Singh and colleagues [1] reported that trimethoprim-sulfamethoxazole was effective for preventing spontaneous peritonitis in patients with cirrhosis. They also argued that this drug has a lower cost and fewer associations with selection of gram-positive bacteria than norfloxacin. Although I agree with their arguments, their results may lead to a different conclusion regarding the efficacy of prophylactic trimethoprim-sulfamethoxazole. Their results appear impressive; 8 of 30 controls and 1 of 30 patients receiving prophylaxis developed spontaneous peritonitis or spontaneous bacteremia. However, two controls had enterococcal peritonitis, and one had Klebsiella pneumoniae bacteremia, conditions that one would not expect to be prevented by trimethoprim-sulfamethoxazole. If these three episodes are excluded from the analysis, the difference between the two groups is no longer statistically significant, whether or not the patients are kept in the analysis (P = 0.19 and 0.09, respectively; Fisher exact test). Assuming that the Klebsiella strain was sensitive to trimethoprim-sulfamethoxazole, the difference between the two groups remains nonsignificant if the two episodes of enterococcal infection are excluded (P = 0.10); however, the difference is marginally significant if these 2 patients are excluded from the analysis (P = 0.048). This lack of statistically significant differences may be the result of the study's relatively small sample size; however, until further studies confirm Singh and colleagues' findings, one must be careful when recommending long-term chemoprophylaxis with trimethoprim-sulfamethoxazole to patients with cirrhosis.

It would be interesting to know the results of an analysis of only those patients who had previously had spontaneous bacterial peritonitis. Long-term chemoprophylaxis may be more clearly beneficial for such patients.