A 10-year profile of enterococcal bloodstream infections at a tertiary-care hospital in Japan.

OBJECTIVES: The first aim of this study is to characterize the epidemiology of enterococcal bloodstream infections (BSIs) at a Japanese tertiary-care hospital. The second aim is to identify predictive factors for 30-day mortality.

METHODS: We conducted a single center retrospective observational study. All patients with enterococcal BSI between 2005 and 2014 were enrolled. Univariate and multivariate analysis were performed to evaluate predictive factors for 30-day mortality.

RESULTS: A total of 410 patients with enterococcal BSI were enrolled. Enterococcus faecalis was identified in 200 patients (48.8%) and Enterococcus faecium in 124 patients (30.2%). Isolates were susceptible to ampicillin and vancomycin (67.3% and 97.8%, respectively). Isolates that were not susceptible to vancomycin were either Enterococcus casseliflavus or Enterococcus gallinarum. All strains of E. faecalis and 10.8% of E. faecium strains were susceptible to ampicillin. Thirty-day mortality was 24.8%. Predictive factors for 30-day mortality were Charlson Comorbidity Index (CCI) 1-2 (adjusted odds ratio [OR] 6.07, 95% confidence interval [CI]: 1.22-30.2), CCI 3-4 (adjusted OR 8.79, 95% CI: 1.70-45.4), CCI ≥5 (adjusted OR 17.6, 95% CI: 3.52-88.2), E. faecium bacteremia (adjusted OR 2.19, 95% CI: 1.27-3.76), Pitt Bacteremia Score (PBS) ≥5 (adjusted OR 15.1, 95% CI: 6.41-35.6), and source control (adjusted OR 0.39, 95% CI: 0.22-0.72).

CONCLUSION: Vancomycin-resistant strains of E. faecalis and E. faecium were not seen in this cohort. In addition, all strains of E. faecalis and 10.8% of E. faecium strains were susceptible to ampicillin. Predictive factors for 30-day mortality were CCI score, E. faecium bacteremia, PBS score, and source control.