Comparison of three methods to recover vancomycin-resistant enterococci (VRE) from perianal and environmental samples collected during a hospital outbreak of VRE

OBJECTIVE: To establish an efficient and sensitive technique for recovering vancomycin-resistant enterococci (VRE) from perianal and environmental sam ples collected during implementation of control measures for an outbreak of VRE. DESIGN: Perianal and environmental samples were collected in triplicate on sterile swabs. One swab was used to inoculate a selective broth medium cont aining 6 pg of vancomycin and 8 mug of ciprofloxacin per mt, one to inocula te Campylobacter agar containing 10 mug/mL of vancomycin, and one to inocul ate Enterococcosel agar containing 8 mug/mL of vancomycin. SETTING: Samples were collected in the intensive care units of a 600-bed un iversity hospital over a period of 2 months. SAMPLE SELECTION: Patients and their immediate environment were sampled ii they resided in a ward with a patient known to be colonized or infected wit h VRE. RESULTS: Of the 88 perianal samples obtained from 63 patients, 37 were posi tive for VRE by broth culture, with 36 also recovered on both types of soli d media (sensitivity, 97.3%; negative predictive value, 98.1%). Of the init ial samples collected from each of the 63 patients, 20 were positive for VR E by all methods. Of the 500 environmental samples cultured, 139 were posit ive for VRE in broth, with only 33 recovered on Campylobacter agar (sensiti vity, 23.7%; negative predictive value, 77.2%) and 22 on Enterococcosel aga r (sensitivity, 15.8%; negative predictive value, 75.2%). CONCLUSIONS: Our data indicate that, when performing surveillance cultures during an outbreak of VRE, use of an enrichment broth medium is required to recover VRE contaminating environmental surfaces; however, direct inoculat ion to selective solid medium is adequate to recover VRE in patient periana l specimens.