Control of methicillin-resistant Staphylococcus aureus infections in a university hospital: critical analysis of results.

The aim of this study was to assess the effectiveness of the control progra m for MRSA in our hospital. Our control strategy involved screening for car riers and widespread nasal decontamination. The incidence of MRSA was 0.99 cases per 100 patients and 1.4 cases per 1,000 hospital days in 1993 (506 c ases) and 0.65 cases per 100 patients and 1.02 cases per 1,000 hospital day s in 1997 (345 cases). Thus there was a reduction of 34.3% in the incidence per patient and of 21.7% in the incidence per day between 1993 and 1997 On e group of wards, in which each patient underwent nasal screening on admiss ion, reported a 30.5% decrease in the number of new cases between 1993 and 1997 Other wards, in which no such systematic screening was undertaken, rep orted a 55.2% increase in MRSA cases over the same period. The risk of acqu iring MRSA for uninfected patients in both wards (the exposure coefficient) was calculated based partly on the mean length of stay and partly on the n umber of cases of skin colonization or wound infections. The mean number of new cases was highest in the wards with the highest exposure coefficients. Thus any universal MRSA control program should involve nasal and wound scr eening of patients on admission. (C) 1999 Elsevier, Paris.