The accuracy of four commercial broth microdilution tests in the determination of the minimum inhibitory concentration of colistin.

Colistin is considered as one of the last-resort antibiotics and reliable antimicrobial susceptibility testing is therefore crucial. The reference standard for AST according to EUCAST and CLSI is broth microdilution (BMD). However, BMD is labor intensive to perform. Commercial antimicrobial susceptibility tests derived from BMD method are available. We investigated the performance of four different commercial tests: Sensititre™, SensiTest™ Colistin, Micronaut MIC Strip Colistin and UMIC Colistin using 70 clinical isolates (half of them was deemed by VITEK2 as resistant), including isolates from cystic fibrosis patients and mcr-1 bearing isolates. We used two reference standards: BMD and composite MIC as determined by all four tests. Sensititre™ had essential agreement (EA, defined as minimum inhibitory concentration within ± 1 dilution) of 87% and 89% compared to BMD and composite reference standard, respectively. For SensiTest™, the EA's were 93% and 90%. For UMIC, 87% and 90%, and for Micronaut, 83% and 84%. All four tests demonstrated categorical agreement (CA) above 90%. CA for SensiTest™ and Micronaut was both 96%, UMIC 94%, and Sensititre™ 93%. All tests were reproducible as tested in two quality control isolates. In conclusion, in clinical isolates from a large referral center, the four commercial tests for determination of colistin minimum inhibitory concentrations showed acceptable performance.