Abstract ▼ AI-Helper

Background: Vancomycin-resistant enterococci (VRE) have become one of the most important antibiotic-resistant bacteria in the hospitalized patients. Nowadays, worldwide spread of clonal complex 17 (CC17) Enterococcus faecium is an important problem in the control of nosocomial VRE infection. The end...

Background: Vancomycin-resistant enterococci (VRE) have become one of the most important antibiotic-resistant bacteria in the hospitalized patients. Nowadays, worldwide spread of clonal complex 17 (CC17) Enterococcus faecium is an important problem in the control of nosocomial VRE infection. The endemic state of VRE in the intensive care unit (ICU) is achieved by both influx of VRE from newly admitted patients who are colonized already and cross-transmission between inpatients. To detect an influx of VRE, active surveillance culture program has been operated. Considering the time lag of 3-5 days to detect VRE, which leads to a delayed identification of VRE and early transmission to others, scoring system to predict pre-colonization of VRE is required. In addition, we intended to compare the clonality and virulence factors between imported vancomycin-resistant E. faecium (VREfm) and ICU acquired VREfm.

Methods: Between January 2008 and February 2009, active surveillance culture of VRE was carried out at the time of admission and weekly in the ICU patients in Korea University Guro Hospital. Medical records were reviewed retrospectively and patients who transferred from general wards were excluded in this study. Risk factors for VRE pre-colonization were found out and a scoring system which predicts the possibility of VRE pre-colonization at admission to ICU was designed. The clonality between the VREfm pre-colonized on admission and VREfm acquired in the ICU from January 2008 to December 2008 was comparatively analyzed. Pulsed-field gel electrophoresis (PFGE) typing of VREfm isolates was performed and multi-locus sequence typing (MLST) was conducted for the representative isolates with different PFGE type. Additionally, genes for virulence factors (esp, hyl) of VREfm were investigated by the polymerase chain reaction (PCR).

Results: High Charlson score (OR 1.71, 95% CI 1.23-2.38) and use of cephalosporin within 3 months (OR 6.89, 95% CI 1.46-32.48) were significant risk factors for pre-colonization of VRE among patients admitted to ICU. The equation of scoring was as follows: -3.841 + 0.539 x Charlson score + 1.930 x cephalosporin use within recent 3 months. Given the cut off value as -1.955, sensitivity of this scoring system was 94.1% and specificity was 72.5%. Among 13 VREfm isolates of pre-colonized group and 14 VREfm isolates of ICU acquired group, all isolates were vanA and esp positive. Hyl was positive at 76.9% of pre-colonized group and 100% of ICU acquired group (p=.098). PFGE showed diverse clonal patterns in each group; total eighteen clonal types were found and five same clonal types were found at both groups. VREfm isolates of 2 types identified at pre-colonized group were found later in the ICU acquired group. MLST was taken for the 10 representative isolates. According to the result of MLST, 8 isolates were related to CC17; sequence type (ST) 192 and ST262 were found in the pre-colonized group while ST192 and ST78 were found in the ICU acquired group. New allelic profiles were indentified and assigned as a new ST, ST471 and 472; these belonged to CC17 also.

Conclusion: High Charlson score and use of cephalosporin within 3 months were predictive factors for the pre-colonization of VRE. VRE transmission was mainly through the horizontal spread and CC17, worldwide hospital-adapted clones were found in the ICU of Korea University Guro Hospital. Pre-emptive isolation of high risk patients with VRE pre-colonization based on the scoring system and molecular diagnostic tools for the early detection of CC17 would helpful to control influx and spread of VRE in ICU.

Keyword