Summary

The first hospital outbreak of a vancomycin-resistant enterococcus (VRE) in Western Australia (WA) started in the Royal Perth Hospital in July 2001 and initially involved the Intensive Care Unit (ICU) and the Nephrology and Dialysis Units. The outbreak was caused by vancomycin-resistant Enterococcus faecium (VREF) of the vanB genotype. Pulsed-field gel electrophoresis and plasmid analysis of the isolates demonstrated a single-strain outbreak. Despite the isolation of carriers and implementation of all the additional precautions recommended to control VRE, VREF spread rapidly. Two months after the index patient was detected, the epidemic strain had spread to 22 wards and units and one outpatient unit (Satellite Dialysis). Four patients were infected and 64 were colonized.

A Hospital VRE Executive Group, which included the Chief Executive and Directors of Clinical Services and Nursing, was formed to eradicate the outbreak and to prevent the epidemic strain from becoming endemic in the hospital. The WA Department of Health agreed to provide substantial extra funding to enable the hospital to use expensive enhanced infection control practices, as follows.

1. Cohorting of all positive and ward contact patients in separate wards with dedicated nursing staff for each cohort.

2. All inpatients (ward contacts without known exposure) were screened during a one-week period to identify the total reservoir of VREF carriage. Thirty-nine previously unknown carriers were found.

3. Establishment of a dedicated VRE ward-cleaning service for the duration of the outbreak.

4. Environmental cultures after terminal cleaning and disinfection to check the efficacy of terminal cleaning after carriers had been discharged.

5. Electronic flagging of the medical records of VREF carriers and ward contacts.

6. Screening of ward contacts after discharge from hospital.

7. Special arrangements for the discharge of geriatric VREF carriers and ward contacts to residential care facilities.

Control was handicapped by the slowness of conventional laboratory methods, which took four to five days to identify VRE and allowed environmental contamination and nosocomial transmission to occur before carriers were detected and isolated. A laboratory procedure to make rapid provisional identification of VRE within 30–48 h was developed by performing multiplex polymerase chain reaction (PCR) for vanA and vanB genes directly on 24-h selective enrichment broth cultures.

On average, four rectal swabs, each collected on separate days, were needed to detect >90% of carriers.

In total, 1977 ward contacts were screened after discharge from hospital and 54 (2.73%) were found to be carrying VREF. The electronic labelling and active follow-up of ward contacts resulted in a significant number of carriers being detected who otherwise posed a risk of initiating further outbreaks in hospital if they were re-admitted.

The outbreak was terminated after five months and the cost of the enhanced infection control practices was AUD$ 2 700 000 (£1 000 000).

Ongoing control has been facilitated by targeted active surveillance cultures: on admission to high-risk units (ICU, Burns, Nephrology, Haematology, Bone Marrow Transplant Unit), on transfer out of the ICU to other hospital units, by monthly screening of patients regularly attending Dialysis Units, and by opportunistic laboratory screening of inpatient faecal specimens submitted for Clostridium difficile culture and toxin.

Vigilance needs to be maintained as the epidemic strain of VREF remains in the Perth community. Ward contacts of the first outbreak have caused small outbreaks in two hospitals, and seven to 19 sporadic new carriers have been detected annually since the first outbreak.

The key elements of the VRE control programme are as follows:

1. Targeted active surveillance cultures of patients on high-risk hospital units.

2. Rapid provisional laboratory identification of VRE by multiplex PCR directly on 24-h selective enrichment broth cultures.

3. Contact isolation of known carriers and unscreened ward contacts of outbreaks.

4. Electronic flagging of medical records of known carriers and unscreened ward contacts of outbreaks.

5. Eradication of single-strain (epidemic strain) outbreaks in hospitals by enhanced infection control practices.

6. Surveillance typing of VRE isolates from all new carriers in WA to identify epidemic strains and trace their epidemiology.

To date, this programme has prevented VRE from becoming established in any WA hospital.