Assessment of the comparability of CLSI, EUCAST and Stokes antimicrobial susceptibility profiles for Escherichia coli uropathogenic isolates

Abstract

Background: As many clinical laboratories convert between Stokes, Clinical and Laboratory Standards Institute (CLSI) and European Committee for Antimicrobial Susceptibility Testing (EUCAST) methods, the problem of comparing differently derived sets of antimicrobial susceptibility testing (AST) data with each other arises, owing to a scarcity of knowledge of inter-method comparability. The purpose of the current study was to determine the comparability of CLSI, EUCAST and Stokes AST methods for determining susceptibility of uropathogenic Escherichia coli to ampicillin, amoxicillin-clavulanate, trimethoprim, cephradine/cephalexin, ciprofloxacin and nitrofurantoin.

Methods: A total of 100 E. coli isolates were obtained from boric acid urine samples from patients attending GP surgeries. For EUCAST and CLSI, the Kirby-Bauer disc diffusion method was used and results interpreted using the respective breakpoint guidelines. For the Stokes method, direct susceptibility testing was performed on the urine samples.

Results: The lowest levels of agreement were for amoxicillin-clavulanate (60%) and ciprofloxacin (89%) between the three AST methods, when using 2017 interpretive guidelines for CLSI and EUCAST. A comparison of EUCAST and CLSI without Stokes showed 82% agreement for amoxicillin-clavulanate and 94% agreement for ciprofloxacin. Discrepancies were compounded by varying breakpoint susceptibility guidelines issued during the period 2011–2017, and through the inclusion of a definition of intermediate susceptibility in some cases.

Conclusions: Our data indicate that the discrepancies generated through using different AST methods and different interpretive guidelines may result in confusion and inaccuracy when prescribing treatment for urinary tract infection.

Keywords:

• EUCAST
• CLSI
• stokes
• antimicrobial susceptibility testing
• comparability assessment
• Escherichia coli

Acknowledgements

The authors also wish to acknowledge their colleagues in the Clinical Microbiology Department of Cork University Hospital for their support of this work, and particularly that of Mr. Eddie McCullagh, Senior Medical Scientist.