Objectives:

To determine the rate of colonisation and the incidence of transmission of methicillin resistant Staphylococcus aureus (MRSA) on an Intensive Care Unit (ICU) and control MRSA transmission.

Methods:

A nine bed ICU was studied for an initial observation period of 8 months, all patients admitted to ICU for >24 h were screened for MRSA within 48 h then three times a week. Demographic data and antibiotic usage was recorded. An intervention period of 8 months when antibiotic prescribing was restricted followed. A second observation period (6 months) was instituted. Monthly environmental screening (29 sites) with swabs from three areas in each bed space (bed floor, monitor and workstation) plus two from the nurses workstation took place. Patient and environmental isolates were typed using pulse field gel electrophoresis (PFGE).

Results

Fifty-seven of 215 patients (26%) were colonised with MRSA during the initial phase, 50% of these acquired MRSA on ICU. Eighty-eight per cent of patients received >1 antibiotic, cefuroxime and metronidazole being the most heavily prescribed primarily as prophylaxis. During the intervention period prophylactic antibiotics were restricted to one dose and the need for treatment antibiotics reviewed daily. Postintervention the total antibiotics used was reduced from 132.5 to 104.1 DDD/100 patient days and a reduction in metronidazole/cefuroxime from 33.1 to 12.0 and 20.6 to 3.8 DDD/100 patient days respectively. Despite this there was no reduction in the number of patients colonised with MRSA on ICU, but the percentage of patients acquiring MRSA on ICU fell to 43.2%. Results from 19 environmental screens of 29 sites in the ICU yielded MRSA from 1 to 11 sites on every screen (mean 4.6). On two occasions no patients on ICU were colonised with MRSA, but MRSA was isolated from two and five environmental sites respectively. Typing showed EMRSA-15 variants were predominant in patients and environment with the environmental isolates reflecting both the current patient types and previously discharged patients.

Conclusion:

Despite reducing the total antibiotic usage the rates of MRSA colonisation remained the same. The continuous isolation of MRSA from the environment together with our typing data indicates that environmental sources of MRSA have a role to play and are important in controlling the endemic MRSA infection in hospitals.