Hospital-acquired infection with vancomycin-resistant Enterococcus faecium transmitted by electronic thermometers.

To describe an epidemic of vancomycin-resistant Enterococcus faecium causing bacteremia and bacteriuria, to identify the source of infection, to delineate risk factors associated with acquisition of the organism, and to determine antibiotic sensitivities for the organism. Investigation of an epidemic, including a case-control study. Medical-surgical intensive care unit and ward in a university medical center. Nine patients infected or colonized with vancomycin-resistant Enterococcus faecium and 20 noninfected controls. Clinical data, environmental surveillance cultures, and in-vitro microbiologic studies. Colonization or infection by vancomycin-resistant E. faecium was associated with an increased duration of treatment with ceftazidime, 13.2 compared with 4.6 days, and a greater number of nonisolated days of hospitalization in the intensive care unit, 19.9 compared with 6.4 days for infected and noninfected patients, respectively (P less than 0.05). Environmental surveillance cultures recovered the organism repeatedly from the rectal probe handles of three electronic thermometers used exclusively on nonisolated patients in the intensive care unit. Restriction endonuclease analysis of plasmid DNA showed that all clinical and environmental isolates were identical. Infection control measures, including isolation of colonized or infected patients and removal of the rectal thermometer probes suspected to be responsible for transmission, resulted in termination of the outbreak. In-vitro, time-kill studies showed that the combination of ciprofloxacin, rifampin, and gentamicin resulted in bactericidal activity against the organism. This nosocomial outbreak of infection due to a highly vancomycin-resistant strain of Enterococcus is the first epidemic in which an electronic thermometer has been implicated as the vehicle of transmission for an infectious agent.