The critical influence of the intermediate category on interpretation errors in revised EUCAST and CLSI antimicrobial susceptibility testing guidelines

Abstract

Erroneous assignments of clinical isolates to the interpretative categories susceptible, intermediate and resistant can deprive a patient of successful antimicrobial therapy. The rate of major errors (ME) and very major errors (vME) is dependent on: (i) the precision/standard deviation (σ) of the antibiotic susceptibility testing (AST) method, (ii) the diameter distributions, (iii) clinical breakpoints, and (iv) the width of the intermediate zone. The European Committee on AST (EUCAST) has abandoned or decreased the intermediate zone for several drug/species combinations. This study focused on the effects of discontinuing the intermediate category on the rate of interpretation errors. In total, 10,341 non-duplicate clinical isolates were included in the study. For susceptibility testing the disc diffusion method was used. Error probabilities were calculated separately for diameter values flanking the interpretative category borders. Error probabilities were then applied to the actual numbers of clinical isolates investigated and expected rates of ME and vME were calculated. Applying EUCAST AST guidelines, significant rates of ME/vME were demonstrated for all drug/species combinations without an intermediate range. Virtually all ME/vME expected were eliminated in CLSI guidelines that retained an intermediate zone. If wild-type and resistant isolates are not clearly separated in susceptibility distributions, the retaining of an intermediate zone will decrease the number of ME and vME. An intermediate zone of 2-3 mm avoids almost all ME/vME for most species/drug combinations depending on diameter distributions. Laboratories should know their epidemiology settings to be able to detect problems of individual species/drug/clinical breakpoint combinations and take measures to improve precision of diameter measurements.