Purpose: Community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) strains are classically characterised by susceptibility to most non-β-lactam antimicrobial agents. We sought to determine whether antimicrobial susceptibility (AMS)-based algorithms could be used to presumptively identify CA-MRSA in a hospital MRSA collection.
Methods: Over a three-month period, all MRSA were tested for AMS, staphylococcal cassette chromosome mec (SCCmec) type, presence of the Panton-Valentine leukocidin (PVL) genes and spa type. CA-MRSA isolates were defined genotypically using a combination of spa and SCCmec type. AMS based algorithms were developed and tested for their sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV).
Results: Ciprofloxacin susceptibility (p < 0.001) and fusidic acid resistance (p = 0.044) were independent predictors of CA-MRSA in a multivariate model. Although 98.5% of HA-MRSA were ciprofloxacin resistant, so too were 36.6% of CA-MRSA. Algorithms based on ciprofloxacin-susceptibility and fusidic acid resistance performed best, with specificity and NPV >90% and sensitivity and PPV >70%.
Conclusions: Our data indicate that while ciprofloxacin-susceptible isolates are likely to be CA-MRSA, the use of ciprofloxacin-susceptibility as a marker of CA-MRSA would miss approximately one third of CA-MRSA isolates. Therefore, AMS patterns have limited utility for the identification of genetically-defined CA-MRSA in our setting.