Rapid detection of ESBL, carbapenemases, MRSA and other important resistance phenotypes within 6-8 h by automated disc diffusion antibiotic susceptibility testing.

Abstract

In principle, automated systems allow rapid reading of disc diffusion AST (rAST) within 6-8 h. This study analysed whether rAST can discriminate resistance phenotypes such as ESBL, carbapenemases and MRSA/methicillin-resistant Staphylococcus epidermidis from WT populations. We describe species-drug combinations that may require clinical breakpoint adaptions for early reading due to zone diameter changes during the incubation period. In total, 1852 clinical strains [Escherichia coli (n = 475), Klebsiella pneumoniae (n = 375), Enterobacter cloacae (n = 301), Staphylococcus aureus (n = 407) and S. epidermidis (n = 294)] were included in this study comprising WT populations and important resistance phenotypes, e.g. ESBL, carbapenemases and MRSA. We assessed (i) separation of resistance phenotypes and WT populations after 6, 8 and 12 h as compared with the 18 h standard, and (ii) diameter changes of WT populations and associated putative epidemiological cut-offs during the incubation period. Disc diffusion plates were automatically streaked, incubated and imaged using the WASPLabTM system. We demonstrated that important resistance phenotypes could reliably be separated from WT populations at early reading times for the most prevalent bacterial pathogens encountered in the clinical laboratory. Current AST expert rules and algorithms for identification of resistance mechanisms can readily be applied for rAST, e.g. EUCAST recommended rules for detection of ESBL, AmpC, carbapenemases and MRSA/methicillin-resistant S. epidermidis. However, several species-drug combinations may require clinical breakpoint adaptations when using rAST as the diameter, and hence the epidemiological cut-off, changes during the incubation period.