Journal of Practical Hepatology ›› 2023, Vol. 26 ›› Issue (5): 686-689.doi: 10.3969/j.issn.1672-5069.2023.05.021

• Liver cirrhosis • Previous Articles     Next Articles

Prediction of re-bleeding in patients with hepatitis B cirrhosis complicated by esophageal gastric variceal bleeding

Tu Jianjun, Fang Xiang, Jiang Qi, et al   

  1. Department of Gastroenterology, First Hospital, Affiliated to Fujian Medical University, Nanping 353000,Fujian Province, China
  • Received:2023-01-06 Online:2023-09-10 Published:2023-09-13

Abstract: Objective The aim of this study was to analyze the influencing factors of rebleeding in patients with hepatitis B liver cirrhosis complicated by esophageal gastric variceal bleeding (EGVB), and to establish a prediction model for early intervention. Methods 119 patients with hepatitis B cirrhosis and EGVB were encountered in our hospital between July 2020 and July 2022, and all patients received comprehensive internal medical treatment and even endoscopic intervention for hemostasis. The occurrence of re-bleeding within 3 months after hemostasis was recorded and statistically analyzed. The risk factors of re-bleeding were analyzed by multivariate Logistic regression analysis, and the ROC curve was applied to reveal the predictive performance of the model we established for re-bleeding. Results Out of our series, 48 patients(40.3%) had re-bleeding within three months after hemostasis, and four patients died 72 hours after re-bleeding; there were significant differences as respect to the percentage of Child-Pugh class C(97.7% vs. 84.5%), splenic thicknesses [(67.5±5.3)mm vs.(59.1±5.7)mm], diameters of portal vain [(16.8±2.1)mm vs.(14.5±2.3)mm], diameters of splenic vain [(10.9±2.3)mm vs.(8.4±1.8)mm], esophageal and gastric varices grade G3(81.8% vs. 46.5%), endoscopic red-color sign(88.6% vs. 38.0%), ascites (38.6% vs. 18.3%), serum albumin levels [(29.1±4.2)g/L vs. (33.5±4.8)g/L] and prothrombin time (PT)[(16.3±2.6)s vs. (12.7±2.1)s] between 44 patients with re-bleeding and 71 patients without (P<0.05); the Logistic regression analysis showed that the Child-Pugh class C [OR(95%CI):2.3(1.3-3.8)], splenic thickness [OR(95%CI):1.8(1.2-2.9)], diameter of portal vain [OR(95%CI):2.6(1.0-6.4)], diameter of splenic vain [OR(95%CI):1.4(1.0-1.9)], varices grade G3[OR(95%CI):2.4(1.0-5.5)], red-color sign[OR(95%CI):2.1(1.0-4.4)] and prolonged PT [OR(95%CI):1.8(1.1-2.9)] were the risk factors for re-bleeding; we built up a formula as follows: P=1/[1+e(0.814X1+0.597X2+0.939X3+0.328X4+0.863X5+0.739X6+0.592X7-1.164)] based on the Logistic analysis, and a good fitting degree between the prediction model and the standard curve was found by the Hosmer-Lemeshow test (x2=3.711, P=0.882), and its AUC was 0.966 (95%CI:0.938-0.995) with the specificity of 87.3% and the sensitivity of 97.8% in predicting re-bleeding. Conclusion The patients with liver cirrhosis and EGVB might have a high tendency of re-bleeding after emergent hemostasis, and the early prediction as well as appropriate management might decrease the mortality, and needs further investigation.

Key words: Liver cirrhosis, Esophageal gastric variceal bleeding, Re-bleeding, Influencing factors, Prediction