Methicillin-susceptible/methicillin-resistant Staphylococcus aureus (MSSA/MRSA) can be directly identified from positive blood culture bottles using molecular methods. This provides faster results than traditional phenotypic testing, but discrepancies between the two are occasionally found. We sought to determine the incidence and clinical impact of such discrepancies.
Positive blood culture bottles are routinely tested in the hospital clinical laboratory for mecA via Xpert MRSA/SA BC (PCR), and antimicrobial susceptibility testing (AST) via MicroScan PC33 is performed on recovered S. aureus isolates; discrepancies between PCR and AST are resolved by repeat and supplemental (Kirby-Bauer) testing. A retrospective review of medical and laboratory data from 1/2015-12/2017 was performed on all patients that had discordant PCR and AST results.
Approximately 1,200 PCR assays were performed from 1/2015-12/2017, and there were 5 (0.4%) cases with discordant AST results. 4 cases were classified as MSSA by PCR but MRSA by AST, and 1 case was classified as MRSA by PCR but MSSA by AST. For the former group, antimicrobial therapy was changed in 2 patients to cover MRSA and 1 patient was readmitted, while the remaining 2 patients were already being treated for MRSA; for the latter case, this patient was treated for MRSA during the initial hospitalization, but was readmitted with disseminated MSSA and subsequently deceased. Based on genetic targets identified by PCR and cefoxitin and oxacillin AST, discrepancies were likely due to borderline oxacillin resistance (BORSA) (n=1), presence of an SCCmec variant not detected by PCR (n=1), or undetermined (n=3).
Rapid identification of MRSA bacteremia via PCR provides actionable information to direct empiric treatment. While highly accurate, PCR results are infrequently not corroborated by AST. This rare possibility should be considered when modifying therapy based on initial PCR results, and there should be close communication between the clinical team and laboratory for these challenging cases.
D. Chen, None