Risk of Bacterial Infection in Patients with Cirrhosis and Acute Variceal Hemorrhage, Based on Child-Pugh Class, and Effects of Antibiotics.
Clin Gastroenterol Hepatol. 2014 Nov 21;
Authors: Tandon P, Abraldes JG, Keough A, Bastiampillai R, Jayakumar S, Carbonneau M, Wong E, Kao D, Bain VG, Ma M
Abstract
BACKGROUND & AIMS: Antibiotics are frequently overused and associated with serious adverse events in patients with cirrhosis. However, these drugs are recommended for all patients presenting with acute variceal hemorrhage (AVH). We investigated whether patients should be stratified for antibiotic prophylaxis based on Child-Pugh scores, to estimate risks of bacterial infection, rebleeding, and mortality, and whether antibiotics have equal effects on patients of all Child Pugh classes. We performed sensitivity analysis using model for end-stage liver disease (MELD) scores.
METHODS: In a retrospective study, we analyzed data from 381 adult patients with cirrhosis and AVH (70% male; mean age, 56 years), admitted from 2000 through 2009 to 2 tertiary care hospitals in Edmonton, Alberta. We excluded patients with bacterial infection on the day of AVH. The association between antibiotic prophylaxis and outcomes was adjusted by liver disease severity and by a propensity score.
RESULTS: The patients included in the study had mean MELD scores of 16, and 54% received antibiotic prophylaxis. Overall, antibiotic therapy was associated with lower risks of infection (adjusted odds ratio, 0.37; 95% confidence interval, 0.91-0.74) and mortality (adjusted odds ratio, 0.63; 95% confidence interval, 0.31-1.29). Among patients of Child-Pugh class A given antibiotics, only 2% developed infections and mortality was 0.4%. Among patients of Child-Pugh class B given antibiotics, 6% developed infections, compared to 14% of patients who did not receive antibiotics; antibiotics did not affect mortality. Administration of antibiotics to patients of Child-Pugh class C reduced infections and mortality by ∼50%, compared to patients who did not receive antibiotics. The MELD was not as useful as Child-Pugh class in identifying patients at risk for infection.
CONCLUSIONS: Based on a retrospective analysis of patients with cirrhosis and AVH, those of Child-Pugh class A have lower rates of bacterial infection and lower mortality, in the absence of antibiotic prophylaxis, than patients of classes B or C. The recommendation for routine antibiotic prophylaxis for this subgroup requires further evaluation.
PMID: 25460564 [PubMed - as supplied by publisher]