UTI guidelines are moving away from complicated or uncomplicated infections to localized or systemic classification terminologies. Guidelines author Zachary Nelson, PharmD, MPH, BCIDP, provides furthe...

Antibiogram Role in Antimicrobial stewardship is a key instrument in working to improve the use of microbiologic data in stewardship

To the  Editor—We read with interest the article by Goto, Bandyopadhyay, and colleagues describing a multi-task XGBoost framework for predicting antimicrobial resistance in Escherichia coli and Klebsiella spp. [1]. The “personalized antibiogram” concept represents a significant advancement, reducing false-negative rates for carbapenem resistance from 39.9% with conventional antibiograms to 21.7% with the multi-task model for E. coli. We highlight an opportunity to further characterize the mechanistic basis underlying the multi-task learning gains.

Most children hospitalized in the United States for AHO and ABA receive empiric MRSA therapy, although few have MRSA identified. Hospital-wide antibiograms consistently have a higher MRSA rate than seen...

CONCLUSIONS: Most children hospitalized in the United States for AHO and ABA receive empiric MRSA therapy, although few have MRSA identified. Hospital-wide antibiograms consistently have a higher MRSA rate than seen in patients with AHO and ABA, and development of AHO- and ABA-specific antibiograms may improve targeted empiric therapy for children with osteoarticular infections.

AbstractIntroductionAntimicrobial resistance in Gram-negative bacilli represents a growing clinical challenge that demands innovative tools to optimize antibiogram interpretation and support safe therapeutic decisions.ObjectiveTo develop and validate a custom generative pre-trained transformer (GPT) model for antibiogram interpretation, comparing its performance with an alternative model based on Gemini (Gems).MethodsA retrospective, cross-sectional diagnostic accuracy study was conducted following the STARD-artificial intelligence (AI) guideline. A total of 210 antibiograms were analysed, balanced across controls and five resistance mechanisms (penicillinases, broad-spectrum β-lactamases, extended-spectrum β-lactamases, AmpC and Carba-R). The custom GPT, configured through prompt engineering and retrieval-augmented generation, was evaluated against expert interpretation and compared with Gems. Performance metrics (sensitivity, specificity, precision, F1-score, receiver operating characteristic/precision–recall area under the curve and Cohen’s Kappa) were calculated, and errors were classified by clinical impact as overtreatment or undertreatment.ResultsThe GPT achieved an accuracy of 95.7% (95% confidence interval: 92.0–98.0) and a Kappa of 0.948, outperforming Gems (91.9% and 0.902, respectively). Both models showed good performance, but GPT displayed greater consistency across metrics, with sensitivities and F1-scores >90% in all classes. Clinically relevant errors were fewer and less severe with GPT (9 versus 17 in Gems), with markedly lower spectrum penalty and treatment failure risk indices.ConclusionsThis study provides proof-of-concept evidence that a custom GPT model can accurately interpret antibiograms of Gram-negative bacilli under controlled retrospective condition. Although these findings are promising, external multicentre validation, expansion to other pathogens and resistance mechanisms and further assessment as a clinical decision-support tool are required before routine clinical implementation.

Conventional hospital antibiograms summarize aggregated resistance rates, limiting their utility for individualized antimicrobial selection. Existing statistical and machine learning models predict each phenotype separately, ignoring correlations among resistance profiles. We developed novel multitask extreme gradient boosting (XGBoost) models utilizing structured data in electronic health records (EHRs) to predict resistance to 8 antimicrobial classes simultaneously and evaluated their performance within the Veterans Health Administration (VHA).MethodsWe conducted a retrospective multicenter study of Escherichia coli and Klebsiella spp. isolates collected at 127 hospitals and >1400 clinics from January 2017 to September 2024. Data from January 2017 to September 2023 were used for model development, while data from October 2023 to September 2024 were used for simulated prospective testing. Model performances were compared to hospital antibiograms and single-target XGBoost models.ResultsThe training cohort included 536 252 E. coli and 246 898 Klebsiella spp. isolates; the test cohort included 75 138 and 38 015 isolates, respectively. On the test data, the multitask model achieved overall areas under the receiver operating characteristic curve (AUROCs) of 0.779 (E. coli) and 0.810 (Klebsiella spp.), with good to excellent per-class performance (AUROC range, 0.743–0.847). A multitask approach improved calibration and decreased false-negative rates for carbapenem resistance while predicting individualized resistance probabilities for all target antimicrobials simultaneously (“personalized antibiograms”).ConclusionsA multitask XGBoost framework can accurately predict individualized resistance profiles for common Gram-negative pathogens, outperforming conventional antibiograms and single-target models. Personalized antibiograms may enhance the selection of empiric therapy, including the detection of carbapenem resistance in low-endemicity settings.

AbstractObjectivesAntimicrobial resistance (AMR) is a serious threat to global health. However, current stewardship approaches rely on static antibiograms that overlook temporal changes in resistance. We developed AMROrbit, an explainable trajectory-based scorecard to proactively identify emerging AMR patterns using routinely generated antimicrobial susceptibility data.MethodsAs a proof of concept, AMROrbit was applied to the Pfizer ATLAS dataset, including ESKAPE pathogens and Escherichia coli across four samples. For each antibiotic–microorganism–sample–country combination, yearly resistance percentages were modelled in rolling windows to generate estimates of resistance amplitude (baseline resistance) and velocity (rate of change). The estimates were plotted in a 2D space stratified using global medians into four quadrants. Q1 (ideal; low amplitude, low velocity), Q2 (high, low), Q3 (low, high) and Q4 (concerning; high, high). Temporal trajectories were classified as containment or spiral-in (moving towards Q1), persistent (stable within quadrant) or spiral-out (moving towards Q4).ResultsAnalysis revealed 62.4% of the combinations demonstrate increasing baseline resistance. Among the 5920 country-level trajectories, 17.7% depicted spiralling-out patterns, compared with only 13.4% spiralling-in, suggesting a concerning global imbalance. The patterns differed across sample-microorganism–antibiotic combinations, highlighting heterogeneity in trajectories across settings. An open-access dashboard and analytical suite are provided to generate trajectory-based stewardship insights from routine antimicrobial susceptibility testing datasets (https://amrorbit.tavlab.iiitd.edu.in:3002/).ConclusionsAMROrbit extends conventional surveillance by jointly characterizing resistance levels and their rate of change within an interpretable trajectory-based framework. This allows emerging resistance patterns to be identified earlier and may support more timely stewardship responses. Its reliance on routinely collected microbiology data supports scalability across diverse health system settings.

Both unit-specific combination antibiograms and patient-specific factors help improve the appropriateness of empiric antibiotic selection for intensive car

YouTube video by Epicentre - MSF

We present complete genome assemblies of two S. capitis type strains (subspecies capitis: DSM 20326; subspecies urealyticus: DSM 6717) and 20 clinical isolates (NRCS-A: 10) from England. Each genome is...

Physician knowledge, attitudes, and perceptions of antibiograms: a pre-implementation study in southern Sri Lanka - Volume 5 Issue 1