Precision limits for MIC data required to set epidemiological cut-off values

Abstract

If the standard deviation of the distribution of MIC values for a set of wild-type isolates exceeds 1.18 log2 μg/mL when calculated by NRI analysis or 1.11 log2 μg/mL when calculated by ECOFFinder that data set should be considered imprecise and not be used to set an epidemiological cut-off value.