ICU患者亚谵妄综合征风险预测模型的构建及验证  

Development and validation of a risk prediction model for subdelirium syndrome in adult ICU patients

作　　者：宋亚亚 江淑敏[2] 王淑玲[2] 祝燕杰 刘亭亭 Song Yaya;Jiang Shumin;Wang Shuling;Zhu Yanjie;Liu Tingting(Nursing Department,The First Affiliated Hospital of Shandong First Medical University/Qianfoshan Hospital of Shandong Province,Jinan 250014,Shandong Province,China;Intensive Care Unit,The First Affiliated Hospital of Shandong First Medical University/Qianfoshan Hospital of Shandong Province,Jinan 250014,Shandong Province,China)

机构地区：[1]山东第一医科大学第一附属医院(山东省千佛山医院)护理部,山东济南250014 [2]山东第一医科大学第一附属医院(山东省千佛山医院)重症监护室,山东济南250014

出　　处：《现代临床护理》2025年第1期16-23,共8页Modern Clinical Nursing

基　　金：山东第一医科大学青年科学基金培育资助计划项目,项目编号为202202-022。

摘　　要：目的 探讨成人重症监护室(intensive care unit,ICU)患者亚谵妄综合征(subsyndromal delirium,SSD)的危险因素,构建风险预测模型并进行验证。方法 纳入2022年6月—2024年1月山东省某三级甲等综合医院的ICU患者380例,将其中70%(n=266)作为建模组,30%(n=114)作为验证组,通过比较发生SSD的患者与未发生SSD的患者一般资料、疾病与治疗、实验室检查指标等,筛选出危险因素,并使用Logistic回归构建风险预测模型。结果 构建的风险预测模型包括年龄(OR=1.023)、急性生理与慢性健康状况评分Ⅱ(acute physiology and chronic health evaluationⅡ,APACHEⅡ)(OR=1.093)、重症监护疼痛观察工具(critical care pain observation tool,CPOT)评分(OR=2.216)、机械通气时间(OR=1.003)、约束(OR=2.615)、脓毒症(OR=2.081)6个风险因素。模型预测效果分析结果显示,Calibration图示校准曲线与理想曲线接近重合;受试者工作特征(receiver operating characteristic,ROC)曲线下面积为0.816,预测临界值85分。模型内部验证结果显示,Calibration图示校准曲线接近理想曲线,ROC曲线下面积0.808。结论 成人ICU患者亚谵妄风险预测模型具有较好的一致性及预测效能,可为临床早期筛查亚谵妄综合征并进行干预提供依据。Objective To investigate the risk factors of subdelirium syndrome in adult ICU patients,then establish a risk prediction model and have it verified.Method A total of 380 adult ICU patients in our hospital between June 2022 and January 2024 were selected in this study.Among the patients,224(70%)were assigned to the model group and 114(30%)to the validation group.Independent risk factors were screened by comparison of the general data,disease and treatment,laboratory indicators and other relevant data between the patients with and without subdelirium.A risk prediction model was established with Logistic regression.Results A risk prediction model was finalised and established.It composed six risk factors of age(OR=1.023),APACHE II score(OR=1.093),critical care pain observation tool(OR=2.216),duration of mechanical ventilation(OR=1.003),constraint(OR=2.615)and sepsis(OR=2.081).In internal validation,it was found that the calibration curve closely overlapped with the ideal curve and the area under the ROC curve was 0.816,with a predictive cut off value of 85 points.In external validation,the calibration curve was found closely overlapped with the ideal curve and the area under ROC curve was 0.808.Conclusion The risk prediction model for subdelirium syndrome in adult ICU patients established in the study has good consistency and prediction efficiency,thereby it provides a basis for early clinical screening and intervention of subdelirium syndrome.

关 键 词：重症监护室 亚谵妄综合征 预测模型 

分 类 号：R47[医药卫生—护理学]