EUVIMEDEUVIMED entdecken
Multizentrische StudieKohortenstudieOpen Access · CC BY 4.0

Machine learning-based prediction of bloodstream infection in rectal carbapenem-resistant Enterobacterales carriers: a multicenter retrospective cohort study

Machine-learning-basierte Vorhersage von Blutstrominfektionen bei rektalen Träger:innen von carbapenemresistenten Enterobacterales: multizentrische retrospektive Kohortenstudie

Kibum Jeon, Wonkeun Song, Dong Hoon Shin, Han-Sung Kim, Hyun Soo Kim, Jacob Lee, Seok Hoon Jeong, Seri Jeong

Frontiers in Cellular and Infection Microbiology · 06.08.2026 · 16

DOI: 10.3389/fcimb.2026.1791859 ↗

Auf einen Blick

Bei 639 rektalen CRE-Träger:innen sagte ein XGBoost-Modell Blutstrominfektionen innerhalb von 30 und 14 Tagen gut voraus. Die wichtigste Einflussgröße war je nach Zeitraum unterschiedlich.

Fragestellung

Welche Merkmale tragen bei rektalen CRE-Träger:innen zur Entwicklung einer Blutstrominfektion bei, und lässt sich das Risiko individuell vorhersagen?

Methode

Multizentrische retrospektive Kohortenstudie in 3 Krankenhäusern in Korea; XGBoost-Modelle mit 173 Merkmalen, SHAP als XAI; Aufteilung in Training und Test.

Ergebnisse

639 CRE-Träger:innen; 72 (11,3 %) entwickelten innerhalb von 30 Tagen und 46 (7,2 %) innerhalb von 14 Tagen eine BSI. AUROC 0,917 (95 %-KI 0,889–0,943) für 30 Tage und 0,854 (95 %-KI 0,821–0,885) für 14 Tage.

Schlussfolgerung der Autor:innen

Die Beiträge der Merkmale unterschieden sich je nach Zeitraum und Interaktionen. Die Studie interpretierte komplexe Faktoren mit ML und XAI und erreichte eine hohe Vorhersageleistung.

Kernaussagen

  • 30-Tage-AUROC 0,917 (95 %-KI 0,889–0,943)
  • 14-Tage-AUROC 0,854 (95 %-KI 0,821–0,885)
  • Wichtigste Merkmale: MIC bzw. Neutrophilenzahl je nach Zeitraum

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Multizentrische Studie in der Zeitschrift Frontiers in Cellular and Infection Microbiology, online erschienen am 06.08.2026. Verfasst von Kibum Jeon und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Laboratory Medicine, Hallym University Hangang Sacred Heart Hospital, Hallym University College of Medicine, Seoul, Republic of Korea. Hauptthemen laut MEDLINE: Bacteremia, Carbapenem-Resistant Enterobacteriaceae, Carrier State, Enterobacteriaceae Infections. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Carbapenem-resistant Enterobacterales (CRE) carriers are at risk of developing subsequent CRE bloodstream infection (BSI). This study aimed to identify features contributing to BSI development in rectal CRE carriers and interpret their interactions using machine learning (ML) and explainable artificial intelligence (XAI), while implementing individualized BSI risk prediction.

Methods

We included patients with first-time positive rectal CRE surveillance cultures during hospitalization at three hospitals in Korea from January 2014 to December 2023. A total of 173 features—laboratory data including the carbapenem minimum inhibitory concentrations (MIC) of rectal CRE isolates and complete blood cell count, demographics, vital signs, comorbidities, medications, transfusions, and procedures —were collected for the 30-day and 14-day follow-up periods, of which 151 remained after removing redundant features. The entire dataset was randomly split into training (80%) and test (20%) sets, and XGBoost ML models were developed. Model performance was evaluated using the area under the receiver operator characteristic curve (AUROC) with 1, 000 bootstrap iterations. SHAP (SHapley Additive exPlanations) was employed as the XAI method.

