Objective: To determine the accuracy and utility of colistin susceptibility testing from the agar dilution (AD), agar gradient diffusion (AG), and disk diffusion (DD) methods compared to broth microdilution (BMD), which is considered the reference method for colistin susceptibility testing of E. coli and K. pneumoniae clinical isolates.

Materials and Methods: E. coli and K. pneumoniae isolates were evaluated for colistin susceptibility by the AD, AG, DD, and BMD methods, and the AD, AG, and DD results were compared with the minimum inhibitory concentration (MIC) of colistin from BMD. The reference value breakpoints for colistin susceptibility from BMD were 1 or less and 2 mg/L or less.

Results: Three hundred twenty-six non-duplicate clinical isolates of E. coli and K. pneumoniae were included. Of those, 16 and 214 were carbapenem-resistant E. coli and K. pneumoniae, respectively. Colistin MIC by AD was in 100% agreement with BMD. Overall agreement of AG and DD with BMD was moderate, but some values of AG and DD were found to be useful.

Conclusion: Colistin MIC measured by AG (2 or less and 4 or more mg/L), and inhibition zone diameters of colistin disk by DD (14 or more and 11 or less mm) are useful for determining colistin susceptibility in carbapenem-resistant E. coli and K. pneumoniae in settings where colistin MIC measured by BMD is unavailable.

Keywords: Colistin, Antibiotic susceptibility test, Broth microdilution, Agar dilution, Agar gradient diffusion, Disk diffusion, E. coli, K. pneumoniae