Establishment of a nomogram model to predict sepsis in patients with multiple trauma
白丽爽 1王兴义 1杨立山1
扫码查看
点击上方二维码区域，可以放大扫码查看
作者信息
1. 宁夏医科大学总医院急诊科,银川 750000
折叠
摘要
目的 探讨预测多发伤患者并发脓毒血症的危险因素,并构建列线图预测模型.方法 回顾性收集2021年1月至2022年4月入住宁夏医科大学总医院急诊重症监护病房(emergency intensive care unit,EICU)的多发伤患者的资料.入人标准:(1)符合多发伤的诊断标准;(2)伤后到入院时间≤24 h;(3)年龄＞18岁;(4)所有检查或抢救措施均获得患者或患者家属同意;(5)患者临床资料完整.以EICU住院28 d为节点,根据Sepsis 3.0定义将患者分为脓毒血症组和非脓毒血症组.绘制受试者工作特征曲线,采用Logistic回归模型中的影响因素预测脓毒血症的发生,并构建列线图.结果 共纳入291例患者,其中脓毒血症组102例,非脓毒血症组189例.多因素Logistic分析发现年龄、入院24 h内首次的急性生理与慢性健康评分(acute physiology and chronic health evaluation,APACHE)-Ⅱ、格拉斯哥昏迷评分(Glasgow Coma Scale,GCS)、创伤严重度评分(injury severity score,ISS)、序贯器官功能衰竭评分(sequential organ failure assessment,SOFA)、输血次数、是否应用去甲肾上腺素药物、是否接受机械通气治疗、病原学培养结果和糖尿病史是发生脓毒症的独立影响因素.结合以上变量构建列线图模型(AUC=0.913,95％CI:0.847～0.942),模型具有良好的拟合校准曲线.结论 年龄、APACHE-Ⅱ评分、GCS评分、SOFA评分、ISS评分、输血次数、机械通气、去甲肾上腺素药物使用、病原学培养、糖尿病构建出的列线图对多发伤患者后期是否并发脓毒血症具有较好的预测价值,值得推广.
Abstract
Objective To explore the risk factors of sepsis in patients with multiple trauma and construct a nomogram prediction model.Methods The data of patients with multiple injuries admitted to the emergency intensive care unit(EICU)of the General Hospital of Ningxia Medical University from January 2021 to April 2022 were respectively collected.Inclusion criteria:(1)meet the diagnostic criteria for multiple injuries;(2)the time from injury to admission≤24 hours;(3)age＞18 years old;(4)all examination or rescue measures were approved by the patient or the patient's family;(5)the patient's clinical data were complete.The patients were divided into sepsis group and non-sepsis group according to the definition of Sepsis 3.0 at the 28-day of EICU hospitalization.The receiver operating characteristic curve was drawn.Logistic regression analysis was applied to determine the independent predictors for sepsis,and the nomogram was constructed.Results A total of 291 patients were included,including 102 in the sepsis group and 189 in the non-sepsis group.Multivariate logistic analysis revealed that age,acute physiology and chronic health status score(APACHE)Ⅱ,Glasgow Coma Scale(GCS),injury severity score(ISS),sequential organ failure assessment(SOFA)within 24 hours after admission,blood transfusion frequency,the application of norepinephrine,mechanical ventilation,pathogenic culture results,and history of diabetes were independent factors influencing the occurrence of sepsis.A nomogram model was constructed by combining these variables(AUC=0.913,95％CI:0.847-0.942),and the model had a good fitting calibration curve.Conclusions The nomogram constructed by age,APACHE-Ⅱ,GCS score,SOFA score,ISS score,number of blood transfusions,mechanical ventilation,norepinephrine drug use,pathogenic culture and diabetes has a good predictive value for sepsis in patients with multiple trauma in the later stage,which is worth promoting.
关键词
多发伤/脓毒血症/列线图/危险因素/急诊重症监护室
Key words
Multiple trauma/Sepsis/Nomogram/Risk factors/Emergency intensive care unit
维普
万方数据