ABSTRACT
OBJECTIVES: The aim of this study was to evaluate the accuracy of various susceptibility methods when testing cefuroxime against a collection of Escherichia coli isolates with MIC values close to the breakpoint. METHODS: 80 E. coli strains with a cefuroxime MIC value of 16 mg/L obtained by broth microdilution with Vitek 2 were selected. Microdilution was considered the reference standard and was performed in duplicate, as were disc and gradient diffusion tests using two different manufacturers in each case. EUCAST 8.0 breakpoints were used for MIC interpretation. RESULTS: All strains were resistant according to Vitek 2 (MIC 16 mg/L) but 72.5% (58/80) were classified as susceptible by reference standard microdilution. Categorical and essential agreements between Vitek 2 and reference standard microdilution were 27.5% (95% CI 1.9-1.4) and 86.3% (95% CI 0.8-0.9), respectively. Differences are statistically significant when isolates are classified as 'susceptible' or 'resistant' according to EUCAST breakpoints between diffusion methods (disc and gradient) and reference standard microdilution. Using BioMérieux (BM) and Liofilchem (LF) gradient testing, 24.1% (14/58) and 13.8% (8/58) of results were identified as false susceptible and 4.5% (1/22) and 40.9% (9/22) were found to be false resistant, respectively. Using Oxoid (OX) and Bio Rad (BR) cefuroxime discs, 22.5% (13/58) and 17.2% (10/58) of results were false susceptible and 9.1% (2/22) and 13.6% (3/22) were false resistant, respectively. DISCUSSION: Intertechnique variation around the cefuroxime breakpoint was a considerable source of disagreements and seriously affected the clinical classification of the isolates. We propose that the definition of the area of technical uncertainty (ATU) be modified to include the variability between approved AST methods.
