Establishing piperacillin-tazobactam susceptibility in ceftriaxone non-susceptible Enterobacterales: comparing disk diffusion, Etest, and VITEK 2 automated MIC measurements vs. broth microdilution.

Abstract

Objective: Post-hoc analyses of the MERINO trial highlight the uncertainty associated with establishing piperacillin tazobactam (PTZ) susceptibility in extended-spectrum beta lactamase (ESBL) producing Enterobacterales. Herein, we compare the concordance of susceptibility for PTZ between the VITEK 2, disk diffusion, and Etest with broth microdilution (BMD) as the reference standard.

Methods: Ninety-four consecutive ceftriaxone non-susceptible E. coli and K. pneumoniae bloodstream isolates were identified from patients at 3 hospitals in Montréal, Quebec. BMD was used as the reference standard against which disk diffusion, VITEK 2 (AST-N391), and E test susceptibility testing were compared. Errors were categorized as very major (false susceptible), major (false resistant), and minor (other).

Results: Overall, 68/94 (72.3%) of isolates were susceptible to PTZ by BMD. Disk diffusion made no major or very major errors (0%; 97.5%CI 0-3.8%). The VITEK 2 system had a major error rate of 2.5% (95% CI 0.003- 0.089) and a very major error rate of 26.7% (95% CI 0.08- 0.55); however, all isolates with VITEK 2 minimal inhibitory concentrations (MICs) of ≤4ug/mL were susceptible. Finally, the Etest had a major error rate of 6.3% (95% CI 0.02- 0.14), but no very major errors. Combining VITEK 2-determined susceptibility with a second test led to an increase in the number of correctly classified susceptible organisms.

Conclusions: The VITEK 2 system, and to a lesser extent the Etest, risk major errors. Used alone, the VITEK 2 system also risks very major errors if the estimated MIC is >4ug/mL. Combining VITEK 2 with disk diffusion in isolates with an estimated MIC of 8-16 μg/ml could prevent both major and very major errors.