832. Profile of enterococcal bloodstream infection at a tertiary-care hospital in Japan over a 10-year period: Characteristics, outcomes, and predictors for 30-day mortality
Session: Poster Abstract Session: Bacteremia and Endocarditis
Friday, October 9, 2015
Room: Poster Hall
Posters
  • Revise1.2@ID week 2015 poster .pdf (480.2 kB)
  • Background: Enterococcusspecies, a major cause of bloodstream infection (BSI), are associated with high mortality, especially in vancomycin-resistant enterococci (VRE) cases. However, few large-scale data exist on enterococcal BSI in Japan. This study aimed to characterize the background, clinical manifestations, microbiological characteristics, and mortality of enterococcal BSI in a Japanese tertiary-care hospital, and identified predictive factors for 30-day mortality.

    Methods: In our hospital-based retrospective case analysis, patients with enterococcal BSI from 2005 to 2014 at a tertiary-care hospital with 925 beds were identified from a laboratory database. Clinical data were taken from electronic medical records. Predictive factors for 30-day mortality were evaluated using univariate and multivariate analyses. 

    Results: In all, 443 patients with enterococcal BSI were noted (median age, 75 [interquartile range 64-82]). Enterococcus faecalis was noted in 226 patients (51.0%); E. faecium in 140 (31.6%); and nosocomial BSI in 251 (56.7%). A solid tumor was the most common comorbidity (130 patients [29.3%]). Intra-abdominal infection was the most frequent source (176 patients [39.7%]), but the source was not identified in 74 (16.7%). Ampicillin and vancomycin resistance rates were 32.3% and 2.0%, respectively. VRE cases were either E. casseliflavus or E. gallinarum. The 30-day mortality rate was 22.6%. Multivariate analysis revealed predictive factors for 30-day mortality: Charlson Comorbidity Index (CCI) score 1-2 (adjusted odds ratio [OR] 5.53, 95% confidence interval [CI] 1.09-28.0); CCI 3-4 (adjusted OR 7.83, 95% CI 1.49-41.3); CCI >4 (adjusted OR 13.5, 95% CI 2.65-68.8); nosocomial bacteremia (adjusted OR 2.14, 95% CI 1.17-3.92); E. faecium bacteremia (adjusted OR 1.78, 95% CI 1.03-3.08); and Pitt Bacteremia Score [PBS] >4 (adjusted OR 17.7, 95% CI 7.61-41.1).

    Conclusion: Our cohort’s unique characteristic was the low VRE rate. Neither E. faecalis nor E. faecium was resistant to vancomycin. The mortality rates were similar to those observed in previous studies. CCI score, nosocomial bacteremia, E. faecium bacteremia, and PBS were independent predictors for 30-day mortality.

    Hiroyuki Suzuki, MD1, Naoto Hosokawa, MD, PhD1, Yoshihito Otsuka, PhD2, Ryota Hase, MD1, Daisuke Suzuki, MD1, Kazuyasu Miyoshi, MD1, Shunsuke Uno, MD1, Akihiro Amma, MD1 and Akihiko Shimizu, MD1, (1)Department of Infectious Diseases, Kameda Medical Center, Kamogawa, Chiba, Japan, (2)Department of Laboratory Medicine, Kameda Medical Center, Kamogawa, Chiba, Japan

    Disclosures:

    H. Suzuki, None

    N. Hosokawa, None

    Y. Otsuka, None

    R. Hase, None

    D. Suzuki, None

    K. Miyoshi, None

    S. Uno, None

    A. Amma, None

    A. Shimizu, None

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.