Spontaneous bacterial peritonitis (SBP) is a common and frequently fatal complication of end-stage liver disease with a mortality of up to 10%, primarily due to the development of kidney failure. Current standard practice is to treat this infection with broad spectrum antibiotics and salt-poor albumin administration on day one and three of treatment. In this study the investigators test the hypothesis that the administration of a second dose of albumin at 48 hours only to patients with renal insufficiency, is as effective at preventing kidney failure as administering the second dose to all patients at 72 hours.

Spontaneous bacterial peritonitis (SBP), infection of the peritoneal fluid(ascites) without evidence of a surgically treatable source, is a common and frequently fatal complication of patients with endstage hepatic cirrhosis. It originates with the passage of bacteria from the intestinal lumen to the systemic circulation and then to the ascitic fluid. Early diagnosis with paracentesis (aspiration of an ascites fluid sample to assess for evidence of infection) and the development of nonnephrotoxic third generation cephalosporin antibiotics have decreased the in hospital mortality from nearly 100% to approximately 30%. Mortality in patients with SBP is invariably associated with the development of functional renal failure. Recently, the administration of two large doses of human serum albumin at diagnosis and at 72 hours has been reported to further reduce mortality and renal failure to 10%. These findings have lead to the recommendation that patients with SBP be treated with albumin. However, no study has evaluated the necessary amount and timing of albumin administration required for its beneficial action.

In this study we test the hypothesis that the administration of a second dose of albumin at 48 hours only to patients with renal insufficiency, is as effective at preventing kidney failure as administering the second dose to all patients at 72 hours.

80 consecutive patients with cirrhosis and SBP who are at risk for renal failure will be enrolled at either the Columbia University Medical Center or The New York Hospital Weill Cornell Medical Center. Baseline clinical and biochemical data will be obtained for etiology and severity of liver disease. All patients will receive antibiotics and salt poor albumin at 1.5g/kg at time of diagnosis and diuretics discontinued (current standard of care). Patients will be randomized to receive the second dose (1.0 gm/kg) at 72 hours (group 1, standard of care) or at 48 hours only if renal function remains elevated after two days of therapy (Group 2). For the latter group of patients, albumin will be administered if the Cr is > 1.0 mg/dl or if the BUN or creatinine levels are higher than admission levels at 48 hours. If albumin is not administered at 48 hours, renal function will be monitored daily, and it will be administered should the BUN or creatinine increase to levels greater than those on admission. Renal failure rates, duration of transient azotemia, mortality, and albumin utilization rated will be compared between the groups who receive albumin in the usual manner at 72 hours versus those who receive it at 48 hours based on renal function.

Inclusion Criteria:

• Age 18 to 75
• End Stage Liver Disease / Cirrhosis
• Documented SBP (ANC > 250 or positive ascites culture)
• Ability to provide informed consent
• Serum Creatinine > 1.0 and/or Total Bilirubin > 4.0

Exclusion Criteria:

• Nonportal hypertensive ascites (i.e. malignancy)
• Hepatic Encephalopathy precluding informed consent