RESUMEN
BACKGROUND: The current increase in nosocomial infections caused by vancomycin-resistant enterococci (VRE) warrants improvement of detection methods and hygiene measures. Knowledge of the local epidemiology is important for monitoring compliance of medical personnel with hygiene measures. AIM: To evaluate semi-automated repetitive element palindromic polymerase chain reaction (rep-PCR) for rapid molecular typing of VRE. METHODS: Primary VRE isolates were collected during an observation period of one year and retrospectively typed by rep-PCR. Molecular typing was performed on isolates from two departments with elevated VRE rates and patients with increased risk for systemic VRE infections. Typing results were correlated with temporal and spatial information on patient moves, VRE laboratory results and multi-locus sequence typing (MLST). FINDINGS: Approximately 70% of VRE isolates within a department could be assigned to similarity clusters. Spread of VRE was limited to the individual departments. There was no evidence for spread of endemic VRE strains within the geographical catchment area of the hospital. Our results demonstrate the utility of rep-PCR typing on a department level. However, a Diversilab® threshold of ≥98% had to be applied to claim similarity, and suspected transmissions needed to be confirmed by vanA/B genotyping and compiled information on spatial and temporal patient contact. MLST verified the findings. CONCLUSION: Spread of predominantly detected vancomycin-resistant Enterococcus faecium was limited to the department level with no evidence for wider dissemination within the hospital. Well-standardized and validated (semi-)automated rep-PCR systems are useful for rapid detection of possible VRE transmission. However, suspected transmissions need to be confirmed by clinical and microbiological parameters.