decision of independent monitoring committee:Risk of death at 3 months higher in Albumin group than in control group, without reaching significance level.

Patients with cirrhosis present an increased susceptibility to bacterial infections. Spontaneous bacterial peritonitis (SBP) is the most frequent infection and induces severe circulatory dysfunction associated with renal failure in about 30% of cases. Renal failure is a reliable surrogate marker of in-hospital mortality in patients with SBP or with non-SBP infections. Albumin, as an adjuvant to antibiotherapy reduces significantly the rate of renal failure, in-hospital mortality, and overall mortality (Sort P, et al. NEJM 1999). However, little is known regarding the effect of albumin administration in patients with non-SBP infections. Two recent prospective studies demonstrated that non-SBP infections are associated with impairment of the effective circulating volume and precipitate renal failure whatever the presence of ascites.

The aim of this randomized clinical trial is to evaluate the effects of albumin, associated with appropriate antibiotic therapy, on occurrence or deterioration of renal failure and survival in septic (SIRS criteria required) cirrhotic patients with non-SBP infections and presenting with a Child-Pugh score > 8.

• Cirrhosis defined by clinical, laboratory or ultrasonographic findings
• Child-Pugh > 8
• Sepsis defined by the presence of proved or suspected infection with two of the four SIRS (systemic inflammatory response syndrome) criteria (Wong F, et al. Sepsis in cirrhosis: report on the 7th meeting of the international ascites club. Gut 2005)
• Creatinine < 160 µmol/L

Inclusion Criteria:

• Cirrhosis defined by clinical, laboratory or ultrasonographic findings
• Child-Pugh > 8
• Sepsis defined by the presence of proved or suspected infection with two of the four SIRS (systemic inflammatory response syndrome) criteria (Wong F, et al. Sepsis in cirrhosis: report on the 7th meeting of the international ascites club. Gut 2005)
• Creatinine < 160 µmol/L
• Written informed consent
• Absence of the exclusion criteria

Exclusion Criteria:

• Spontaneous bacterial peritonitis
• Difficult to treat infections such as : endocarditis, septic arthritis, osteomyelitis
• Heart insufficiency (YHA III-IV)
• Digestive bleeding during the week preceding the study
• Septic shock
• Hepatocellular carcinoma : stage D
• Use of antibiotics during the week preceding the study, except noroxin used for long-term antibioprophylaxy
• Diseases which can influence the short term survival