Bottom line: Primary prophylaxis is recommended for patients with cirrhosis and upper gastrointestinal hemorrhage.
Dynamed cites the recommendations of the International Ascites Club (J Hepatol. 2000 Jan;32(1):142)
Recommendation for prophylaxis in patients with cirrhosis and GI bleeding based mortality benefit demonstrated by at least one RCT
For nonbleeding patients with cirrhosis and ascites without history of SBP, prophylaxis not recommended if ascitic fluid protein > 10 g/L . These recommendations report no consensus on the need for prophylaxis if ascitic fluid protein is less than 10 g/L.
A systematic review (Cochrane Database Syst Rev 2009 Apr 15;(2):CD004791) of 9 small RCTs of antibiotic prophylaxis in patients with cirrhosis but without GI bleeding, showed mortality benefit, but due to poor methodology and lack of significant benefit in any of the individual studies, authors recommend against prophylaxis in this population.
For summary and critique of published evidence on prophylaxis in nonbleeding patients with low ascitic fluid, see the Prevention and Screening section of Bacterial Peritonitis in Dynamed.
Repost of January 27, 2011