Prevalence and risk factors of bacteriuria in cirrhotic patients: a prospective case-control multicenter study in 244 patients.

Background/aims: The prevalence and risks factors of bacteriuria in cirrhotics have not been assessed by case-control study, and there are conflicting data concerning the role of liver failure and of ascites. The aims of this study were: i) to evaluate the prevalence of bacteriuria in cirrhotics, ii) to search for associated factors, iii) to evaluate the role of bladder post-void residual volume, and iv) to test the sensitivity of isolated bacteria to norfloxacin.
Methods: The prevalence and risk factors of bacteriuria on admission were determined by a multicenter prospective case-control study.
Results: Two hundred and forty-four cirrhotic patients and 240 controls were studied. Bacteriuria was present in 38 patients (15.6%; IC 5%: 11%-20%) and 18 controls (7.5%; IC 5%: 4.2%-11%; p<0.001). By univariate analysis, female sex and ongoing diuretic treatment were associated with bacteriuria (p<0.0001 and p<0.04, respectively). Pugh's grade, ascites and bladder residual volume were not associated with bacteriuria. By multivariate analysis, female sex (p<0.0001) and Child-Pugh score (p<0.03) were predictors of bacteriuria. Sensitivity of bacteria to norfloxacin was observed in 94.7%; sterile urine cultures were noted in 95.2% of patients treated with this antibiotic.
Conclusion: Bacteriuria is twice as frequent in cirrhotic patients as in matched controls, and there is a trend to association with female sex and liver insufficiency.