ICU持续炎症-免疫抑制-分解代谢综合征风险早期危险因素分析及预测  

Early risk factor analysis and prediction of the risk of persistent inflammatory-immunosuppressive-catabolic syndrome in the ICU

作　　者：袁成 邓晰明[1] 徐阳 牛凯旋 YUAN Cheng;DENG Ximing;XU Yang;NIU Kaixuan(Department of Intensive Care Unit,the First Afiliated Hospital of Bengbu Medical College,Bengbu,Anhui,233000,China)

机构地区：[1]蚌埠医学院第一附属医院重症医学科,安徽蚌埠233000

出　　处：《临床急诊杂志》2023年第12期631-636,共6页Journal of Clinical Emergency

摘　　要：目的:探讨重症医学科(intensive care unit,ICU)滞留患者发生持续炎症-免疫-代谢综合征(persistent inflammatory-immunosuppression-catabolic syndrome,PICS)的早期危险因素并构建预测模型。方法:回顾性分析2021年7月-2022年7月蚌埠医学院第一附属医院ICU收治的住院时间≥14 d患者(共计242例)的临床资料。根据是否发生PICS分为PICS组(107例)和非PICS组(135例)进行组间比较。再按照4:1比例,随机将242例患者分为模型建立组(194例)和模型验证组(48例),通过分析模型建立组的临床数据指标,构建PICS的早期预测模型。结果:纳入分析的242例患者中,PICS发生率为44.21%。对模型建立组数据进行单因素分析显示:年龄、手术、血小板计数、总胆红素、肌酐、C-反应蛋白、前白蛋白、高密度脂蛋白与PICS的发生有关(P<0.05)。筛选患者入院初期即可获得的危险因素并经二元logistic回归分析显示年龄、手术、肌酐、总胆红素、血小板为PICS发生的独立危险因素(P<0.05),依据早期数据构建ICU滞留患者发生PICS的列线图预测模型,并对预测模型进行评估,模型建立组与验证组的曲线下面积分别为:0.756(95%CI:0.588~0.824)、0.780(95%CI:0.650~0.910)。结论:年龄、手术、肌酐、总胆红素和血小板计数是急危重症患者早期即可获得的危险因素指标,基于此构建的PICS风险预测模型有着良好的区分度及校准度,可以有效评估急危重症患者PICS发病风险,为重症患者的治疗提供参考意见。Objective To explore the early risk factors of persistent inflammatory-immunosuppression-catabolic syndrome(PICS)in stranded patients in intensive care unit(ICU)and construct a prediction model.Methods Clinical data of 242 patients(in total)admitted to the ICU of a large tertiary Class A hospital from July 1,2021 to July 31,2022 was retrospectively analyzed.They were divided into PICS group and non-PICS groups according to whether PICS had occurred.According to the ratio of 4:1,242 patients were randomly divided into model building group(n=194)and model verification group(n=48).By analyzing the clinical data indicators of the model building group,construct the early prediction model of PICS was constructed.Results The 242 patients included in the analysis were divided into the PICS group(107 patients)and in the non-PICS group(135 patients),and the incidence of PICS was 44.21%.Univariate analysis to the data of the model building group showed that age,surgery,platelet,total bilirubin,creatinine,C-reactive protein,pro-albumin,and high-density lipoprotein were related with the occurrence of PICS(P<0.05).Screen those risk factors that could be obtained at early admission period.Binary Logistic regression analysis showed that,age,surgery,creatinine,total bilirubin,and platelets were independent risk factors for PICS(P<0.05).Constructed the nomogram prediction model of PICS in ICU retention patients based on early data,and evaluated the prediction model.AUC of the model building group and the model verification group were 0.756(95%CI:0.588-0.824),0.780(95%CI:0.650-0.910),respectively.Conclusion Age,surgery,creatinine value,total bilirubin and platelet are risk factors that can be obtained in the early stage in patients with critical diseases.The PICS risk prediction model constructed based on this has good differentiation and calibration,which can effectively assess the risk of PICS in acute and critical patients and provide reference for the treatment of severe patients.

关 键 词：重症医学科 持续炎症免疫抑制分解代谢综合征 危险因素 列线图模型 

分 类 号：R459.7[医药卫生—急诊医学]