入ICU 12h内相关指标构建危重症患者预后预测模型  

The Prognostic Model of Critically Ill Patients was Established within 12 Hours after Admission to ICU

作　　者：李淑娴[1] 翁慧纯[1] 洪春燕[1] LI Shuxian;WENG Huichun;HONG Chunyan(Department of Critical Care,the Central Hospital of Shantou City,Shantou City,Guangdong Province 515031)

机构地区：[1]广东省汕头市中心医院重症医学科,515031

出　　处：《医学理论与实践》2025年第1期34-37,共4页The Journal of Medical Theory and Practice

基　　金：汕头市医疗卫生科技计划项目(181203164010457)。

摘　　要：目的:探讨危重症患者28d死亡的影响因子,以入ICU 12h内指标构建危重症患者预后预测模型。方法:回顾性分析2021年1月—2022年6月我院ICU接收的危重症患者资料,经多因素二元logistic回归模型筛选影响因子,构建预测模型,并采用Nomogram模型展示影响因子对预后影响可视化。结果:共纳入259例危重症患者,入ICU 28d存活180例,死亡79例。多因素logistic回归分析显示年龄(OR=1.021,95%CI:1.002~1.040)、收缩压(OR=0.990,95%CI:0.981~0.999)、血清肌酐(OR=1.142,95%CI:1.018~1.281)、氧合指数(OR=0.998,95%CI:0.996~1.000)、格拉斯哥昏迷评分(OR=0.992,95%CI:0.868~0.981)是危重症患者28d死亡的影响因子(P<0.05);建立危重症患者28d死亡的预测模型为P=1/(1^(+)e^(-z)),Z=0.021×年龄-0.010×SBP^(+)0.133×SCr-0.002×OI-0.081×GCS,该模型对危重症患者预后预测AUC为0.752(95%CI:0.685~0.819,P<0.001),敏感度为0.658,特异度为0.767。结论:入ICU 12h内的年龄、收缩压、血清肌酐、氧合指数、格拉斯哥昏迷评分是本组危重症患者28d死亡的影响因子,依托入ICU 12h内建立的预测模型在早期预测危重症患者有良好的价值。Objective:To explore the influencing factors of 28-day death in critically ill patients,and to build a prognostic model of critically ill patients with indicators within 12 hours after admission to ICU.Methods:The data of critically ill patients were received from the ICU of our hospital from January 2021 to June 2022 were analyzed retrospectively.The influencing factors were screened by multivariate binary logistic regression model,and a prediction model was constructed to visualize the influence of influencing factors on prognosis by Nomogram model.Results:A total of 259 critically ill patients were included,of whom 180 survived and 79 died after 28 days in ICU.The multivariate logistic regression analysis showed that age(OR=1.021,95%CI:1.002~1.040),systolic blood pressure(OR=0.990,95%CI:0.981~0.999),serum creatinine(OR=1.142,95%CI:1.018~1.281),oxygenation index(OR=0.998,95%CI:0.996~1.000),Glasgow coma score(OR=0.992,95%CI:0.868~0.981)were the influential factors for 28-day death in critically ill patients(P<0.05).The prediction model of 28-day death in critically ill patients was established as P=1/(1+e^(-z)),Z=0.021×age-0.010×SBP+0.133×SCr-0.002×OI-0.081×GCS.The predictive AUC of this model for critically ill patients was 0.752(95%CI:0.685~0.819,P<0.001),sensitivity was 0.658,and specificity was 0.767.Conclusion:Age,systolic blood pressure,serum creatinine,oxygenation index and Glasgow coma score within 12 hours of ICU admission are the influential factors for 28-day death of critically ill patients in this study.The predictive model established within 12 hours of ICU admission has good value in early prediction of critically ill patients.

关 键 词：危重症患者 重症监护室 12h内相关指标 预测模型 

分 类 号：R195.4[医药卫生—卫生统计学]