Evaluating the necessity of gradient diffusion confirmations for piperacillin-tazobactam susceptibility in bacteremia

Abstract Routine confirmatory testing of automated susceptibility results using gradient diffusion strips is common practice in many centers, particularly for agents like piperacillin-tazobactam. We evaluated whether such confirmatory testing meaningfully improves clinical decision-making in bloodstream infections in an intermediate ESBL prevalence setting. We retrospectively evaluated the clinical relevance of routine gradient diffusion strip (GDT) confirmations for piperacillin–tazobactam (TZP) susceptibility in 672 Gram-negative bloodstream isolates tested in parallel with VITEK2 and GDT. Discrepancies were categorized and reviewed for their potential therapeutic impact. The overall categorical agreement between methods was high, with concordant classification in 652/672 isolates (97.02%); 597/613 (97.4%) among Enterobacterales and 55/59 (93.2%) among Pseudomonas spp. Very major discrepancies were observed in four E. coli isolates. Major discrepancies occurred in 12 Enterobacterales and in four Pseudomonas spp. isolates. Clinical relevance of the very major discrepancies was limited, as all four cases involved borderline MICs and/or ESBL expression and clinical management was ultimately guided by source of infection and local stewardship recommendations rather than TZP susceptibility alone. In this cohort, VITEK2 and GDT showed strong categorical agreement for TZP susceptibility in bloodstream isolates of Enterobacterales and Pseudomonas spp., with few clinically relevant discrepancies and no identified adverse therapeutic consequence attributable to the discrepant results. Our findings support moving from universal confirmation toward more targeted strategies, such as restricting confirmation to isolates with borderline MIC values.