In recent years, because of carbapenemase spreading in Klebsiella pneumoniae strains, the antibiotic of reserve group, colistin, is increasingly prescribed. In vitro testing of colistin susceptibility in everyday practice has a number of difficulties due to the cationic properties of molecule and weak diffusion into agar. Therefore it is recommended to use the reference Broth Microdilution Method (BMD) for determination of the Minimum Inhibitory Concentration (MIC) for colistin. The purpose of the study was to determine susceptibility to colistin in 119 carbapenem-resistant K. pneumoniae (CRKp) which were isolated from the patients at three hospitals in Moscow in 2012-2016 by the broth microdilution method (BMD) and to compare these data with the ones obtained by epsilometer test (E-test) and VITEK 2 Compact. The proportion of resistant isolates (MIC>2 mg/L) was 52%, 39%, 35% respectively. Both commercial methods demonstrated a high level of the very major error (VME) that was 26% for the E-test method and 34% for the VITEK 2 Compact. The values of categorical agreement and essential agreement (CA, EA) were less than 90%. A single major error (ME) was detected for the VITEK 2 Compact. In conclusion, results of both commercial tests for determination of MIC for colistin showed differences with the results of the reference BMD. It is necessary for clinical laboratories to be aware about this discrepancy and to use E-tests and VITEK 2 Compact with caution to determine colistin susceptibility.
Keywords: E-test; Klebsiella pneumoniae; Vitek 2; broth microdilution method; colistin.