Background & aims: Antibiotic prophylaxis is a cornerstone in the management of acute variceal bleeding. However, emergence of multiresistant bacteria and antibiotic-associated complications is a growing problem in cirrhosis. It has been proposed that certain low-risk populations may have good outcomes without antibiotic. We aimed to analyse the stratified risk of bacterial infection after a variceal bleeding in previously considered low-risk patients.
Methods: We analysed prospective data from all consecutive cirrhotic patients with acute variceal bleeding admitted to our tertiary hospital between 2004 and 2012. All patients received somatostatin, variceal ligation and antibiotic prophylaxis. Patients were followed until day 42 or death. Patients were stratified based on Child-Pugh class. Within the low-risk strata, further one-step stratification was performed using baseline risk factors identified through logistic regression multivariate adjustment.
Results: Two hundred and fifteen patients were included. Twenty-seven patients (12.5%) developed 32 bacterial infections within 6 weeks after the index bleeding. Multivariate adjustment identified alcohol consumption as a significant risk factor for infection. Within previously considered low-risk patients (Child-Pugh A), the risk of infection was significantly higher in patients with active alcohol consumption (21.4% vs. 0% in non-drinkers, P = 0.015). The risk of infection in Child-Pugh A and B patients with non-alcohol cirrhosis receiving antibiotics was very low (2/81, 2.5%).
Conclusions: Alcohol consumption significantly increased the risk of infection after a variceal bleeding in previously considered low-risk patients. Within Child-Pugh A class, patients with active alcohol consumption should not be considered at low risk of infection.
Keywords: alcohol-antibiotic resistance; cirrhosis; risk stratification; variceal bleeding.
© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.