Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 3 min ago
Peer-reviewedCC BY 4.0npj Digital Medicine

Predicting persistent fever in patients with cancer receiving antibiotics: a machine learning approach

  • Gernot Pucher1,2,3
  • Sanjat Kanjilal4,5
  • Kevin Kopp1,2,3
  • Aman Deep1,2,3
  • Till Rostalski2
  • Martin Schuler3,6,7
  • Felix Nensa2,3
  • Jens Kleesiek2,3,8,9,10
  • H. Christian Reinhardt3,7,8,11,12
  • Christopher M. Sauer1,2,3
12 affiliations
  1. 1Laboratory for Clinical Research and Real-World Evidence, Department of Hematology & Stem Cell Transplantation, West German Cancer Center, University Hospital Essen, Essen, Germany
  2. 2Institute for Artificial Intelligence in Medicine, University Hospital Essen, Essen, Germany
  3. 3National Center for Tumor Diseases (NCT-West), Campus Essen, Essen, Germany
  4. 4Department of Medical Microbiology and Infection Prevention, Amsterdam University Medical Centers, Amsterdam, The Netherlands
  5. 5Brigham and Women’s Hospital, Division of Infectious Diseases, Harvard Medical School, Boston, MA USA
  6. 6Department of Medical Oncology, West German Cancer Center, University Hospital Essen, Essen, Germany
  7. 7West German Cancer Center, University Hospital Essen, Essen, Germany
  8. 8DKTK Partner Site Essen/Düsseldorf, University Hospital Essen, Essen, Germany
  9. 9Department of Physics, TU Dortmund University, Dortmund, Germany
  10. 10Faculty of Medicine, University of Duisburg-Essen, Essen, Germany
  11. 11Department of Hematology & Stem Cell Transplantation, West German Cancer Center, University Hospital Essen, Essen, Germany
  12. 12Center for Molecular Biotechnology, University Hospital Essen, Essen, Germany
Read full textDOICite

Abstract

Persistent fever after initiating intravenous antibiotics is common in hospitalized patients with cancer and often leads to additional imaging and antimicrobial escalation. Early identification of patients unlikely to remain febrile could provide additional information during clinical reassessment. We developed and externally validated machine learning models to predict persistent fever 48–72 h after antibiotic initiation using electronic health records from two academic hospitals. Adult cancer patients with fever within 24–48 h of treatment were included. Models incorporated demographics, comorbidities, laboratory results, vital signs, and temperature patterns. The best-performing model achieved an AUROC of 0.82 (95% CI 0.79–0.84) internally and 0.78 (95% CI 0.77–0.79) externally. Temperature-related variables were the dominant predictors. Decision curve analysis demonstrated positive net benefit across much of the evaluated threshold range. Retrospective assessment suggested that approximately one quarter of antibiotic escalations and chest CT scans occurred in patients correctly predicted not to remain febrile, highlighting the potential contextual value of fever predictions. These findings indicate that short-term fever persistence can be predicted with clinically meaningful accuracy using routine data. However, clinical use should be limited to an adjunctive role rather than standalone rule-out decision-making, and prospective studies are required to evaluate safety, clinician uptake, and real-world impact.

Related

  1. Predicting infection risk in rheumatoid arthritis patients receiving biological or targeted synthetic disease-modifying anti-rheumatic drugs: an application of machine learning and healthcare big data.
    Kuan Peng, Deliang Yang, Jiaqi Wang +11
    BMC medical informatics and decision makingPeer-reviewedCC BY-NC-NDFull textGeneral medicineEurope PMCPMC
  2. 994 Predicting immune-related inflammatory arthritis in patients with cancer receiving immunotherapy
    Namrata Singh, Pankti Reid, Wooyoung Kim +3
    Journal for ImmunoTherapy of CancerPeer-reviewedCC BYOncologyDOAJ
  3. Explainable machine learning for predicting grade ≥3 treatment-related adverse events in resectable non-small cell lung cancer receiving neoadjuvant chemoimmunotherapy: predominant contribution of hematologic toxicity.
    Fei-Hang Zhi, Yunzeng Zhang, Tengfei Ge +14
    Translational lung cancer researchPeer-reviewedCC BY-NC-NDFull textOncologyEurope PMC
  4. MRI-based intratumoral and peritumoral heterogeneity for predicting intracranial progression in EGFR-mutant lung adenocarcinoma patients with treatment-naive brain metastases receiving first-line third-generation EGFR-TKIs.
    Guangchuan Deng, Houying Li, Zhengjun Dai +5
    Journal of translational medicinePeer-reviewedCC BY-NC-NDFull textMedical imaging & radiologyEurope PMCPMC
  5. Explainable machine learning and deep learning models for predicting in-hospital mortality in burn patients
    Reza Zarei, Maral Taghizadeh Shalmaei, Masiha Mobayen +2
    Scientific ReportsPeer-reviewedCC BY-NC-ND 4.0Full textEpidemiology & public healthCrossref

Cite

BibTeXRIS
BibTeX
@article{pucher2026predicting,
  title = {{Predicting persistent fever in patients with cancer receiving antibiotics: a machine learning approach}},
  author = {Gernot Pucher and Sanjat Kanjilal and Kevin Kopp and Aman Deep and Till Rostalski and Martin Schuler and Felix Nensa and Jens Kleesiek and H. Christian Reinhardt and Christopher M. Sauer},
  journal = {npj Digital Medicine},
  year = {2026},
  month = oct,
  doi = {10.1038/s41746-026-03449-w},
  url = {https://doi.org/10.1038/s41746-026-03449-w},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Predicting persistent fever in patients with cancer receiving antibiotics: a machine learning approach
AU  - Gernot Pucher
AU  - Sanjat Kanjilal
AU  - Kevin Kopp
AU  - Aman Deep
AU  - Till Rostalski
AU  - Martin Schuler
AU  - Felix Nensa
AU  - Jens Kleesiek
AU  - H. Christian Reinhardt
AU  - Christopher M. Sauer
PY  - 2026
DA  - 2026/10/01
JO  - npj Digital Medicine
DO  - 10.1038/s41746-026-03449-w
UR  - https://doi.org/10.1038/s41746-026-03449-w
AB  - Persistent fever after initiating intravenous antibiotics is common in hospitalized patients with cancer and often leads to additional imaging and antimicrobial escalation. Early identification of patients unlikely to remain febrile could provide additional information during clinical reassessment. We developed and externally validated machine learning models to predict persistent fever 48–72 h after antibiotic initiation using electronic health records from two academic hospitals. Adult cancer patients with fever within 24–48 h of treatment were included. Models incorporated demographics, comorbidities, laboratory results, vital signs, and temperature patterns. The best-performing model achieved an AUROC of 0.82 (95% CI 0.79–0.84) internally and 0.78 (95% CI 0.77–0.79) externally. Temperature-related variables were the dominant predictors. Decision curve analysis demonstrated positive net benefit across much of the evaluated threshold range. Retrospective assessment suggested that approximately one quarter of antibiotic escalations and chest CT scans occurred in patients correctly predicted not to remain febrile, highlighting the potential contextual value of fever predictions. These findings indicate that short-term fever persistence can be predicted with clinically meaningful accuracy using routine data. However, clinical use should be limited to an adjunctive role rather than standalone rule-out decision-making, and prospective studies are required to evaluate safety, clinician uptake, and real-world impact.
SN  - 2398-6352
ER  -