Reliability of various antimicrobial susceptibility testing methods for piperacillin/tazobactam in challenging Escherichia coli isolates

Objective: 

Piperacillin/tazobactam antimicrobial susceptibility testing (AST) against Enterobacterales can be challenging. The aim of this study was to assess the reproducibility of various automated (VITEK 2) and nonautomated AST methods (broth microdilution (BMD), minimum inhibitory concentration (MIC) test strip, and disk diffusion) for piperacillin/tazobactam in ‘challenging’ E. coli isolates. 

Methods: 

We performed 20 repeated ASTs for seven clinical E. coli isolates: Two resistant to piperacillin/tazobactam but susceptible to amoxicillin/clavulanic acid, four isolates with various β-lactamase coding genes (two bla_TEM-1, one bla_OXA-1, and one with plasmidal bla_ampC), and one isolate where VITEK 2 initially could not produce MIC measurements for piperacillin/tazobactam (i.e. no results generated).

Results: 

Upon repetition, the same MIC as the mode value (i.e. the most frequent MIC value of each AST method) was found between 21% and 87% (BMD), 46% and 100% (VITEK 2), and 48% and 100% (gradient test) of the repetitions. The range of essential agreement percentage (i.e. ±1 doubling dilution from this mode value) was 53–100% (BMD), 63–100% (VITEK 2), and 100% (gradient test). Percent categorical agreement (same susceptible of resistant category using EUCAST breakpoint v. 14.0) was 71–100% (BMD), 85–92% (VITEK 2), 76–100% (gradient test) and 100% (disk diffusion). 

Conclusions: 

: In conclusion, this study provides insight on the reliability of AST results for piperacillin/tazobactam in challenging E. coli isolates. While the results indicate that most methods are generally reproducible, certain isolates may present inconsistent MIC results.