肝硬化合并消化道出血影响因素及Fbg、D-D、IL-6相关预测模型构建

The influencing factors of cirrhosis combined with gastrointestinal hemorrhage and construction of prediction models associated with Fbg,D-D and IL-6

目的探讨肝硬化合并上消化道出血(UGB)影响因素及纤维蛋白原(Fbg)、D-二聚体(D-D)、白细胞介素-6(IL-6)在预测出血风险中的价值。方法选取2021年4月至2023年5月在本院接受治疗的82例肝硬化合并UGB患者为出血组,104例单纯肝硬化患者为非出血组。采用Kendall's tau-b法分析血清Fbg、D-D、IL-6与肝硬化患者临床特征的相关性;采用Logistic回归分析影响肝硬化患者并发UGB的危险因素;采用ROC曲线分析Fbg、D-D、IL-6三项联合及联合预测指数预测肝硬化患者并发UGB的AUC值、敏感度、特异度。结果出血组与非出血组在饮酒史、Child-Pugh分级、食管静脉曲张程度、门静脉内径、脾脏厚度、白蛋白、腹水、胃底静脉曲张等资料比较中,差异具有统计学意义(P<0.05)。出血组血清Fbg低于非出血组,而D-D、IL-6高于非出血组(P<0.05)。Kendall's tau-b相关性分析显示,Fbg与Child-Pugh分级呈负相关,食管静脉曲张程度与D-D、IL-6呈正相关(P<0.05)。Logistic回归分析表明,饮酒史、脾脏厚度>40 mm、胃底静脉曲张、Fbg含量下降及D-D、IL-6水平升高是影响肝硬化患者并发UGB的危险因素(P<0.05)。ROC曲线分析显示,Fbg、D-D、IL-6三项联合及列线图预测模型预测肝硬化患者并发UGB的AUC值分别为0.728、0.745、0.732、0.916、0.948,均对肝硬化患者并发UGB有一定预测价值,列线图预测模型更高。当设定cut-off值时,列线图预测模型敏感度和特异度分别为86.60%、86.50%。Bootstrap内部验证(B=1000)显示,Bias-corrected曲线与Ideal曲线基本重合,表明预测模型效能较好且稳定。结论饮酒史、脾脏厚度>40 mm、胃底静脉曲张、Fbg含量下降及D-D、IL-6水平升高是影响肝硬化患者并发UGB的危险因素。基于Fbg、D-D、IL-6水平以及其他危险因素构建的风险列线图预测模型对UGB有良好且稳定的预测价值。

Objective To investigate the influence factors of liver cirrhosis combined with upper gastrointestinal hemorrhage(UGB)and the predictive value of fibrinogen(Fbg),D-dimer(D-D),interleukin-6(IL-6)on bleeding risk.Methods A total of 82 patients with cirrhosis combined with UGB treated in our hospital from April 2021 to May 2023 were selected as the bleeding group,while 104 patients with simple cirrhosis without UGB as the non-bleeding group.Kendall's tau-b method was used to analyze the correlation between serum Fbg,D-D,IL-6 and clinical features of patients with cirrhosis.Logistic regression was used to analyze the risk factors of UGB in patients with cirrhosis.Receiver operating characteristic(ROC)curve was used to analyze the area under the curve(AUC)value,sensitivity and specificity of Fbg,D-D and IL-6 and combined prediction index to predict UGB in patients with cirrhosis.Results There were statistically significant differences in drinking history,Child-Pugh grade,esophageal varices degree,portal vein diameter,spleen thickness,albumin,ascites and gastric fundus varices(P<0.05).Serum Fbg in the bleeding group was lower than that in the non-bleeding group,while D-D and IL-6 were higher than that in the non-bleeding group(P<0.05).Kendall's tau-b correlation analysis showed that Fbg was negatively correlated with Child-Pugh grade,and the degree of esophageal varices was positively correlated with D-D and IL-6(P<0.05).Logistic regression analysis showed that drinking history,spleen thickness>40 mm,gastric varices,decreased Fbg content and increased D-D and IL-6 levels were risk factors for UGB in patients with liver cirrhosis(P<0.05).ROC curve analysis showed that the AUC values of Fbg,D-D,IL-6,the combination of the three and the nomogram prediction model for predicting UGB in patients with liver cirrhosis were 0.728,0.745,0.732,0.916 and 0.948,respectively,which all had certain predictive value for UGB in patients with liver cirrhosis,but the nomogram prediction model was higher.When the cut-off value was taken,the sensitivity and specificity of the nomogram prediction model were 86.60%and 86.50%,respectively.Bootstrap internal validation(B=1000)showed that the Bias-corrected curve basically coincided with the Ideal curve,suggesting that the prediction efficiency of the model was good and relatively stable.Conclusion Drinking history,spleen thickness>40 mm,gastric varices,decreased Fbg content and increased D-D and IL-6 levels are the risk factors of UGB in patients with liver cirrhosis,indicating that the identified factors and the nomogram prediction model could be valuable tools in clinical practice for assessing the risk of UGB in cirrhotic patients.