Phenotypic detection of mec A-positive staphylococcal blood stream isolates: High accuracy of simple disk diffusion tests

Detection of oxacillin-resistance in staphylococci, by phenotypic methods remains problematic. Although standardized susceptibility test methods are adequate for Staphylococcus aureus, many are less satisfactory for the coagulase-negative staphylococci (CNS). We have studied 108 consecutive blood culture isolates of staphylococci. The mec A gene was detected by PCR in one S. aureus and 55 CNS isolates. Susceptibility testing was performed as follows: oxacillin (1-mu g), ceftizoxime (30-mu g), and cephalothin (30-mu g) by disk diffusion; oxacillin, ceftizoxime, cephalothin, methicillin, ampicillin, ampicillin/sulbactam, penicillin, cefazolin, imipenem, and meropenem by the broth microdilution method. In addition, isolates were tested by the oxacillin agar screen plate method. The single oxacillin-resistant S. aureus strain was detected by all oxacillin susceptibility test methods and by the ceftizoxime disk and MIC methods. Two oxacillin-susceptible S. aureus were intermediate (minor error) by ceftizoxime broth microdilution (MIC, 16 mu g/mL). The most sensitive, simple phenotypic methods for detection of oxacillin-resistant CNS (mec A positive) were as follows: oxacillin disk diffusion at 98%, oxacillin screen plate at 91%, oxacillin broth microdilution at 87%, ceftizoxime disk diffusion at 100%, ceftizoxime broth microdilution at 87%, and methicillin broth microdilution at 83%. These results indicate that oxacillin and ceftizoxime disk diffusion tests are the most accurate phenotypic methods in routine clinical use for detection of oxacillin-resistant CNS. Oxacillin broth microdilution MIC testing (2% NaCl supplement) would perform more satisfactorily (100% sensitivity) with an adjusted interpretive breakpoint at less than or equal to 0.5 mu g/mL, in contrast to the lower accuracy of the ''so-called'' reference agar screen test.