We would like to thank Thalheimer and Burroughs for their interest in our review on cardiovascular complications in cirrhosis. We agree that bacterial infections and the systemic inflammatory response in cirrhosis have variably been a topic of interest throughout the last decades. A considerable number of studies have dealt with the impact of, in particular, Gram-negative bacteria and endotoxins on the development of systemic circulatory, renal and portosystemic complications.^{1 2} However, the successes in identifying and correlating the infectious and inflammatory agents with the relevant organ dysfunction have been somewhat unpredictable.^3-5 In two conditions, however, a direct relationship between bacterial infection and increased mortality have been proved, namely infections in relation to bleeding from oesophageal varices and spontaneous bacterial peritonitis.^6-8 In both of these complications of cirrhosis and portal hypertension, treatment with antibiotics has significantly ameliorated the course of the disease and the prognosis, and there is little doubt that . . .