Cross-database adaptation of machine learning models for predicting sustained antibiotic exposure among patients with intracerebral hemorrhage in intensive care units
- 1Department of Health Science, California State University, Long Beach, Long Beach, CA, USA
Abstract
BackgroundPatients with intracerebral hemorrhage (ICH) in intensive care units (ICUs) frequently receive antibiotics early in critical illness. Such exposure reflects suspected infection, physiological deterioration, monitoring intensity, and clinical management and may vary across institutions and electronic health record databases. We evaluated target-domain adaptation for modeling sustained antibiotic exposure in critically ill patients with ICH.
MethodsWe analyzed derived ICH cohorts from MIMIC-IV and the eICU Collaborative Research Database. MIMIC-IV included 2,024 ICU stays and served as the source cohort. The eICU cohort comprised 1,277 stays and was divided, at the ICU-stay level, into a development subset (766 stays) and a held-out evaluation subset (511 stays). The eICU outcome was sustained antibiotic exposure on at least two ICU days. Candidate models incorporated demographic, comorbidity, physiological, laboratory, neurologic, respiratory, temporal, and measurementprocess features. Timestamped observations for static and temporal tabular families were censored at the recorded first-antibiotic time when available, whereas the compact sequence component and explicit measurement-density features used the full first 24 hours. For base-model families using eICU development data, meta-learner inputs were generated by five-fold stratified cross-fitting. A 42-input L2-penalized logistic meta-learner combined 16 model probabilities with model-summary, clinical, availability, and measurement-density features.
ResultsThe held-out eICU subset contained 48 positive outcomes (9.39%). The adapted ensemble achieved an area under the receiver operating characteristic curve (AUROC) of 0.841 (95% CI 0.775-0.899) and an average precision (AP) of 0.540 (95% CI 0.399-0.676), with a Brier score of 0.063 and expected calibration error of 0.029. The MIMIC-IV source-only stack achieved an AUROC of 0.782; the paired AUROC difference was 0.059 (95% CI 0.003-0.112). In a separate patient-grouped five-fold out-of-fold sensitivity analysis of 503 linked eICU stays, AUROC was 0.807 (95% CI 0.691-0.910).
ConclusionsTarget-domain adaptation improved held-out discrimination relative to a source-only stack for sustained antibiotic exposure. Because some model components and measurement-process features used information from the broader first 24 hours, the model should be interpreted as a first-24-hour treatment-association model rather than a strictly prospective pre-treatment infection predictor.
Related
- Machine learning for predicting 365-day mortality in spontaneous intracerebral hemorrhage patients in the intensive care unit
- Head-to-head comparison of glucose-to-albumin ratio against five albumin-derived indices for predicting 28-day ICU mortality in critically ill patients with spontaneous intracerebral hemorrhage: a multicenter retrospective study.
- Cross-trait genetic architecture of intracerebral hemorrhage highlights convergent vascular, hematologic and lipid pathways
- PINK1 agonist MTK458 attenuates cerebral edema by protecting blood-brain barrier integrity in intracerebral hemorrhage mice
- An automated, explainable, NCCT-based clinical decision-support system for spontaneous intracerebral hemorrhage.
BibTeX
@misc{pathak2026crossdatabase,
title = {{Cross-database adaptation of machine learning models for predicting sustained antibiotic exposure among patients with intracerebral hemorrhage in intensive care units}},
author = {S. Pathak and Y. Si and J. Sanjaya and M. Haghi and N. T. Kudrot and G. Placencia and K. Alaei and M. Pishgar},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.11.26365235},
url = {https://www.medrxiv.org/content/10.64898/2026.10.11.26365235v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Cross-database adaptation of machine learning models for predicting sustained antibiotic exposure among patients with intracerebral hemorrhage in intensive care units
AU - S. Pathak
AU - Y. Si
AU - J. Sanjaya
AU - M. Haghi
AU - N. T. Kudrot
AU - G. Placencia
AU - K. Alaei
AU - M. Pishgar
PY - 2026
DA - 2026/10/11
PB - medRxiv
DO - 10.64898/2026.10.11.26365235
UR - https://www.medrxiv.org/content/10.64898/2026.10.11.26365235v1
AB - BackgroundPatients with intracerebral hemorrhage (ICH) in intensive care units (ICUs) frequently receive antibiotics early in critical illness. Such exposure reflects suspected infection, physiological deterioration, monitoring intensity, and clinical management and may vary across institutions and electronic health record databases. We evaluated target-domain adaptation for modeling sustained antibiotic exposure in critically ill patients with ICH. MethodsWe analyzed derived ICH cohorts from MIMIC-IV and the eICU Collaborative Research Database. MIMIC-IV included 2,024 ICU stays and served as the source cohort. The eICU cohort comprised 1,277 stays and was divided, at the ICU-stay level, into a development subset (766 stays) and a held-out evaluation subset (511 stays). The eICU outcome was sustained antibiotic exposure on at least two ICU days. Candidate models incorporated demographic, comorbidity, physiological, laboratory, neurologic, respiratory, temporal, and measurementprocess features. Timestamped observations for static and temporal tabular families were censored at the recorded first-antibiotic time when available, whereas the compact sequence component and explicit measurement-density features used the full first 24 hours. For base-model families using eICU development data, meta-learner inputs were generated by five-fold stratified cross-fitting. A 42-input L2-penalized logistic meta-learner combined 16 model probabilities with model-summary, clinical, availability, and measurement-density features. ResultsThe held-out eICU subset contained 48 positive outcomes (9.39%). The adapted ensemble achieved an area under the receiver operating characteristic curve (AUROC) of 0.841 (95% CI 0.775-0.899) and an average precision (AP) of 0.540 (95% CI 0.399-0.676), with a Brier score of 0.063 and expected calibration error of 0.029. The MIMIC-IV source-only stack achieved an AUROC of 0.782; the paired AUROC difference was 0.059 (95% CI 0.003-0.112). In a separate patient-grouped five-fold out-of-fold sensitivity analysis of 503 linked eICU stays, AUROC was 0.807 (95% CI 0.691-0.910). ConclusionsTarget-domain adaptation improved held-out discrimination relative to a source-only stack for sustained antibiotic exposure. Because some model components and measurement-process features used information from the broader first 24 hours, the model should be interpreted as a first-24-hour treatment-association model rather than a strictly prospective pre-treatment infection predictor.
ER -