Comperative Evaluation of Vitek 2 and Etest Methods with the Referance Broth Microdilution for Antimicrobial Susceptibility Testing of Colistin Among Multi-Drug Resistant Gram-Negative Bacteria

Abstract

Introduction: Despite the increased need for colistin, especially in serious infections caused by carbapenem resistant gram-negative bacteria, problems and challenges regarding colistin susceptibility testing remain. The aim of this study was to evaluate the performance of Vitek 2, one of the commonly used automated systems, and Etest for colistin susceptibility testing compared with reference broth microdilution method (BMD). Materials and Methods: This study included 657 multi-drug resistance (MDR) Gram negative bacteria obtained from clinical samples; Negative control, Escherichia coli ATCC 25922 and Positive control, Escherichia coli NCTC 13846. The collected MDR isolates were performed colistin BMD according to ISO standard 20776-1, prospectively. Categorical agreement (CA), Very Major Error (VME), and Major Error (ME) rate were calculated. Acceptable performance was evaluated as; CA >= 90%; VME <1.5% and ME <3%. Results: Colistin resistance rates were detected by Vitek 2, Etest and BMD; 40.3%, 48.7%, 53.9%, respectively. CA rates were as fol- lows: Vitek 2 92.4% and Etest 71.9%. While the compatibility of Vitek 2 and BMD was observed (kappa value= 0.85) to be 'excellent agreement'; the agreement of Etest and BMD was found to be 'moderate' (kappa value= 0.45). Although CA varied from 85.7% to 100% for Vitek 2, it ranged 63.6% to 80% for Etest depending on bacterial species. Alarming high rates of VME were determined for Vitek 2 (14.5%) and Etest (36.5%). While MEs were 1.7% by Vitek 2; there was no false resistant isolate with Etest. Conclusion: It may be recommended for laboratories not to rely on Vitek 2 and Etest colistin susceptibility results. Additionally, colistin resistant isolates will be underestimated by reducing colistin susceptibility studies to a specific minimum inhibitory concentration (MIC). In this regard, our suggestion is that laboratories would improve their infrastructure and staff skills to apply BMD routinely.