列线图模型预测肝硬化门静脉血栓发生的风险

Nomogram model for predicting the risk of portal vein thrombosis in liver cirrhosis

目的 分析肝硬化患者门静脉血栓(portal vein thrombosis, PVT)发生的危险因素,采用列线图模型预测肝硬化PVT发生的风险。方法 回顾性分析2018年6月至2022年10月于我院消化内科及感染病科住院治疗的208例肝硬化患者的临床资料,依据是否合并PVT,将其分为PVT组(n=85)和无PVT组(n=123)。利用SPSS 25.0软件对数据进行单因素及多因素Logistic回归分析,筛选出肝硬化PVT发生的危险因素。基于多因素分析结果应用R语言软件建立列线图模型,并通过绘制ROC曲线以及校准曲线对该列线图模型的准确性及临床实用性进行评价。结果 两组间在脾切除术、内镜治疗、白细胞计数、纤维蛋白原、D-二聚体、门静脉直径方面比较,差异有统计学意义(P<0.05)。其中门静脉直径增宽、脾切除术、内镜下食管胃底静脉套扎或硬化剂注射治疗是肝硬化患者PVT发生的独立危险因素(P<0.05)。基于Logistic回归分析结果构建的列线图模型具有较高的准确度及区分度,有一定的临床效用。结论 对于该列线图模型评分较高的患者,应注意定期监测随访,在排除出血风险后必要时给予预防性抗凝治疗,减少血栓的发生风险。

Objective To analyze the risk factors of portal vein thrombosis(PVT)in patients with liver cirrhosis,and to predict the risk of PVT in liver cirrhosis by Nomogram model.Methods The clinical data of 208 patients with liver cirrhosis who were hospitalized in the Department of Gastroenterology and Infectious Diseases of our hospital from Jun.2018 to Oct.2022 were retrospectively analyzed.According to whether they were combined with PVT,they were divided into PVT group(n=85)and non-PVT group(n=123).Univariate and multivariate Logistic regression analysis was performed on the data using SPSS 25.0 software to screen out the risk factors for PVT in liver cirrhosis.Based on the results of multivariate analysis,R language software was used to establish a Nomogram model,and the accuracy and clinical practicability of the Nomogram model were evaluated by drawing ROC curve and calibration curve.Results There were significant differences in splenectomy,endoscopic treatment,WBC,FBG,D-dimer and portal vein diameter between the two groups(P<0.05).Among them,portal vein diameter widening,splenectomy,endoscopic esophagogastric variceal ligation or sclerosing agent injection were independent risk factors for PVT in patients with liver cirrhosis(P<0.05).The Nomogram model based on the results of Logistic regression analysis had high accuracy and discrimination.It had certain clinical utility.Conclusion For patients with a high score of the Nomogram model,regular monitoring and follow-up should be paid attention to,and preventive anticoagulant therapy should be given when necessary after excluding the risk of bleeding,so as to reduce the risk of thrombosis.