冠状动脉粥样硬化性心脏病合并心力衰竭患者院内死亡风险预测模型的构建  

Construction of a predictive model for in-hospital mortality risk in patients with coronary artery atherosclerotic heart disease and heart failure

作　　者：刘成[1] 刘森 金梦龙 刘紫阳 付真彦[1] 马依彤[1] LIU Cheng;LIU Sen;JIN Menglong;LIU Ziyang;FU Zhenyan;MA Yitong(Heart Center,the First Affiliated Hospital of Xinjiang Medical University,Urumqi 830054,China)

机构地区：[1]新疆医科大学第一附属医院心血管病中心,乌鲁木齐830054

出　　处：《中国分子心脏病学杂志》2025年第1期6553-6559,共7页Molecular Cardiology of China

基　　金：新疆维吾尔自治区重点研发计划项目(2022B03022,2022B03022-4)。

摘　　要：目的 构建冠状动脉粥样硬化性心脏病(CAD)合并心力衰竭患者院内死亡风险的预测模型。方法 回顾性分析2012年4月至2023年9月新疆医科大学第一附属医院收治的CAD合并心力衰竭的患者共4 215例。收集患者的临床资料,以7∶3随机划分为训练集(n=2 950)与验证集(n=1 265)。采用二元logistic回归分析筛选CAD合并心力衰竭患者院内死亡的独立危险因素,并构建受试者操作特征(ROC)曲线、校准曲线、决策分析曲线(DCA)及列线图预测模型。再用验证集对预测模型进行验证。结果 多因素logistic回归分析结果显示,年龄≥65岁、有脑梗死史、血红蛋白水平降低、D-二聚体≥280 ng/mL、甘油三酯-葡萄糖指数(TyG)升高是CAD合并心力衰竭患者院内死亡的独立危险因素(P均<0.05)。采用以上5个指标构建列线图预测模型,校准曲线评估模型的有效性χ^(2)=4.21,P=0.84。训练集模型ROC曲线下面积为0.770,灵敏度为79.5%,特异度为62.0%,DCA曲线显示在一定区间内该模型净获益率具有临床实用价值。进一步用验证集对预测模型进行评估,结果显示,ROC曲线下面积为0.774,灵敏度为80.7%,特异度为65.4%,校准曲线χ^(2)=11.10,P=0.20。结论 年龄≥65岁、有脑梗死史、血红蛋白水平降低、D-二聚体≥280 ng/mL、TyG升高是CAD合并心力衰竭患者院内死亡的独立危险因素,以此构建的预测模型具有较高的区分度及有效性,临床实用性较好,可作为预测CAD合并心力衰竭患者院内死亡的有效工具。Objective To develop a predictive model for estimating the risk of in-hospital mortality among patients with coronary artery atherosclerotic heart disease(CAD) complicated by heart failure. Methods This study enrolled 4 215 patients diagnosed with CAD and heart failure during hospitalization at the First Affiliated Hospital of Xinjiang Medical University from April 2012 to September 2023. Clinical data were collected and the cohort was randomly divided into a training set(n=2 950) and a validation set(n=1 265) at a ratio of 7∶3. Binary logistic regression was employed to analyze independent risk factors for mortality in CAD patients with heart failure, subsequently constructing receiver operating characteristic(ROC) curves, calibration curves, decision curves(DCA), and nomograms. The model was then validated using the validation set. Results Multivariate logistic regression analysis demonstrated that advanced age(≥ 65 years old), a history of cerebral infarction, decreased hemoglobin level, D-dimer level ≥280 ng/mL, and increased triglyceride-glucose index(TyG) were independent risk factors for in-hospital death in patients with CAD complicated with heart failure(all P<0.05). A nomogram prediction model was constructed using the above five indicators, and a calibration curve was drawn to evaluate the validity of the model(χ^(2)=4.21, P=0.84). In the training set, the area under the receiver operating characteristic(ROC) curve of the model was 0.770, with a sensitivity of 79.5% and a specificity of 62.0%. The decision curve analysis(DCA) showed that the net benefit rate of this model had clinical practical value within a certain range. Furthermore, the model was evaluated using the validation set, and the results showed that the area under the ROC curve was 0. 774, the sensitivity was80. 7%, the specificity was 65. 4%, and the calibration curve (χ^(2)= 11.10,P= 0. 20).Conclusions Advanced age (≥ 65 yearsold), a history of cerebral infarction, decreased hemoglobin level, D-dimer level ≥280 ng/mL, and in

关 键 词：冠状动脉粥样硬化性心脏病 心力衰竭 列线图 预测模型 

分 类 号：R541.4[医药卫生—心血管疾病]