Dynamic near-term risk prediction of clinically significant immune-related adverse events using longitudinal electronic health records
- 1Columbia University
Abstract
ObjectiveTo develop and evaluate a dynamic model that repeatedly estimates the 30-day risk of a first grade [≥]2 immune-related adverse event (irAE) during immune checkpoint inhibitor (ICI) therapy.
Materials and MethodsAdults initiating ICI therapy during 2015-2022 comprised the development cohort; 2023 initiators formed a nonoverlapping internal temporal evaluation cohort. Every 21 days during the first treatment year, models used available EHR information to predict a first structured phenotype-defined grade [≥]2 irAE during the next 30 days. Evaluation included pooled and within-landmark discrimination, precision-recall performance, calibration, and patient-clustered bootstrap confidence intervals. Feature-state ablations and alternative update schedules examined the contribution of dynamic information.
ResultsDevelopment included 1,522 patients and 18,744 evaluable landmarks; temporal evaluation included 526 patients and 6,263 landmarks, of which 184 were positive (2.94%). The multiscale XGBoost model achieved pooled AUROC 0.716 (95% CI, 0.676-0.755), AUPRC 0.067, within-landmark AUROC 0.672 (95% CI, 0.631-0.716), Brier score 0.0281, and log loss 0.1245. The time-aware baseline achieved AUROC 0.631 and within-landmark AUROC 0.536. Updated treatment exposure and laboratory trajectories accounted for most of the gain. A complete-state extension achieved AUROC 0.719 and AUPRC 0.073. The 21-day schedule had the highest discrimination among tested cadences.
ConclusionLongitudinal EHR updating improved near-term irAE prediction beyond pretreatment characteristics and treatment time. Phenotype adjudication and independent evaluation of calibration, transportability, and clinical utility are needed before implementation.
Related
- Dynamic Prediction to Predict Virologic Failure in Antiretroviral Treated People Living with HIV: Cohort Analysis of Routine Electronic Health Records Data in the Western Cape, South Africa
- Characterization and Antimicrobial Susceptibility Patterns of Clinically Significant Non-fermenting Gram-Negative Bacilli Isolated From Clinical Specimens
- Transcatheter atrial septal defect closure in infants weighing 3 to 6 kg with clinically significant bronchopulmonary dysplasia
- Validation of the discriminatory performance of the iPREFACE score for clinically significant fetal acidemia: an outcome-enriched nested case–control study
- Clinically significant neck pain and tablet tilt angles in real-world educational settings: A nationwide cross-sectional study of university students
BibTeX
@misc{gong2026dynamic,
title = {{Dynamic near-term risk prediction of clinically significant immune-related adverse events using longitudinal electronic health records}},
author = {L. Gong and A. Agyekum and N. Aswani and H. R. Nieva and C. Hur},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.06.26364907},
url = {https://www.medrxiv.org/content/10.64898/2026.10.06.26364907v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Dynamic near-term risk prediction of clinically significant immune-related adverse events using longitudinal electronic health records
AU - L. Gong
AU - A. Agyekum
AU - N. Aswani
AU - H. R. Nieva
AU - C. Hur
PY - 2026
DA - 2026/10/06
PB - medRxiv
DO - 10.64898/2026.10.06.26364907
UR - https://www.medrxiv.org/content/10.64898/2026.10.06.26364907v1
AB - ObjectiveTo develop and evaluate a dynamic model that repeatedly estimates the 30-day risk of a first grade [≥]2 immune-related adverse event (irAE) during immune checkpoint inhibitor (ICI) therapy. Materials and MethodsAdults initiating ICI therapy during 2015-2022 comprised the development cohort; 2023 initiators formed a nonoverlapping internal temporal evaluation cohort. Every 21 days during the first treatment year, models used available EHR information to predict a first structured phenotype-defined grade [≥]2 irAE during the next 30 days. Evaluation included pooled and within-landmark discrimination, precision-recall performance, calibration, and patient-clustered bootstrap confidence intervals. Feature-state ablations and alternative update schedules examined the contribution of dynamic information. ResultsDevelopment included 1,522 patients and 18,744 evaluable landmarks; temporal evaluation included 526 patients and 6,263 landmarks, of which 184 were positive (2.94%). The multiscale XGBoost model achieved pooled AUROC 0.716 (95% CI, 0.676-0.755), AUPRC 0.067, within-landmark AUROC 0.672 (95% CI, 0.631-0.716), Brier score 0.0281, and log loss 0.1245. The time-aware baseline achieved AUROC 0.631 and within-landmark AUROC 0.536. Updated treatment exposure and laboratory trajectories accounted for most of the gain. A complete-state extension achieved AUROC 0.719 and AUPRC 0.073. The 21-day schedule had the highest discrimination among tested cadences. ConclusionLongitudinal EHR updating improved near-term irAE prediction beyond pretreatment characteristics and treatment time. Phenotype adjudication and independent evaluation of calibration, transportability, and clinical utility are needed before implementation.
ER -