Background:Coagulase-negative staphylococci (CNS) has been considered as a major causative agent of nosocomial infections. A prompt and accurate detection of methicillin resistance (MR) in staphylococci is a current issue of clinical microbiology laboratories. This study was purposed to evaluate various methods for detecting MR from CNS.

Methods:We selected 78 CNS strains obtained from blood cultures from April 1999 through July 2001 including 20 strains of Staphylococcus epidermidis, 20 S. hominis (SHO), 19 S. capitis, 9 S. haemolyticus, 3 S. saccharolyticus, 1 S. saprophyticus (SAP), 2 S. warneri (SWA), 2 S. xylosus, 1 S. lugdunensis, and 1 S. auricularis. In addition, one SAP strain received from World Health Organization for proficiency tests was also studied. The following methods were compared to the mecA gene PCR: MicroScan PosCombo 12, oxacillin salt agar containing 6 ㎍/mL (OSA-6) or 0.6 ㎍/mL (OSA0.6) of oxacillin, oxacillin disk diffusion (ODD), and MRSA-Screen latex agglutination (LA) for detecting penicillin binding protein 2a.

Results:One SWA was failed in mecA-PCR and fifty-nine of 78 (75.6%) CNS were positive for mecA gene. The agreement rates, sensitivities, and specificities for each test were as follows: for MicroScan, 97.3%, 98.2%, 88.9%; for OSA-6 and OSA-0.6 at 24-h incubation, 79.5%, 74.6%, 94.7% and 79.5%, 72.9%, 100%, respectively, and at 48-h incubation, 91.0%, 91.5%, 89.5% and 91.0%, 96.6%, 73.7%, respectively; ODD, 84.6%, 84.7%, 84.2%; LA, 80.8%, 76.3%, 94.7%. One SHO and one SAP that were mecA-negative showed resistance in the MicroScan, ODD, and OSA.

Conclusions:MicroScan appears a reliable method to detect MR in all species of CNS except SHO and SAP. ODD and LA were not appropriate in detecting MRCNS due to a low sensitivity. Although OSA-0.6 at 48-h incubation showed a high sensitivity, the low specificity may limit a routine use in clinical laboratory. (Korean J Clin Microbiol 2002;5:110-118)