心力衰竭患者出院后90天内不良事件预测模型构建  

Establishment of a risk prediction model for adverse events within 90 days after discharge in the patients with heart failure

作　　者：赵继军 田娜 马磊 Zhao Jijun;Tian Na;Ma Lei(Department of Emergency,the General Hospital of Ningxia Medical University,Yinchuan 750004,China)

机构地区：[1]宁夏医科大学总医院急诊科,宁夏银川750004 [2]宁夏医科大学临床医学院,宁夏银川750004

出　　处：《中国急救医学》2023年第10期788-793,共6页Chinese Journal of Critical Care Medicine

摘　　要：目的构建心力衰竭患者出院后90天内发生不良事件的预测模型。方法回顾性分析发布于PhysioNet的自贡市第四人民医院2016年12月至2019年6月的2008例心力衰竭患者资料,根据出院后90天是否发生不良事件(心源性非计划性再入院或全因死亡)分为事件组和非事件组。采用Lasso-Cox回归分析法筛选影响90天内不良事件的影响因素,以此构建预测模型。结果纳入的1959例患者中524例(26.75%)出现不良事件,全因死亡35例(1.79%),再入院489例(24.96%);基于Lasso-Cox回归筛选出收缩压、纽约心脏病协会(NYHA)分级、查尔森共病指数(CCI)、红细胞分布宽度、肌酐、血清钾和氯化物等7个预测变量构建预测模型,模型验证结果显示,受试者工作特征(ROC)曲线下面积(AUC)为0.653,校准曲线趋于理想,模型一致性良好。Kaplan-Meier风险分层结果显示,高风险组不良事件发生风险高于其他组。结论收缩压、NYHA分级、CCI指数、肌酐、红细胞分布宽度、血清钾和氯化物是心力衰竭患者出院后90天发生不良事件的影响因素,基于此构建的预测模型可有效评估不良事件的风险,帮助临床医师早期合理评估患者病情。Objective To construct a risk prediction model for adverse events within 90 days of discharge in the patients with heart failure.Methods The data of 2008 heart failure patients from Zigong Fourth People’s Hospital from December 2016 to June 2019 published in the PhysioNet network were retrospectively analyzed,and divided into event group and non-event group based on whether adverse events(all-cause death and unplanned cardiac readmission)occurred 90 days after hospital discharge.Lasso-Cox regression analysis was used to screen risk factors for predicting adverse events within 90 days in order to establish a prediction model.Results In the included 1959 patients,524 patients(26.75%)had adverse events,including 35 patients(1.79%)with all-cause death,and 489 patients(24.96%)with readmission;seven predictor variables of systolic blood pressure,NYHA classification,Charlson comorbidity index(CCI),red blood cell distribution width,creatinine,serum potassium and chloride screened by Lasso-Cox regression were used to construct the prediction model.The model verification results showed that the area under the ROC curve(AUC)was 0.653,the correction curve was close to the ideal curve,so the model had good consistency.The Kaplan-Meier risk stratification results showed that the risk of adverse events was higher in the high-risk group than in other risk groups.Conclusions Systolic blood pressure,NYHA classification,CCI index,creatinine,red blood cell distribution width,serum potassium and chloride are influencing factors affecting the occurrence of adverse events in heart failure patients 90 days after discharge.The prediction model established based on the above indicators can assess early the risk of adverse events,and help clinicians to assess early patients′conditions.

关 键 词：心力衰竭 不良事件 风险预测模型 影响因素 查尔森共病指数 红细胞分布宽度 血清钾 氯化物 

分 类 号：R541.6[医药卫生—心血管疾病]