Clinical outcomes of spontaneous bacterial peritonitis due to extended-spectrum beta-lactamase-producing Escherichia coli or Klebsiella pneumoniae: a retrospective cohort study.

Abstract

The aim of this study was to (1) evaluate the clinical outcomes of spontaneous bacterial peritonitis (SBP) due to extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli or Klebsiella pneumoniae (EK) and (2) investigate the relationship between the adequacy of initial antibiotic treatments and patient outcomes. We conducted a retrospective cohort study of cirrhotic patients with SBP caused by EK. We evaluated the 30-day mortality rate and used Cox proportional hazard models to identify risk factors for mortality. Between January 2006 and December 2012, a total of 231 episodes of SBP due to EK were recorded. Among them, 52 were caused by ESBL-producing EK (ESBL-EK). The 30-day mortality rate was significantly higher in patients with SBP due to ESBL-EK than in those with non-ESBL-producing EK (non-ESBL-EK) (34.6 vs. 18.4%, respectively; p=0.013). Multivariate analysis revealed that ESBL production [adjusted HR (aHR) 1.82, 95% confidence interval (CI) 1.00-3.31], nosocomial infection (aHR 2.24, 95% CI 1.26-3.95), septic shock (aHR 4.84, 95% CI 2.70-8.65), higher Child-Pugh score (aHR 1.57, 95% CI 1.28-1.92), and higher Charlson comorbidity index (aHR 1.37, 95% CI 1.15-1.64) were independent risk factors for 30-day mortality in the total cohort. When we analyzed patients with SBP due to ESBL-EK separately, septic shock (aHR 3.64, 95% CI 1.40-9.77), accompanying bacteremia (aHR 3.71, 95% CI 1.37-10.08), and hepatocellular carcinoma (aHR 3.21, 95% CI 1.20-8.56) were independent risk factors. Both 7- and 30-day mortalities for SBP due to ESBL-EK were significantly higher than for SBP due to non-ESBL-EK. Initial antibiotic choice was not associated with poor clinical outcomes in patients with SBP due to ESBL-EK.