[Analysis of risk factors for bleeding as a complication of ultrasound-guided percutaneous liver biopsy examination].

Objective: To explore the independent risk factors for bleeding in patients following percutaneous liver biopsy examination. Methods: The clinicopathological data of patients who underwent percutaneous liver biopsy examination at Nanjing Second Hospital from January 2012 to December 2021 were retrospectively collected. Univariate and multivariate logistic regression analysis were used to investigate the effect of age, gender, lesion type (diffuse liver parenchymal lesions, focal liver lesions), number of biopsies, tissue length, presence or absence of cirrhosis, presence or absence of portosystemic shunt, erythrocytes, white blood cells, hemoglobin, platelets, prothrombin time, fibrinogen, international normalized ratio, and liver biochemical indicators on bleeding following liver biopsy, as well as to screen independent risk factors. Results: A total of 3 331 patients were examined by percutaneous liver biopsy, and 3 060 cases were actually included by excluding 271 cases who took consultation from other hospitals. The overall postoperative hemorrhagic rate was 1.6% (49/3 060). Of which, forty-four cases (1.4%) had overt bleeding (hemodynamic changes or hemoglobin decreased by more than 20 g/L), five cases (0.2%) had minor bleeding, three cases had subcapsular hepatic hemaotma, and two cases had local bleeding from liver biopsy. Among the overt bleeding cases, two cases were in the off-label group (platelet<50×10⁹/L or international normalized ratio>1.5), and the rest were in the non-off-label group. The results of univariate analysis showed that factors such as focal liver lesions, portosystemic shunt, prolonged prothrombin time, increased international normalized ratio, bilirubin, and alkaline phosphatase were associated with bleeding after liver biopsy in the non-off-label group. The multivariate collinearity diagnosis revealed statistically significant differences for the indicators. Multivariate logistic regression analysis finally included factors such as lesion type, portosystemic shunt, international normalized ratio, total bilirubin, and alkaline phosphatase. The results showed that patients with focal liver lesions were more prone to bleed after surgery than patients with diffuse liver parenchymal lesions (OR=3.396, P=0.002, 95%CI: 1.596-7.228). Patients with portosystemic shunt were more prone to bleed than those without portosystemic shunt (OR=3.301, P=0.018, 95%CI: 1.232-8.845). Patients were more likely to experience bleeding following liver biopsy when their total bilirubin levels were elevated (OR=1.006, P<0.001, 95%CI:1.003-1.008). Conclusion: Focal liver lesions, portosystemic shunts, and elevated total bilirubin are independent risk factors for bleeding after percutaneous liver biopsy.