How stable are antimicrobial susceptibility profiles in serial clinical isolates? A retrospective study informing...

Repeat antimicrobial susceptibility testing is frequently performed whenever the same bacterial species is repeatedly isolated from hospitalized patients, consuming considerable laboratory resources. Although referral to prior antimicrobial susceptibility testing (AST) results may reduce laboratory workload, evidence guiding safe omission of repeat testing is limited. Therefore, we aimed to quantify short-term stability of antimicrobial susceptibility in serial clinical isolates within 7 days and to derive an evidence-informed repeat-testing strategy.

We conducted a retrospective observational study using routine AST data generated in July 2023-October 2025 with the BD Phoenix™ M50 system. Serial AST results from the same patient and species were compared within a 7-day interval, both within and across-specimen types. Evaluated antibiotics reflected local and national reporting and treatment guidelines. Clinically relevant changes were defined as transitions to resistant. To account for analytical variability, transitions accompanied by ≥2 twofold MIC increases were separately analysed. Instability patterns were integrated into a risk-stratified AST referral framework.

A total of 11 831 repeat comparisons across 12 bacterial species were included. Overall categorical agreement exceeded 90% for most organism-antibiotic combinations. Transitions to resistance remained below 5% for most Enterobacterales and Staphylococcus aureus. Higher instability was observed for Pseudomonas aeruginosa and Enterobacter cloacae complex, particularly for β-lactam/β-lactamase inhibitor combinations such as piperacillin-tazobactam (up to 13.9% and 9%, respectively) as well as for coagulase-negative staphylococci, e.g. for flucloxacillin (up to 11%). Across-specimen comparisons demonstrated slightly higher variability than within-specimen analyses, thereby supporting specimen-restricted referral strategies which were based on integrating quantitative instability thresholds with clinical risk considerations.

Antimicrobial susceptibility of serial clinical isolates remains highly stable within 7 days for most species. A conservative species- and specimen-restricted repeat-testing strategy may safely reduce unnecessary AST while preserving detection of clinically relevant resistance. These findings support the implementation of evidence-informed AST referral policies in clinical microbiology laboratories.