Variceal bleeding in portal hypertension: bacterial infection and comparison of efficacy of intravenous and per-oral application of antibiotics--a randomized trial.

OBJECTIVE: To determine the prevalence of bacterial infection in patients admitted to hospital with variceal bleeding in comparison with patients with liver cirrhosis admitted because of another reason. To compare the effect of orally administered antibiotics vs. intravenous antibiotics.
METHODS: Bacteriological investigation of blood culture, urine, throat smear, perianal smear and ascites (polymorphonuclear count as well in ascites) was made in 46 cirrhotic patients admitted to hospital with variceal bleeding and 48 cirrhotic patients admitted because of another reason. Bleeders were treated endoscopically (sclerotization) and pharmacologically (terlipressin 1 mg every 4 h for 5 days), and were randomly allocated to the treatment with oral norfloxacin (25 patients) or intravenous ampicillin/sulbactam (21 patients). Early and late mortalities were evaluated.
RESULTS: The incidence of infection was high in both groups (63.0% bleeders vs. 54.2% controls), but bleeding patients more often had positive blood culture (17.3% vs. 8.6%) and statistically significantly more positive findings in the throat smears (36.9% vs. 17.3%, P=0.04), which gives the evidence of increased pathological colonization in these patients. No difference in survival was seen in patients with per-oral or intravenous administration of antibiotics.
CONCLUSION: Bacterial infection was demonstrated in high percentage in patients with liver cirrhosis admitted to hospital. The administration of antibiotics is indicated in these patients. Intravenous application is probably of the same efficacy as per-oral one.

RCT Entities:   Population Interventions Outcomes