction model for the risk of bleeding in patients with coronary heart disease(CHD)and atrial fibrillation after percutaneous coronary intervention(PCI)with dual antibody therapy for 1 year.Methods A retrospective study was conducted on 428 patients with CHD complicated with atrial fibrillation admitted to our hospital from June 2019 to April 2022.After admission,all patients underwent PCI surgery,and after surgery,dual antiplatelet therapy was used and followed up for 1 year.The incidence of bleeding events was statistically analyzed.Results During the follow-up process,60 patients experienced bleeding events,with an incidence rate of 14.02%(60/428).Through inter group comparison,correlation analysis and LASSO regression,10 continuous variables and categorical variable of previous bleeding history were screened,including age,serum creatinine(Scr),C-reactive protein(CRP),CRUSADE score,HAS-BLED score,prothrombin time(PT),operation time,D-dimer,maximum amplitude(MA value),coagulation index(CI value).Multifactor Logistic regression analysis showed that previous bleeding history,Scr,CRUSADE score,HAS-BLED score,D-dimer and MA value were independent risk factors for bleeding after PCI(P＜0.05).The area under the receiver operating characteristic of the above risk factors to predict the bleeding risk of patients was 0.734,0.742,0.772,0.843,0.793,0.782,respectively.At the same time,the C-index of the nomogram model constructed to predict the bleeding risk of patients was0.968(95%CI0.941～0.996).The internal validation method of repeated sampling was used to verify the nomogram model and draw a calibration curve.The goodness of fit test value of Hosmer Lemeshow was 1.764,P =0.984.The DCA decision curve showed that the nomogram model had a higher net return rate in predicting patient bleeding risk.Conclusion Previous bleeding history,Scr,CRUSADE score,HAS-BLED score,D-dimer,MA value are independent risk factors for postopera-tive bleeding in CHD patients with atrial fibrillation after PCI.The nomogram model based on the above variables can effectively pre-dict the occurrence of bleeding events.
维普
万方数据