Results

Among 639 included CRE carriers, 72 (11.3%) and 46 (7.2%) developed BSI within 30 and 14 days, respectively. The 30-day and 14-day models demonstrated excellent performance with AUROCs of 0.917 (95% CI: 0.889-0.943) and 0.854 (95% CI: 0.821-0.885), respectively. In the 30-day model, the most contributory feature was minimum inhibitory concentration (MIC) of imipenem or meropenem against rectal CRE, with BSI risk increasing when MIC ≥8 μg/mL. In the 14-day model, increased neutrophil count (≥7.50×10³/μL) was the most contributory feature. Several features showed nonlinear relationships with BSI risk and contrasting contributions depending on feature interactions. Individual 30-day and 14-day prediction models required at least 12 top-contributing features, achieving AUROCs of 0.873 (95% CI: 0.728-0.969) and 0.864 (95% CI: 0.727-0.959), respectively.

Conclusions

The influence of contributing features varied by time periods and interactions. This study comprehensively interpreted complex factors using ML and XAI, achieving excellent predictive performance. Individualized BSI risk stratification can facilitate determining appropriate timing for preemptive therapy and inform antimicrobial stewardship decisions.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

BacteremiaCarbapenem-Resistant EnterobacteriaceaeCarrier StateEnterobacteriaceae InfectionsMachine LearningRectumFemaleHumansMaleAnti-Bacterial AgentsBoosting Machine Learning AlgorithmsCarbapenemsMicrobial Sensitivity TestsPredictive Learning ModelsRepublic of KoreaRetrospective StudiesROC Curve

Stichwörter der Autor:innen

XGBoost ML modelsbloodstream infectioncarbapenem-resistant Enterobacteralesexplainable artificial intelligencemachine learningrisk prediction

Substanzen

Anti-Bacterial AgentsCarbapenems

Autor:innen und Institutionen

  • Department of Laboratory Medicine, Hallym University Hangang Sacred Heart Hospital, Hallym University College of Medicine, Seoul, Republic of Korea.
  • Department of Laboratory Medicine, Hallym University Kangnam Sacred Heart Hospital, Hallym University College of Medicine, Seoul, Republic of Korea.
  • Department of Laboratory Medicine, Hallym University Chuncheon Sacred Heart Hospital, Hallym University College of Medicine, Chuncheon, Republic of Korea.
  • Department of Laboratory Medicine, Hallym University Sacred Heart Hospital, Hallym University College of Medicine, Anyang, Republic of Korea.
  • Department of Laboratory Medicine, Hallym University Dongtan Sacred Heart Hospital, Hallym University College of Medicine, Hwaseong, Republic of Korea.
  • Department of Internal Medicine, Hallym University Kangnam Sacred Heart Hospital, Hallym University College of Medicine, Seoul, Republic of Korea.

Zitieren

APA

Jeon, K., Song, W., Shin, D. H., Kim, H. S., Kim, H. S., Lee, J., Jeong, S. H., & Jeong, S. (2026). Machine learning-based prediction of bloodstream infection in rectal carbapenem-resistant Enterobacterales carriers: a multicenter retrospective cohort study. Frontiers in Cellular and Infection Microbiology, 16. https://doi.org/10.3389/fcimb.2026.1791859

Vancouver

Jeon K, Song W, Shin DH, Kim HS, Kim HS, Lee J, et al. Machine learning-based prediction of bloodstream infection in rectal carbapenem-resistant Enterobacterales carriers: a multicenter retrospective cohort study. Frontiers in Cellular and Infection Microbiology. 2026;16. doi:10.3389/fcimb.2026.1791859

BibTeX
@article{Jeon2026,
  author = {Jeon, Kibum and Song, Wonkeun and Shin, Dong Hoon and Kim, Han-Sung and Kim, Hyun Soo and Lee, Jacob and Jeong, Seok Hoon and Jeong, Seri},
  title = {Machine learning-based prediction of bloodstream infection in rectal carbapenem-resistant Enterobacterales carriers: a multicenter retrospective cohort study},
  journal = {Frontiers in Cellular and Infection Microbiology},
  year = {2026},
  volume = {16},
  doi = {10.3389/fcimb.2026.1791859}
}
Kostenlos starten →