重症肺结核死亡的多因素分析及风险模型建立  

Multivariate analysis of death from severe pulmonary tuberculosis and establishment of risk model

作　　者：王鑫[1] 魏云霞 龙雪娟[2] 闫丽静[1] 邵丽娇 郄东磊[1] 李宁[1] WANG Xin;WEI Yunxia;LONG Xuejuan;YAN Lijing;SHAO Lijiao;QIE Donglei;LI Ning(Department of Intensive Care Medicine,Hebei Key Laboratory of Lung Disease,Hebei Chest Hospital,Shijiazhuang,Hebei 050000,China;Department of Emergency,Hebei Key Laboratory of Lung Disease,Hebei Chest Hospital,Shijiazhuang,Hebei 050000,China)

机构地区：[1]河北省胸科医院重症医学科,河北省肺病重点实验室,河北石家庄050000 [2]河北省胸科医院急诊科,河北省肺病重点实验室,河北石家庄050000

出　　处：《临床肺科杂志》2024年第7期1078-1083,共6页Journal of Clinical Pulmonary Medicine

摘　　要：目的分析ICU中重症肺结核患者死亡的危险因素,构建风险模型,协助临床诊治。方法回顾性收集2015年1月至2020年12月河北省胸科医院ICU收治的118例重症肺结核患者的28项指标。包括一般资料[年龄、性别、体重指数(BMI)]和入ICU时相关指标[血常规、动脉血气分析、生化指标、凝血功能、感染指标及APACHEⅡ评分、SOFA评分]。依据入ICU后28天预后情况分为生存组和死亡组,比较两组各项指标之间的差异,对差异有统计学意义的指标应用二元Logistic回归分析,筛选与重症肺结核患者ICU 28天相关的独立危险因素。结果重症肺结核入ICU 28天生存组与死亡组相比,年龄、性别、BMI、APACHEⅡ评分、WBC等13项指标具有统计学差异(P<0.05),进一步行二元Logistic回归分析显示,年龄、性别、BMI、APACHEⅡ、Lac及HCT是患者入ICU 28天死亡的独立影响因素,其中BMI(OR=0.603,95%CI 0.450~0.808,P=0.001)及HCT(OR=0.864,95%CI 0.733~0.976,P=0.022)为保护性因素,男性(OR=6.575,95%CI 1.118~38.679,P=0.037)、年龄(OR=1.070,95%CI 1.027~1.114,P=0.001)、APACHEⅡ评分(OR=1.288,95%CI 1.149~1.444,P<0.001)及Lac(OR=2.742,95%CI 1.425~5.275,P=0.003)为危险因素;构建死亡风险模型logit P=In(P/1-P)=2.616+0.068(年龄)+1.883(男性)+0.253(APACHEⅡ评分)+1.009(Lac)-0.506(BMI)-0.168(HCT)。受试者工作特征曲线(ROC)分析显示,该模型用于重症肺结核入ICU 28天的预后评估,AUC为0.96(95%CI 0.928~0.993),敏感度为68.5%,特异度为89.7%,具有较高的预测准确率。结论重症肺结核患者死亡率高,年龄、性别、BMI、APACHEⅡ、Lac及HCT对重症肺结核患者ICU 28天预后具有较好的预测能力。Objective To analyze the risk factors of death in ICU patients with severe pulmonary tuberculosis,and establish a risk model to assist clinical diagnosis and treatment.Methods The 28 indexes of 118 patients with severe pulmonary tuberculosis admitted to the ICU of Hebei Chest Hospital from January 2015 to December 2020 were retrospectively collected.It included general data[age,sex,body mass index(BMI)]and relevant indicators at ICU admission[blood routine,arterial blood gas analysis,biochemical indicators,coagulation function,infection indicators,APACHE II score,SOFA score].The survival group and death group were divided into survival group and death group according to the 28-day prognosis after admission to the ICU.The differences between the two groups were compared.Binary Logistic regression analysis was applied to the indicators with statistical significance,and independent risk factors related to the 28-day ICU of patients with severe tuberculosis were screened.Results Compared with the death group,age,gender,BMI,APACHEⅡscore,WBC,and other 13 indicators in the 28-day survival group of severe tuberculosis admitted to ICU had statistical differences(P<0.05).Further binary Logistic regression analysis showed that Age,sex,BMI,APACHEⅡ,Lac,and HCT were independent factors of 28-day death,among which BMI(OR=0.603,95%CI 0.45~0.808,P=0.001)and HCT(OR=0.864,95%CI 0.733~0.976,respectively)were independent factors of 28-day death.P=0.022 was the protective factor,male(OR=6.575,95%CI 1.118~38.679,P=0.037),age(OR=1.07,95%CI 1.027~1.114,age(OR=6.575,95%CI 1.118~38.679,P=0.037).P=0.001),APACHEⅡscore(OR=1.288,95%CI 1.149~1.444,P<0.001)and Lac(OR=2.742,95%CI 1.425~5.275,P=0.003)were risk factors.The mortality risk model logit P=In(P/1-P)=2.616+0.068(age)+1.883(male)+0.253(APACHEⅡscore)+1.009(Lac)-0.506(BMI)-0.168(HCT).Receiver operating characteristic curve(ROC)analysis showed that the prediction accuracy of this model was 0.96(0.928~0.993),sensitivity was 68.5%,specificity was 89.7%,and the model had a high predicti

关 键 词：重症 肺结核 ICU 预后 危险因素 

分 类 号：R521[医药卫生—内科学]