Internal Medicine Journal 2004; 34: 510–512 BRIEF COMMUNICATION Rate of nosocomial transmission of vancomycin-resistant enterococci from isolated patients A. C. CHENG, 1,4,5 G. HARRINGTON, 2 P. RUSSO, 2,3 L. LIOLIOS 1 and D. SPELMAN 1,2 1 Infectious Diseases and Microbiology Unit and 2 Infection Control and Hospital Epidemiology Unit, Alfred Hospital and 3 VICNISS Hospital Acquired Infection Surveillance Coordinating Centre, Melbourne, Victoria and 4 Menzies School of Health Research, Charles Darwin University and 5 Northern Territory Clinical School, Flinders University, Darwin, Northern Territory, Australia Abstract To evaluate an isolation policy for patients colonised with vancomycin-resistant enterococci (VRE), we insti- tuted active surveillance for transmission to uncolonised patients. Surveillance rectal swabs were taken and pulsed-field gel electrophoresis was performed on posi- tive isolates. VRE transmission with an identical geno- type occurred in 5 patients, giving a transmission rate of 3.7 per 1000 patient days, or 1 patient per ward each week. The present study provides a baseline for assessment of VRE transmission and will be useful in evaluation of the effectiveness of infection control inter- ventions. (Intern Med J 2004; 34: 510–512) Key words: infection control, cross infection, vancomycin resistance, Enterococcus, disease transmission The past 20 years has seen the emergence of vanco- mycin-resistant Enterococcus faecalis and Enterococcus faecium (VRE) as nosocomial pathogens in hospitals throughout the world. In Australia, VRE is still regarded as an emerging pathogen, with a low hospital (1.6%) and community prevalence (0.2%) of colonisation. 1,2 Despite clinical infection with VRE remaining uncommon, the ease at which the organism may spread and its propensity for chronic colonisation of the gastrointestinal tract poses a significant challenge for infection control policies. Infection control policies have generally been evalu- ated by the comparison of rates before and after an intervention. Although active surveillance has been used to assess the adequacy of infection control measures for VRE 3,4 a transmission rate has not previously been described in acute care centres. In the present study, a surveillance system to detect transmission of VRE from colonised patients to non-colonised patients was established. The Alfred Hospital is a 300-bed teaching hospital and tertiary referral centre in Melbourne, Australia. VRE was first isolated in April, 1998. The present study was designed to define a baseline from which to evaluate future changes in infection control policies in general medical and surgical wards (excluding critical care areas and outpatient clinics) with an average of 30 beds on each ward. During the present study, our isolation policy varied slightly from that as previously reported. 2 Patients were stratified on their perceived risk of spreading VRE: high dispersers were defined as those patients with diarrhoea and/or incontinence, whereas low dispersers, forming the majority of patients with VRE, were those without diarrhoea or incontinence. A patient was regarded as being colonised with VRE if they had the organism isolated from rectal swabs or clinical sites at any time in the past. High dispersers were managed in a single room with a dedicated toilet. The door was closed with a sign detailing contact isolation procedures. A dedicated nurse was assigned to this patient for all shifts, with all visitors to the room wearing disposable gloves. Low dispersers were nursed in a single room with contact isolation as for high dispersers, but the nurse caring for this patient would also nurse other patients depending on the acuity of the patient’s condition. Infection control practitioners undertook a hospital-wide education campaign supple- mented by regular newsletters to promote this policy. We followed the infection control guidelines set out in the 1995 Disease Control and Prevention (CDC) Guidelines. 5 Following the isolation of VRE from a patient (index patient), the ward was notified and contact isolation procedures were instituted as above. At this time, rectal swabs were taken on all patients to determine if trans- mission or colonisation had already occurred. At weekly intervals while the index patient remained on the ward, 15 patients selected by random computer- generated list had rectal swabs performed (surveillance swabs). We collected data regarding the length of time Correspondence to: Allen C. Cheng, Melioid Lab, Sapprasithiprasong Hospital, Sapprasit Road, Muang, Ubon Ratchathani, 34000, Thailand. Email: allenc@menzies.edu.au Received 7 January 2003; accepted 20 October 2003. Funding: None Conflicts of interest: None