[Risk factors of gastrointestinal polypectomy concurrent with bleeding in patients with liver cirrhosis]

Objective: To investigate and analyze the occurrence and the related risk factors of gastrointestinal polypectomy accompanied by bleeding in patients with liver cirrhosis. Methods: 127 cases of gastrointestinal polyps with cirrhosis who had endoscopy at the Endoscopic Center of Tianjin Third Central Hospital between November 2017 and November 2020 were collected. At the same time, 127 cases of gastrointestinal polyps with non-cirrhosis that were treated by endoscopy were collected for comparison. The occurrence of hemorrhagic complications between the two groups was compared. The effects of age, sex, liver function, peripheral blood leukocytes, hemoglobin, platelets, blood glucose, the international normalized ratio (INR), polyp resection method, polyp location, size, number, endoscopic morphology, pathology, the presence or absence of diabetes, portal vein thrombosis, and esophageal varices on polypectomy bleeding in the cirrhosis group were analyzed. The measurement data between groups were compared using the t-test and rank sum test. The χ (2) test or Fisher's exact probability method, and multivariate logistic regression analysis were used for the comparison of categorical data between groups. Results: The number of polypectomy bleeding cases in the cirrhotic group was 21, with a bleeding rate of 16.5%. The number of bleeding cases in the non-cirrhotic group was 3, with a bleeding rate of 2.4%. The bleeding rate was higher in the cirrhosis group when polypectomy was performed (χ (2) = 14.909, P < 0.001). A univariate analysis of the risk factors for gastrointestinal polypectomy associated with bleeding in patients with liver cirrhosis showed that liver function grading, platelets, INR, hemoglobin, degree of esophageal and gastric varices, and the location, shape, size, and pathology of the polyps had a statistically significant impact on bleeding (P < 0.05). Multivariate logistic regression analysis showed that liver function grade, degree of varicose veins, and polyp location were independent risk factors for bleeding. Patients with Child-Pugh B or C grade liver function were more likely to bleed than those with Child-Pugh A grade (OR = 4.102, 95% CI 1.133 ~ 14.856), gastric polyps were more likely to bleed than colorectal polyps (OR = 27.763, 95% CI 5.567 ~ 138.460), and severe esophagogastric varices were more likely to bleed than no varices or mild to moderate varices (OR = 7.183, 95% CI 1.384 ~ 37.275). Conclusion: Cirrhotic population has higher risk of bleeding during endoscopic gastrointestinal polypectomy than the non-cirrhotic population. Cirrhotic patients with Child-Pugh grades B or C liver function, polyps located in the stomach, severe esophagogastric varices, and other high-risk factors should be listed as a relative contraindication for endoscopic polypectomy.