急性心肌梗死患者院内新发心房颤动的影响因素及其风险预测列线图模型构建

Influencing Factors and Construction of Risk Prediction Nomogram Model of New Onset Atrial Fibrillation during Hospitalization in Patients with Acute Myocardial Infarction

背景急性心肌梗死(AMI)患者并发心房颤动会影响其血流动力学,增加患者死亡风险。而整合AMI患者新发心房颤动的危险因素并构建风险预测列线图模型,有助于识别新发心房颤动高风险人群。目的探讨AMI患者院内新发心房颤动的影响因素,并构建其风险预测列线图模型。方法选取2017年1月至2021年3月在青岛市市立医院行急诊经皮冠状动脉介入术(PCI)的513例AMI患者作为研究对象,按照患者院内是否新发心房颤动分为房颤组(n=82)和非房颤组(n=431)。通过医院信息系统收集AMI患者人口学特征及入院时临床资料。比较两组患者人口学特征及入院时临床资料,采用多因素Logistic回归模型分析AMI患者院内新发心房颤动的影响因素,采用R 3.4.3软件包绘制列线图模型,绘制受试者工作特征(ROC)曲线以评估列线图模型的预测效能,并采用Bootstrap方法重复抽样1000次验证列线图模型的预测效能。结果房颤组年龄和左心房内径(LAD)大于非房颤组,有吸烟史者所占比例、糖尿病发生率、Gensini积分、N末端B型利钠肽前体(NT-proBNP)高于非房颤组(P<0.05)。多因素Logistic回归分析结果显示,年龄〔OR=1.047,95%CI(1.011,1.085)〕、吸烟史〔OR=1.828,95%CI(1.042,3.206)〕、糖尿病〔OR=3.073,95%CI(1.622,5.819)〕、LAD〔OR=1.138,95%CI(1.069,1.211)〕、Gensini积分〔OR=1.048,95%CI(1.032,4.064)〕、NT-proBNP〔OR=1.114,95%CI(1.069,2.005)〕均是AMI患者院内新发心房颤动的独立影响因素(P<0.05)。将上述6个独立影响因素作为预测指标,构建AMI患者院内新发心房颤动风险预测列线图模型。ROC曲线分析结果显示,列线图模型预测AMI患者院内新发心房颤动的曲线下面积(AUC)为0.839〔95%CI(0.786,0.892)〕,说明列线图模型的区分能力较好。Bootstrap方法结果显示,校准曲线的平均绝对误差(MAE)为0.019,说明校准曲线与理想曲线贴合良好。结论年龄、吸烟史、糖尿病、LAD、Gensini积分及NT-proBNP是AMI患者院内新发心房颤动的独立影响因素,而基于上述影响因素构建的AMI患者院内新发心房颤动风险预测列线图模型具有较高的预测效能。

Background Acute myocardial infarction(AMI)with atrial fibrillation will affect the patients'hemodynamics and increase the risk of death.Integrating the risk factors of new onset atrial fibrillation in patients with AMI and constructing a risk prediction nomogram model will help identify high-risk groups of new onset atrial fibrillation.Objective To explore the influencing factors of new onset atrial fibrillation during hospitalization in patients with AMI,and establish a risk prediction nomogram model.Methods A total of 513 AMI patients who underwent emergency percutaneous coronary intervention(PCI)in Qingdao Municipal Hospital from January 2017 to March 2021 were selected as the research objects,they were divided into atrial fibrillation group(n=82)and non atrial fibrillation group(n=431)according to whether occured new onset atrial fibrillation during hospitalization.The demographic features and clinical data at admission were collected by hospital information system.The demographic features and clinical data at admission were compared between the two groups.The influencing factors of new onset atrial fibrillation during hospitalization in patients with AMI were analyzed by multivariate Logistic regression model.The nomogram model was drawn by R 3.4.3 software package,and the receiver operating characteristic(ROC)curve was drawn to evaluate the prediction efficiency of the nomogram model,the Bootstrap method(self-sampling 1000 times)was used to verify the prediction efficiency of the nomogram model.Results Age and left atrial diameter(LAD)in atrial fibrillation group were larger than those in non atrial fibrillation group,and the proportion of patients with smoking history,incidence of diabetes,Gensini score and N-terminal pro-B-type natriuretic peptide(NT-proBNP)were higher than those in non atrial fibrillation group(P<0.05).The multivariate Logistic regression analysis results showed that age[OR=1.047,95%CI(1.011,1.085)],smoking history[OR=1.828,95%CI(1.042,3.206)],diabetes[OR=3.073,95%CI(1.622,5.819)],LAD[OR=1.138,95%CI(1.069,1.211)],Gensini score[OR=1.048,95%CI(1.032,4.064)]and NT-proBNP[OR=1.114,95%CI(1.069,2.005)]were independent influencing factors of new onset atrial fibrillation during hospitalization in AMI patients(P<0.05).The above six independent influencing factors were used as predictors to construct a nomogram model for predicting the risk of new onset atrial fibrillation in patients with AMI.ROC curve analysis showed that the area under curve(AUC)of nomogram model in predicting new onset atrial fibrillation during hospitalization in patients with AMI was 0.839[95%CI(0.786,0.892)],indicating that the nomogram model has good discrimination ability.Bootstrap method result showed that the MAE of the calibration curve was 0.019,indicating that the correction curve fit well with the ideal curve.Conclusion Age,smoking history,diabetes,LAD,Gensini score and NT-proBNP are independent influencing factors of new onset atrial fibrillation during hospitalization in AMI patients.The nomogram model based on the above independent influencing factors in predicting the risk of new onset atrial fibrillation during hospitalization in AMI patients has a higher predictive power.