Clinical characteristics and outcomes of patients with Escherichia coli in airway samples.

PURPOSES: Escherichia coli is one of the most common pathogens in nosocomial and community-acquired infections, but is an uncommon respiratory pathogen. However, this pathogen may at times be seen in respiratory secretions. The study aims to determine the clinical and prognostic value of E. coli in respiratory secretions. METHODS: Cultures of respiratory secretions from hospitalized and outpatients between 2009 and 2016 were screened for isolation of E. coli. We defined three groups of patients: "Sensitive (S)"-growth of E. coli sensitive to all antimicrobials tested; Intermediate (I)-resistant to 1-2 antimicrobial classes; and "Resistant (R)"-resistant to at least three antibiotic classes. We compared factors associated with resistant strains and outcomes between the groups. RESULTS: Eighty patients with E. coli isolates from respiratory secretions were identified while screening 177 712 (4.5: 10 000 samples). Of the E. Coli-positive cultures, 11 were from ambulatory patients, 31 patients were hospitalized and 37 were hospitalized and intubated. Ten people had bronchiectasis and 29 had COPD. Patients with resistant E. coli had significantly more hospitalization days prior to positive culture (S = 1.2 ± 1.89 days, I = 1.23 ± 1.5 days and R = 3.7 ± 5.4 days, respectively; P = 0.002). Mortality was higher in patients with a resistant strain (R) versus (I) or (S) (76.7%, 31.8% and 26.7%, respectively; P < 0.0001) and remained significantly elevated after correction for prior hospital days. CONCLUSIONS: Pulmonary infection due to E. coli is uncommon. Isolation of resistant E. coli is associated with length of previous hospitalization, elevated mortality and may be viewed as a nosocomial pathogen.

The association between treatment appropriateness according to EUCAST and CLSI breakpoints and mortality among patients with candidemia: a retrospective observational study.

To evaluate the association between appropriate antifungal treatment and mortality among patients with candidemia using different breakpoint definitions. In a retrospective study, we included all adults with candidemia in a tertiary center between 2009 and 2015. We defined three versions of appropriate (covering) antifungal treatment, according to Clinical and Laboratory Standards Institute (CLSI) 2008, CLSI 2012, and European Committee on Antimicrobial Susceptibility Testing (EUCAST) (2017 update) breakpoints. For empiric treatment, we evaluated the association with 30-day mortality. For definitive treatment, we evaluated the association with 90-day mortality among patients surviving the first week after candidemia onset. Adjusted odds ratios (OR) from a bivariate logistic regression with 95% confidence intervals are reported. We identified 302 patients with 308 separate candidemia episodes. The crude 30-day mortality was 55% (168/308). Resistance to anidulafungin increased from 3.5 to 51.6% and to fluconazole from 15.2 to 44.1%, when applying CLSI 2008 and EUCAST definitions, respectively. Appropriate empirical treatment was significantly associated with lower 30-day mortality using the CLSI 2008 definitions, adjusted OR 0.56 (0.33-0.96). The associations were similar, though not statistically significant for EUCAST, 0.58 (0.33-1.00), and CLSI 2012, OR 0.62 (0.37-1.04). Appropriate definitive treatment according to CLSI 2012 and EUCAST was independently associated with lower 90-day mortality, ORs 0.31 (0.13-0.75) and 0.44 (0.23-0.8), respectively. With CLSI 2008, the association was similar but not statistically significant, OR 0.4 (0.11-1.41), with few isolates classified as resistant. Considering the major shift in resistance prevalence when applying CLSI 2008, CLSI 2012, and EUCAST breakpoint definitions, no major differences were observed in their association with mortality.