冠心病伴心力衰竭老年患者合并肺部感染风险预测模型的构建

Construction of risk prediction model for lung infection in elderly patients with coronary heart disease and heart failure

目的 分析冠心病伴心力衰竭老年患者合并肺部感染的危险因素,并构建模型。方法 收集2020年1月至2021年9月期间于上海市浦东新区浦南医院收治的160例冠心病伴心力衰竭老年患者病历资料,根据肺部感染情况分为感染组(42例)和未感染组(118例),采用单因素及多因素logistic回归模型分析影响冠心病伴心力衰竭老年患者合并肺部感染的危险因素,建立肺部感染发生的风险预测模型,并采用受试者工作特征曲线(ROC)分析风险预测模型对肺部感染的应用价值。结果 感染组与未感染组在年龄、心功能分级、吸烟史、合并糖尿病、白蛋白(Alb)水平和住院时间等方面比较,差异有统计学意义(P<0.05);年龄≥70岁(OR=3.452,P=0.004)、心功能Ⅲ~Ⅳ级(OR=3.788,P=0.002)、吸烟史(OR=2.450,P=0.030)和住院时间≥30 d(OR=3.259,P=0.007)均为老年冠心病伴心力衰竭患者发生肺部感染的独立危险因素;ROC曲线显示,老年冠心病伴心力衰竭患者发生肺部感染的风险预测模型诊断肺部感染的AUC为0.798,95%CI为0.723~0.873(P<0.001)。结论 年龄、心功能、吸烟史及住院时间与冠心病伴心力衰竭老年患者发生肺部感染有关,风险预测模型对肺部感染具有良好的应用价值。

Objective To analyze the risk factors of lung infection in elderly patients with coronary heart disease and heart failure,and to construct the prediction model.Methods A total of 160 elderly patients with coronary heart disease and heart failure admitted to Shanghai Punan Hospital of Pudong New District were enrolled as the research objects between January 2020 and September 2021.According to presence or absence of lung infection,they were divided into infection group(n=42)and non-infection group(n=118).The risk factors of lung infection were analyzed by univariate analysis and multivariate logistic regression model.The risk prediction model for lung infection was constructed,and its predictive value for lung infection was analyze by receiver operating characteristic(ROC)curves.Results There were significant differences in age,cardiac function grading,smoking history,diabetes mellitus,albumin(Alb)level and hospitalization time between infection group and non-infection group(P<0.05).Age≥70 years(OR=3.452,P=0.004),cardiac function grading at gradeⅢ-Ⅳ(OR=3.788,P=0.002),smoking history(OR=2.450,P=0.030)and hospitalization time≥30 d(OR=3.259,P=0.007)were independent risk factors of lung infection.ROC curves analysis showed that AUC and 95%CI of the model for predicting lung infection were 0.798 and 0.723-0.873(P<0.001).Conclusions Age,cardiac function,smoking history and hospitalization time are related to lung infection in elderly patients with coronary heart disease and heart failure.The risk prediction model has good predictive value for lung infection.