BackgroundThe objective of this study is to analyze the factors that are associated with the adequacy of empirical antibiotic therapy and its impact in mortality in a large cohort of patients with extended-spectrum β-lactamase (ESBL) - producing Escherichia coli and Klebsiella spp. bacteremia.MethodsCases of ESBL producing Enterobacteriaceae (ESBL-E) bacteremia collected from 2003 through 2008 in 19 hospitals in Spain. Statistical analysis was performed using multivariate logistic regression.ResultsWe analyzed 387 cases ESBL-E bloodstream infections. The main sources of bacteremia were urinary tract (55.3%), biliary tract (12.7%), intra-abdominal (8.8%) and unknown origin (9.6%). Among all the 387 episodes, E. coli was isolated from blood cultures in 343 and in 45.71% the ESBL-E was multidrug resistant. Empirical antibiotic treatment was adequate in 48.8% of the cases and the in hospital mortality was 20.9%. In a multivariate analysis adequacy was a risk factor for death [adjusted OR (95% CI): 0.39 (0.31-0.97); P = 0.04], but not in patients without severe sepsis or shock. The class of antibiotic used empirically was not associated with prognosis in adequately treated patients.ConclusionESBL-E bacteremia has a relatively high mortality that is partly related with a low adequacy of empirical antibiotic treatment. In selected subgroups the relevance of the adequacy of empirical therapy is limited.