Machine learning and clinician predictions of antibiotic resistance in Enterobacterales bloodstream infections

Kevin Yuan; Augustine Luk; Jia Wei; A Sarah Walker; Tingting Zhu; David W Eyre

doi:10.1016/j.jinf.2024.106388

Machine learning and clinician predictions of antibiotic resistance in Enterobacterales bloodstream infections

J Infect. 2024 Dec 30;90(2):106388. doi: 10.1016/j.jinf.2024.106388. Online ahead of print.

Authors

Kevin Yuan¹, Augustine Luk², Jia Wei², A Sarah Walker³, Tingting Zhu⁴, David W Eyre⁵

Affiliations

¹ Big Data Institute, Nuffield Department of Population Health, University of Oxford, Oxford, UK.
² Nuffield Department of Medicine, University of Oxford, Oxford, UK.
³ Nuffield Department of Medicine, University of Oxford, Oxford, UK; NIHR Health Protection Research Unit in Healthcare Associated Infections and Antimicrobial Resistance, University of Oxford, Oxford, UK; NIHR Oxford Biomedical Research Centre, Oxford, UK.
⁴ Institute of Biomedical Engineering, University of Oxford, Oxford, UK.
⁵ Big Data Institute, Nuffield Department of Population Health, University of Oxford, Oxford, UK; NIHR Health Protection Research Unit in Healthcare Associated Infections and Antimicrobial Resistance, University of Oxford, Oxford, UK; NIHR Oxford Biomedical Research Centre, Oxford, UK; Oxford University Hospitals NHS Foundation Trust, Oxford, UK. Electronic address: david.eyre@bdi.ox.ac.uk.

PMID: 39742978
DOI: 10.1016/j.jinf.2024.106388

Abstract

Background: Patients with Gram-negative bloodstream infections are at risk of serious adverse outcomes without active treatment, but identifying who has antimicrobial resistance (AMR) to target empirical treatment is challenging.

Methods: We used XGBoost machine learning models to predict antimicrobial resistance to seven antibiotics in patients with Enterobacterales bloodstream infection. Models were trained using hospital and community data from Oxfordshire, UK, for patients with positive blood cultures between 01-January-2017 and 31-December-2021. Model performance was evaluated by comparing predictions to final microbiology results in test datasets from 01-January-2022 to 31-December-2023 and to clinicians' prescribing.

Findings: 4709 infection episodes were used for model training and evaluation; antibiotic resistance rates ranged from 7-67%. In held-out test data, resistance prediction performance was similar for the seven antibiotics (AUCs 0.680 [95%CI 0.641-0.720] to 0.737 [0.674-0.797]). Performance improved for most antibiotics when species identifications (available ∼24 h later) were included as model inputs (AUCs 0.723 [0.652-0.791] to 0.827 [0.797-0.857]). In patients treated with a beta-lactam, clinician prescribing led to 70% receiving an active beta-lactam: 44% were over-treated (broader spectrum treatment than needed), 26% optimally-treated (narrowest spectrum active agent), and 30% under-treated (inactive beta-lactam). Model predictions without species data could have led to 79% of patients receiving an active beta-lactam: 45% over-treated, 34% optimally-treated, and 21% under-treated.

Conclusions: Predicting AMR in bloodstream infections is challenging for both clinicians and models. Despite modest performance, machine learning models could still increase the proportion of patients receiving active empirical treatment by up to 9% over current clinical practice in an environment prioritising antimicrobial stewardship.

Keywords: Antibiotic stewardship; Antibiotics; Antimicrobial resistance; Bloodstream infection; Machine learning.