Mupirocin for controlling methicillin-resistant Staphylococcus aureus: lessons from a decade of use at a university hospital.

BACKGROUND: From 1990 to 1995 at Hospital Universitário Clementino Fraga Filho, patients colonized or infected with methicillin-resistant Staphylococcus aureus (MRSA) were treated with mupirocin to eliminate MRSA carriage. In 1995, 65% of MRSA patients at this hospital had mupirocin-resistant isolates. Starting in 1996, mupirocin use was restricted to patients colonized, but not infected, with MRSA. OBJECTIVES: To describe the use of mupirocin for controlling MRSA over a decade and to analyze the molecular epidemiology of mupirocin-resistant MRSA infections at this hospital. SETTING: A 490-bed, tertiary-care university hospital. METHODS: The incidence densities of patients with MRSA and acquisition of mupirocin by the hospital were calculated for the period 1992-2001. S. aureus isolates from 1999-2000 were analyzed by pulsed-field gel electrophoresis. Mupirocin-resistant MRSA isolates from 1994-1995 and 1999-2000 were analyzed for ileS-2 gene background polymorphisms. RESULTS: The incidence density of MRSA patients increased slightly over time, whereas the purchase of mupirocin decreased dramatically. Mupirocin-resistant MRSA infections decreased from 65% in 1994-1995 to 15% in 1999-2000. The MRSA Brazilian clone, detected in 1992, was still highly prevalent. The same ileS-2 encoding plasmid found in 1994-1995 persisted in three identical MRSA isolates from 1999-2000 belonging to the Brazilian clone. CONCLUSIONS: After mupirocin use decreased, the ileS-2 encoding plasmid persisted in only a few Brazilian clone isolates. Our data on mupirocin-resistant MRSA incidence and mupirocin use strongly suggested that restricted use was related to decreased rates of mupirocin resistance at our hospital.