多发伤患者预后列线图模型的构建和研究

Construction and study of nomograph model for prognosis of multiple trauma patients

目的探讨多发伤患者预后危险因素,建立列线图预测模型。方法收集291例入住本院急诊重症监护病房(emergency intensive care unit,EICU)的多发伤患者的临床资料,包括性别、年龄、是否开放伤、是否应用去甲肾上腺素药物、是否机械通气、伤后到医院的时间、到医院的距离、淋巴细胞相对值、血小板计数、乳酸、入院24 h内的创伤严重度评分(injury severity score,ISS)、急性生理与慢性健康评分(acute physiology and chronic health evaluationⅡ,APACHEⅡ)、格拉斯哥昏迷评分(glasgow coma scale,GCS)、输血次数、手术次数、糖尿病、高血压、吸烟既往史。根据在EICU住院期间病情是否恶化为节点,分为恶化组和好转组。采用软件SPSS26.0对数据进行统计分析,单因素及多因素分析影响多发伤患者预后的因素,绘制受试者工作特征(receiver operating characteristic,ROC)曲线及森林图,二元Logistic回归模型中的影响因素制作列线图。结果是否机械通气、是否使用去甲肾上腺素药物、年龄、淋巴细胞相对值、乳酸、APACHEⅡ评分、GCS评分、手术次数对多发伤患者预后预警有意义(P<0.05),二元Logistic回归模型(Enter法)得出独立影响因素:年龄、乳酸、APACHEⅡ评分、手术次数,采用Medcalc软件进行ROC曲线分析,多因素联合预测曲线下面积最大,APACHEⅡ评分次之。每个指标诊断截断值:年龄>58岁、淋巴细胞相对值≤8.62%、乳酸>1.72、APACHEⅡ评分>16分、GCS评分≤6分、手术次数≤0,采用R软件将二元Logistic回归模型中的影响因素建立列线图,有较好的预测价值。结论年龄、淋巴细胞相对值、乳酸、APACHEⅡ评分、GCS评分、手术次数、是否机械通气、是否去甲肾上腺素药物使用构建出的列线图对多发伤预后预警具有较好的预测价值,值得推广。

Objective To explore the prognostic risk factors of patients with multiple injuries and establish a nomogram prediction model.Methods The clinical data of 291 patients with multiple injuries admitted to the Emergency Intensive Care Unit(EICU)of General Hospital of Ningxia Medical University were collected,including sex,age,open injury,norepinephrine use,mechanical ventilation,time to hospital after injury,distance to hospital,relative lymphocyte value,platelet count,lactic acid,injury severity score(ISS),acute physiology and chronic health evaluationⅡ(APACHEⅡ),Glasgow coma scale(GCS),number of blood transfusions,number of operations,and previous history of diabetes,hypertension and smoking within 24 h after admission.According to whether the condition worsened during the hospitalization of EICU,the patients were divided into the deterioration group and improvement group.SPSS26.0 software was used for statistical analysis of the data,univariate and multivariate analysis were used to screen the factors affecting the prognosis of patients with multiple injuries,receiver operating characteristic(ROC)curve and forest chart were drawn,and the influencing factors in binary Logistic regression model were used to make the nomogram.Results Mechanical ventilation,norepinephrine use,age,relative lymphocyte value,lactic acid,APACHE-II score,GCS score,and number of operations were signifi cant for predicting the prognosis of patients with multiple injuries(P<0.05).The independent influencing factors obtained by binary Logistic regression model were age,lactic acid,APACHE-Ⅱscore and number of operations.ROC curve analysis showed that the area under the curve was the largest in multi-factor combined prediction,followed by APACHE-Ⅱscore.The diagnostic cut-off value of each index was as follows:age>58 years old,relative lymphocyte value≤8.62%,lactic acid>1.72,APACHE-Ⅱscore>16,GCS score≤6,and number of operations≤0.The R software was used to establish a nomogram of the infl uencing factors in the binary Logistic regression model,which had good predictive value.Conclusions The nomogram constructed by age,relative lymphocyte value,lactic acid,APACHE-Ⅱscore,GCS score,number of operations,mechanical ventilation,and norepinephrine use has a good predictive value for the prognosis of patients with multiple injuries,and is worthy of promotion..