内科

目的探讨肝硬化患者发生食管胃底静脉曲张（EGV）及其破裂出血（EGVB）的影响因素，并构建风险预测模型。方法回顾性分析2016年1月至2024年1月在广西医科大学第一附属医院消化内科住院治疗的301例肝硬化患者的临床资料。先根据内镜检查结果将所有患者分为EGV组与无EGV组，再根据是否发生EGVB将EGV组患者分为EGVB组与无EGVB组。采用单因素分析及二元logistic回归模型筛选独立影响因素并构建预测模型；采用受试者操作特征（ROC）曲线下面积（AUC）评估模型及各指标的预测效能。结果男性（OR=13.052，95%CI：2.943～57.887）和有腹水（OR=6.073，95%CI：1.455～25.346）是肝硬化患者发生EGV的独立危险因素，高纤维蛋白原水平是其独立保护因素（OR=0.297，95%CI：0.117～0.757）（均P<0.05）；据此构建的EGV预测模型ROC AUC为0.857。球蛋白与血小板比值对预测食管静脉曲张价值中等（AUC=0.707），其与影像学检查联合后效能提升（AUC=0.830）；影像学检查联合球蛋白与白蛋白比值、天门冬氨酸氨基转移酶与血小板比值指数及纤维化-4指数可提高对胃静脉曲张的预测效能（AUC=0.745）。在肝硬化伴EGV患者中，有红色征是发生EGVB的独立危险因素（OR=9.582，95%CI：4.907～17.805），高纤维蛋白原水平是独立保护因素（OR=0.033，95%CI：0.001～0.918）；据此构建的EGVB预测模型AUC高达0.988。结论男性、有腹水是肝硬化患者发生EGV的独立危险因素，有红色征是肝硬化伴EGV患者发生EGVB的独立危险因素，而较高的纤维蛋白原水平对两者均有保护作用；基于上述独立影响因素构建的风险预测模型效能良好。此外影像学检查联合血清学指标策略有助于提高对EGV的非侵入性诊断价值。

Objective To investigate the influencing factors for the esophagogastric varices (EGV) and variceal bleeding (EGVB) in cirrhosis patients, and to construct risk prediction models. Methods Clinical data of 301 cirrhosis patients hospitalized in the Department of Gastroenterology, the First Affiliated Hospital of Guangxi Medical University from January 2016 to January 2024 were retrospectively analyzed. First, all patients were divided into an EGV group or a non-EGV group based on endoscopic findings; then, the EGV group was further divided into an EGVB group or a non-EGVB group based on the occurrence of EGVB. Univariate analysis and binary logistic regression model were used to screen independent influencing factors, and prediction models were constructed. The area under the receiver operating characteristic (ROC) curve (AUC) was used to evaluate the predictive performance of the models and various indicators. Results Male (OR=13.052, 95% CI: 2.943-57.887) and ascites (OR=6.073, 95% CI: 1.455-25.346) were independent risk factors for EGV in cirrhosis patients, while a high fibrinogen level was an independent protective factor (OR=0.297, 95% CI: 0.117-0.757) (all P<0.05); the prediction model for EGV constructed based on the above factors had an ROC AUC of 0.857. The globulin-to-platelet ratio had moderate predictive value for esophageal varices (AUC=0.707), and its performance improved when combined with imaging examinations (AUC=0.830); the combination of imaging examinations with the globulin-to-albumin ratio, aspartate aminotransferase-to-platelet ratio index, and fibrosis-4 index improved the predictive performance for gastric varices (AUC=0.745). Among cirrhosis patients with EGV, the red color sign was an independent risk factor for EGVB (OR=9.582, 95% CI: 4.907-17.805), while a high fibrinogen level was an independent protective factor (OR=0.033, 95% CI: 0.001-0.918); the prediction model for EGVB constructed based on the above factors reached a high AUC of 0.988. Conclusion Male and ascites are independent risk factors for EGV in cirrhosis patients, the red color sign is an independent risk factor for EGVB in cirrhosis patients with EGV, and a higher fibrinogen level has a protective effect against both outcomes; the risk prediction models constructed based on the above-mentioned independent influencing factors show good performance. The strategy of combining imaging examinations with serological indicators helps improve non-invasive diagnostic value for EGV.

ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref
ref