Genomic evaluation of phenotypic antibiotic susceptibility patterns as a surrogate for MRSA relatedness and putative transmission during outbreak investigations.

Abstract

Antibiograms have been used during outbreak investigations for decades as a surrogate for genetic relatedness of Methicillin-resistant Staphylococcus aureus (MRSA). In this study, we evaluate the accuracy of antibiograms in detecting transmission, using genomic epidemiology as the reference standard. We analysed epidemiological and genomic data from 1,465 patients and 1,465 MRSA isolates collected at a single clinical microbiology laboratory in the United Kingdom over a one-year period. A total of 132 unique antibiograms (AB) were identified based on VITEK 2 susceptibility testing, with two profiles (AB1 and AB2) accounting for 698 isolates (48%). We identified MRSA-positive patients with a known hospital or community contact and evaluated the prediction of MRSA transmission based on identical antibiograms. The sensitivity and specificity of identical antibiograms to infer genetically related MRSA isolates (≤25 SNPs) within hospital contacts (presumed transmission events) was 66.4% and 85.5% respectively and 73.8% and 85.7% within community contacts. Reanalysis, where any single drug mismatch in susceptibility results was allowed, increased sensitivity but reduced specificity: 95.2% and 58.8%, respectively, for hospital contacts; and 91.7% and 62.6% for community contacts. Overall, the sensitivity and specificity of identical antibiograms for inferring genetically related MRSA isolates (≤25 SNPs), regardless of epidemiological links, were 49.1% and 87.5%, respectively. We conclude that using an antibiogram with one mismatch can detect most transmission events; however, its poor specificity may lead to an increased workload through the evaluation of numerous pseudo-outbreaks. This study further supports the integration of genomic epidemiology into routine practice for the detection and control of MRSA transmission.