INTRODUCTION: The genus Staphylococcus is of great importance because of its high prevalence in hospital infections and because it presents high rates of resistance to oxacillin and other antimicrobials. Thus, evaluation of the accuracy of the phenotypic methods that are used to determine the profile of antimicrobial resistance is essential to ensure that the most appropriate therapy is chosen. METHODS: One hundred and fourteen strains of Staphylococcus sp (53 S. aureus and 61 CNS) were used to evaluate the accuracy of the methods of disk diffusion, agar microdilution, oxacillin screening agar and automated systems, in comparison with PCR for investigating resistance to oxacillin. RESULTS: The mecA gene was detected in 48 strains (42.1%), and 27 strains (23.7%) showed discrepant results in at least one of the methods (74.1% of CNS, 25.9% of S. aureus). For S. aureus, with the exception of the Microscan Walkaway, all the methods showed 100% specificity and sensitivity. In relation to CNS, the automated system and cefoxitin disk had lower accuracy. CONCLUSIONS: Use of two methods should be the best option for improved accuracy, especially when the diagnostic laboratory only uses an automated system or oxacillin disk diffusion test. Combination of these methods with others presented almost 100% sensitivity and specificity in our study.
Staphylococcus aureus; Oxacillin; Cefoxitin; Coagulase-negative Staphylococcus