OBJECTIVE

To review relevant studies for both primary and secondary antibiotic prophylaxis of spontaneous bacterial peritonitis (SBP) in patients with cirrhosis without gastrointestinal bleeding.

METHODS

A search of PubMed (1980-July 2010) was conducted using the terms prophylaxis, SBP, and antibiotics. A manual review of bibliographies was conducted for inclusion of relevant articles.

METHODS

Prospective studies and meta-analyses published in English were included.

RESULTS

Ten trials and 3 meta-analyses were included. Of the 10 trials, 2 examined the use of secondary prophylaxis for prevention of subsequent episodes of SBP, 4 examined the use of primary prophylaxis to prevent an initial SBP episode, and 4 examined the use of antibiotic prophylaxis in a mixed population. Seven trials evaluated the use of an antibiotic compared to placebo or no treatment. Only 1 trial evaluated norfloxacin versus trimethoprim/sulfamethoxazole. Trial duration varied from 24 days to 12 months. In general, trials examining norfloxacin as secondary prophylaxis found significantly decreased occurrence of SBP but no significant difference in mortality rates. Primary prophylaxis studies found no significant difference in the incidence of infections, including SBP, with norfloxacin or ciprofloxacin treatment but significantly lower incidence of gram-negative infections. Mixed population studies found a significantly decreased incidence of SBP but no significant difference in mortality. In the 3 meta-analyses, a significant decrease in mortality and an overall decrease in SBP incidence in the treatment groups were noted.

CONCLUSIONS

Based on currently available data, the use of prophylactic antibiotic therapy is warranted for the prevention of recurrent SBP in patients with cirrhosis and ascites. In patients with low ascetic fluid protein and at least 1 more risk factor, primary prophylaxis may be considered. Further studies with improved methodology are needed to determine whether prophylactic antibiotic therapy has an impact on mortality.