Johan W Mouton1, Joseph Meletiadis1,2, Andreas Voss3,4, John Turnidge5. 1. Department of Medical Microbiology and Infectious Diseases, Erasmus MC, Rotterdam, The Netherlands. 2. Clinical Microbiology Laboratory, Attikon University Hospital, Medical School, National and Kapodistrian University of Athens, Athens, Greece. 3. Department of Medical Microbiology, Radboud University Medical Centre, Nijmegen, The Netherlands. 4. Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Hospital, Nijmegen, The Netherlands. 5. Adelaide Medical School, University of Adelaide, Adelaide, Australia.
Abstract
Objectives: Although testing of antimicrobial agents for susceptibility has inherent variability like any assay, it is generally held that there are also real differences in susceptibility between strains. In the routine laboratory, variability of the MIC measurement may be sufficient to mask real strain differences. We determined which factors contributed to the variability, using linezolid against Staphylococcus aureus as one example. Methods: Twenty-five S. aureus strains were sent to five different laboratories in quadruplicate in a blinded fashion. Laboratories determined MICs of linezolid using Etest. Results of 22 strains corresponding to 440 observations were available for analysis. Sources of variability were explored and quantified using an ANOVA approach. Results: The overall geometric mean MIC was 1.8 mg/L, comparable to that of the published WT distribution of 1.7 mg/L (www.eucast.org). The total variation amounted to ∼1.3 2-fold dilutions for a one-sided CI of 95% and two 2-fold dilutions for a CI of 99%. Variation between laboratories and variation between strains contributed 10% and 48%, and in a subset analysis averaging 17% and 26%, respectively. Strain-to-strain variation (biological variation) could not be reliably determined, even with four replicates. Conclusions: This analysis serves as an example of an approach to discerning various sources of MIC variation. Here, at best, a single measurement of an MIC may provide an indication of whether it likely belongs to the WT distribution. Only repeated measurements of MICs for individual strains within one laboratory may provide an indication of differences in susceptibility between strains.
Objectives: Although testing of antimicrobial agents for susceptibility has inherent variability like any assay, it is generally held that there are also real differences in susceptibility between strains. In the routine laboratory, variability of the MIC measurement may be sufficient to mask real strain differences. We determined which factors contributed to the variability, using linezolid against Staphylococcus aureus as one example. Methods: Twenty-five S. aureus strains were sent to five different laboratories in quadruplicate in a blinded fashion. Laboratories determined MICs of linezolid using Etest. Results of 22 strains corresponding to 440 observations were available for analysis. Sources of variability were explored and quantified using an ANOVA approach. Results: The overall geometric mean MIC was 1.8 mg/L, comparable to that of the published WT distribution of 1.7 mg/L (www.eucast.org). The total variation amounted to ∼1.3 2-fold dilutions for a one-sided CI of 95% and two 2-fold dilutions for a CI of 99%. Variation between laboratories and variation between strains contributed 10% and 48%, and in a subset analysis averaging 17% and 26%, respectively. Strain-to-strain variation (biological variation) could not be reliably determined, even with four replicates. Conclusions: This analysis serves as an example of an approach to discerning various sources of MIC variation. Here, at best, a single measurement of an MIC may provide an indication of whether it likely belongs to the WT distribution. Only repeated measurements of MICs for individual strains within one laboratory may provide an indication of differences in susceptibility between strains.
Authors: Robin J Svensson; Katarina Niward; Lina Davies Forsman; Judith Bruchfeld; Jakob Paues; Erik Eliasson; Thomas Schön; Ulrika S H Simonsson Journal: Br J Clin Pharmacol Date: 2019-07-25 Impact factor: 4.335
Authors: Laurynas Mockeliunas; Lina Keutzer; Marieke G G Sturkenboom; Mathieu S Bolhuis; Lotte M G Hulskotte; Onno W Akkerman; Ulrika S H Simonsson Journal: Pharmaceutics Date: 2022-03-30 Impact factor: 6.525