利用床旁简易指标及可解释性机器学习算法建立老年重症患者死亡预测模型  

Establishing of mortality predictive model for elderly critically ill patients using simple bedside indicators and interpretable machine learning algorithms

作　　者：孟玉兰 李嘉欣 单新强 路朋宇 黄伟[2,4] Meng Yulan;Li Jiaxin;Shan Xinqiang;Lu Pengyu;Huang Wei(Department of Critical Care Medicine,Tacheng People's Hospital of Ili Kazak Autonomous Prefecture,Tacheng 834300,Xinjiang Uygur Autonomous Region,China;Department of Disease Prevention and Hospital-Acquired Infection Control,First Hospital of Dalian Medical University,Dalian 116012,Liaoning,China;Department of Anesthesiology,Kaifeng Central Hospital,Kaifeng 475000,Henan,China;Department of Critical Care Medicine,First Hospital of Dalian Medical University,Dalian 116012,Liaoning,China)

机构地区：[1]伊犁哈萨克自治州塔城地区人民医院重症医学科,新疆维吾尔自治区伊犁834300 [2]大连医科大学附属第一医院疾病预防与院内感染控制部,辽宁大连116012 [3]河南省开封市中心医院麻醉科,河南开封475000 [4]大连医科大学附属第一医院重症医学科,辽宁大连116012

出　　处：《中华危重病急救医学》2025年第2期170-176,共7页Chinese Critical Care Medicine

基　　金：国家临床重点专科建设项目(2012-475)。

摘　　要：目的探讨基于可解释性的机器学习算法将床旁简易指标纳入老年重症患者死亡预测模型的可行性,为临床的病情评估提供新方案。方法回顾性选择2017年6月至2020年5月在新疆伊犁哈萨克自治州塔城地区人民医院重症监护病房(ICU)住院的年龄≥65岁的老年危重症患者。收集患者人口统计学特征、入院24 h内的基础生命体征、24 h液体出入量等基础数据以及急性生理学与慢性健康状况评分Ⅱ(APACHEⅡ)、格拉斯哥昏迷评分(GCS)和序贯器官衰竭评分(SOFA)等传统评分。根据院内结局将患者分为生存组和死亡组。用基础数据和评分数据分别构成4个数据集,即基线数据集(B),包括年龄、体温、心率、脉搏血氧饱和度、呼吸频率、平均动脉压、尿量、静脉总入量、晶体输液量;B+APACHEⅡ数据集(BA);B+GCS数据集(BG);B+SOFA数据集(BS)。分别用Logistic回归(LR)、极度梯度提升(XGboost)和梯度提升决策树(GBDT)3种机器学习算法在4个数据集训练病死率预测模型,并通过沙普利可加性特征解释方法(SHAP)绘制出每个预测模型的特征重要性柱状图。比较各模型的曲线下面积(AUC)、准确度和F1分数以确定最优预测模型,并绘制模型列线图。结果共纳入392例患者,其中生存组341例,死亡组51例。两组间心率、脉搏血氧饱和度、平均动脉压、静脉总入量、晶体输液量和病因分布差异均存在统计学意义。死亡组前3位病因为休克、脑出血和慢性阻塞性肺疾病。4个数据集经3种机器学习算法训练建立的12个预测模型中,基于B数据集的预后模型总体表现偏后,而以BA数据集训练的LR模型预测效能最优,该模型的AUC为0.767〔95%可信区间(95%CI)为0.692~0.836〕,准确度为0.875(95%CI为0.837~0.903),F1分数为0.190;该模型中排名前3位的变量为24 h晶体输注量、心率和平均动脉压;绘制该模型的列线图结果显示,总分在150~230分预测效能良�Objective To explore the feasibility of incorporating simple bedside indicators into death predictive model for elderly critically ill patients based on interpretability machine learning algorithms,providing a new scheme for clinical disease assessment.Methods Elderly critically ill patients aged≥65 years who were hospitalized in the intensive care unit(ICU)of Tacheng People's Hospital of Ili Kazak Autonomous Prefecture from June 2017 to May 2020 were retrospectively selected.Basic parameters including demographic characteristics,basic vital signs and fluid intake and output within 24 hours after admission,as well acute physiology and chronic health evaluation Ⅱ(APACHEⅡ),Glasgow coma score(GCS)and sequential organ failure assessment(SOFA)were also collected.According to outcomes in hospital,patients were divided into survival group and death group.Four datasets were constructed respectively,namely baseline dataset(B),including age,body temperature,heart rate,pulse oxygen saturation,respiratory rate,mean arterial pressure,urine output volume,infusion volume,and crystal solution volume;B+APACHEⅡ dataset(BA),B+GCS dataset(BG),and B+SOFA dataset(BS).Then three machine learning algorithms,Logistic regression(LR),extreme gradient boosting(XGboost)and gradient boosting decision tree(GBDT)were used to develop the corresponding mortality predictive models within four datasets.The feature importance histogram of each prediction model was drawn by SHapley additive explanation(SHAP)method.The area under curve(AUC),accuracy and F1 score of each model were compared to determine the optimal prediction model and then illuminate the nomogram.Results A total of 392 patients were collected,including 341 in the survival group and 51 in the death group.There were statistically significant differences in heart rate,pulse oxygen saturation,mean arterial pressure,infusion volume,crystal solution volume,and etiological distribution between the two groups.The top three causes of death were shock,cerebral hemorrhage,and chronic obs

关 键 词：危重症患者 老年 病死率 机器学习 预测模型 列线图 

分 类 号：R459.7[医药卫生—急诊医学]