基于超声及实验室参数的Logistic回归模型预测肝硬化食管胃底静脉曲张破裂出血的临床价值  

Clinical value of Logistic regression model based on ultrasound and laboratory parameters in predicting esophageal and gastric variceal bleeding in liver cirrhosis

作　　者：张志远 邓卉 ZHANG Zhiyuan;DENG Hui(Department of Infectious Diseases,Xinchang Hospital Affiliated to Wenzhou Medical University,Zhejiang 312500,China)

机构地区：[1]温州医科大学附属新昌医院感染科,浙江省绍兴市312500 [2]重庆市渝北区人民医院科教科

出　　处：《临床超声医学杂志》2024年第6期491-495,共5页Journal of Clinical Ultrasound in Medicine

摘　　要：目的基于二维超声、剪切波弹性成像及实验室参数建立Logistic回归模型,探讨其预测肝硬化食管胃底静脉曲张破裂出血的临床应用价值。方法选取于我院就诊的肝硬化患者73例,依据胃镜检查结果分为低风险(G1级)组14例和高风险(G2级+G3级)组59例,均接受二维超声、剪切波弹性成像及实验室检查,比较两组上述检查结果的差异。采用Logistic回归分析筛选肝硬化食管胃底静脉曲张破裂出血的独立影响因素,并建立预测模型。绘制受试者工作特征(ROC)曲线评价模型的诊断效能。结果两组血小板计数、凝血酶原时间、胆碱酯酶、血浆白蛋白及脾动脉收缩期峰值流速、脾静脉内径、脾静脉流速、脾脏面积、脾脏弹性均值比较差异均有统计学意义(均P<0.05)。Logistic回归分析显示,血浆白蛋白、脾静脉内径、脾脏弹性均值均为肝硬化食管胃底静脉曲张破裂出血的独立影响因素(均P<0.05)。建立预测肝硬化食管胃底静脉曲张破裂出血的Logistic回归模型,公式为:Logit(P)=2.172+1.453×脾脏弹性均值+0.396×脾静脉内径-0.135×血浆白蛋白。ROC曲线分析显示其预测肝硬化食管胃底静脉曲张破裂出血的曲线下面积为0.965,灵敏度为92.40%,特异度为91.90%。结论基于二维超声、剪切波弹性成像及实验室参数建立的Logistic回归模型对预测肝硬化食管胃底静脉曲张破裂出血具有较好的临床应用价值。Objective To establish a Logistic regression model based on two-dimensional ultrasound,shear wave elastography and laboratory parameters,and to explore its clinical application value in predicting esophageal and gastric variceal bleeding in liver cirrhosis.Methods A total of 73 patients with liver cirrhosis in our hospital were selected and divided into lowrisk(G1 grade)group(14 cases)and high-risk(G2 grade+G3 grade)group(59 cases)according to the gastroscopy results.All patients received two-dimensional ultrasound,shear wave elastography and laboratory examination,and the differences of the above examination results between the two groups were compared.Logistic regression analysis was used to screen the independent influencing factors of esophageal and gastric variceal bleeding in liver cirrhosis,and a diagnositc model was established.Receiver operating characteristic(ROC)curve was drawn to evaluate the diagnostic efficacy of the model.Results There were significant differences in platelet count,prothrombin time,cholinesterase,plasma albumin,peak systolic velocity of splenic artery,splenic vein diameter,velocity of splenic vein,area of spleen,and mean spleen elasticity between the two groups(all P<0.05).Logistic regression analysis showed that the plasma albumin,splenic vein diameter and mean spleen elasticity were the independent influencing factors of esophageal and gastric varices bleeding in liver cirrhosis(all P<0.05).A Logistic regression model of for predicting esophageal and gastric variceal bleeding in liver cirrnosis was established,the formula was:Logit(P)=2.172+1.4534×mean spleen elasticity+0.396×splenic vein diameter-0.135×plasma albumin.ROC curve analysis showed that the area under the curve was 0.965,the sensitivity was 92.40%,and the specificity was 91.90%.Conclusion The Logistic regression model based on two-dimensional ultrasound,shear wave elastography and serological parameters has good clinical application value in predicting esophageal and gastric variceal bleeding in liver cirrhosis.

关 键 词：超声检查 剪切波弹性成像 食管胃底静脉曲张破裂出血 肝硬化 LOGISTIC回归 

分 类 号：R445.1[医药卫生—影像医学与核医学]