摘要
目的 基于重症医学信息数据库(MIMIC-IV)中发生血流感染的危重疾病患者相关数据构建预测危重疾病患者革兰阴性菌血流感染发生风险预测模型,以期为预测危重疾病患者发生革兰阴性菌血流感染风险提供新的方法。方法 收集MIMIC-IV中2 503例发生血流感染的危重疾病患者的临床资料及实验室检查指标[血常规(红细胞、白细胞、血小板等)、血生化(钾离子、钙离子、氯离子、碳酸氢根、阴离子间隙和尿素氮等)、凝血功能指标(INR、PT、PTT)]数据。将所有危重疾病患者以7∶3的比例分为训练集(1 752例)和验证集(751例)。在训练集中使用LASSO回归初步筛选出危重疾病患者发生革兰阴性菌血流感染的影响因素,并将筛选出来的影响因素行多因素Logistic回归分析,建立危重疾病患者革兰阴性菌血流感染风险预测模型(列线图模型)。在训练集和验证集中通过受试者工作特征曲线(ROC)、校准曲线和决策曲线(DCA)分别对列线图模型的区分度、一致性、临床适用性进行评价。结果 年龄、患有肿瘤、肝胆系统疾病、嗜酒史、钾离子、钙离子、碳酸氢根、阴离子间隙和尿素氮为危重疾病患者发生革兰阴性菌血流感染风险的影响因素,基于以上影响因素采用Logistic回归分析,构建列线图模型。训练集和验证集中,列线图模型预测危重疾病患者发生革兰阴性菌血流感染的ROC下面积分别为0.711(95%CI 0.667~0.756)、0.705(95%CI 0.678~0.733);校准曲线表明列线图模型预测革兰阴性菌血流感染发生的结果与实际结果之间具有良好的一致性(P=0.764);DCA显示列线图模型具有良好的临床适用性。结论 成功构建了预测危重疾病患者发生革兰阴性菌血流感染风险的列线图模型,该模型预测性能较好,能有效识别血流感染高风险危重疾病患者。
Objective To construct the prediction model for risk of Gram-negative bacterium(GNB) bloodstream infection(BSI) based on data of patients with blood flow infection from the Medical Information Mart for Intensive Care IV(MIMIC-IV) database,aiming to provide a new method for predicting the risk of GNB-BSI.Methods Clinical data and laboratory indexes [blood routine(red blood cells,white blood cells,platelets,etc.),blood biochemistry(potassium ions,calcium ions,chloride ions,bicarbonate,anion gap and urea nitrogen,etc.),coagulation function indicators(INR,PT,PTT)] of 2 503 patients with bloodstream infection from MIMIC-IV database were collected.These data were divided into the training set(1 752 cases) and the validation set(751 cases) at a ratio of 7:3.The LASSO regression was employed to select the factors influencing the incidence of GNB-BSI in the training dataset.These factors were subjected to multivariate Logistic regression analysis in order to establish a nomogram prediction model for the risk of developing GNBBSI.The differentiation,consistency,and clinical practicality of the model was assessed using receiver operating characteristic(ROC) curve,calibration curve,and decision curve analysis(DCA) in both the training and validation sets.Results Age,cancer,hepatobiliary disease,alcohol abuse,potassium,calcium,bicarbonate,anion gap,and urea nitrogen were independent factors influencing the risk of GNB-BSI.Logistic regression analysis was conducted to develop a risk prediction model for GNB-BSI(nomogram model) based on these factors.The area under ROC curve of the model was 0.711(95%CI = 0.667-0.756) in the training set and 0.705(95%CI = 0.678-0.733) in the validation set.The calibration curve exhibited satisfactory consistency between predicted and actual outcomes for GNB-BSI(P=0.764).The DCA showed that the nomogram model had good clinical practicability.Conclusion A nomogram model for risk of GNB-BSI was established successfully,which had good predictive performance,and effectively identified high-risk patients.
作者
陈秋宇
秦泽辉
刘享田
叶莉萍
田行瀚
CHEN Qiuyu;QIN Zehui;LIU Xiangtian;YE Liping;TIAN Xinghan(The Second School of Clinical Medicine,Binzhou Medical University,Yantai 264000,China)
出处
《山东医药》
CAS
2024年第3期51-55,共5页
Shandong Medical Journal
基金
山东省自然科学基金(ZR2020MH197)
山东省医药卫生科技发展计划项目(2019WS281)。
关键词
感染风险预测模型
列线图
革兰阴性杆菌感染
血流感染
革兰阴性杆菌血流感染
重症医学信息数据库
infection risk prediction model
nomogram
Gram-negative bacterial infection
bloodstream infection
Gram-negative bacterial bloodstream infection
Medical Information Mart for Intensive Care IV database
