Cumulative antibiograms: effects of isolate selection and stratification order on susceptibility estimates and resistance representation
Year: 2026
Type: preprint
Abstract: Cumulative antibiograms depend on how isolates are selected and how data are stratified. Different selection strategies may affect both the number of isolates included and the susceptibility estimates reported. We retrospectively analyzed routine microbiology records from a tertiary-care hospital in Barcelona, Spain (2022-2025). Within each patient-organism-calendar-year unit, we compared three isolate-selection strategies: first-isolate, resistance-phenotype, and susceptibility-profile. We assessed the number of isolates retained, susceptibility at standard or increased exposure, the identification of resistance-phenotype profiles, and the effect of de-duplication before versus after stratification. Annual differences in susceptibility were estimated using paired cluster-bootstrap confidence intervals. For Pseudomonas aeruginosa, susceptibility-profile selection retained 16.9% more isolates than first-isolate selection, with differences exceeding 5 percentage points in 10 of 44 annual antimicrobial comparisons. No annual susceptibility differences exceeding this threshold were observed for the other species. Resistance-phenotype and susceptibility-profile selection criteria identified 468 MDR and 278 XDR P. aeruginosa isolates. Of these, 90 (19.2%) of the MDR and 53 (19.1%) of the XDR isolates were not captured by first-isolate selection. In the broader stratification dataset, stratification-first processing increased first-isolate retention by 21.2% for blood specimens and 11.2% for inpatients. These findings show that modest changes in cumulative susceptibility can coexist with incomplete representation of resistant phenotypes. In conclusion, phenotype- and AST-profile-based selection capture additional resistance information beyond first-isolate selection, despite modest differences in cumulative susceptibility. Clinical validation is needed before their use to guide empirical therapy.
Institution: Hospital de Sant Pau
Language: English
Subfield: Microbiology (medical)
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7236024759
