Risk factors, outcome and impact of empirical antimicrobial treatment in extended-spectrum β-lactamase-producing Escherichia coli bacteraemia

Abstract

Background: To investigate patient characteristics and empirical antimicrobial treatment of extended-spectrum β-lactamase-producing Escherichia coli (ESBL-EC) bacteraemia, to determine risk factors, outcome and impact of empirical antimicrobial treatment.

Methods: We performed a retrospective case-control study of all patients diagnosed with ESBL-EC from January 2011 to September 2012. The control group consisted of patients with non-ESBL E. coli bacteraemia. The groups were compared with respect to empirical treatment, risk factors and outcome, using univariate and multivariate analysis.

Results: The study consisted of 70 consecutive cases of ESBL-producing and 140 controls of non-ESBL-producing E. coli bacteraemia. ESBL-EC prevalence of bloodstream invasive E. coli isolates was 6.1%. The independent risk factor found for ESBL-EC bacteraemia was a prior culture with ESBL production (p < 0.001). A higher frequency of inappropriate empirical antibiotic treatment (p < 0.001) and a trend towards worse outcome was observed in patients infected with ESBL-EC and empirical guidelines were more often not followed (p = 0.013). If the guidelines were followed this was associated with adequate initial antibiotic treatment (p < 0.001).

Conclusions: Patients with ESBL-EC frequently received inappropriate empirical treatment and guidelines were more often not followed. A prior culture of ESBL-producing bacteria was an independent predictor and risk factor for ESBL-EC bacteraemia. Since the prevalence of ESBL-producing E. coli is increasing the importance of adequate guidelines must be emphasized.

Keywords: ESBL; Escherichia coli; antibiotic resistance; bacteraemia/bloodstream infection; empirical antibiotic treatment.