Why the study?
ICU readmission carries poor outcomes and high costs, and traditional scoring methods for discharge readiness have failed to meet expectations.
Does an XGBoost machine learning model improve the prediction of ICU readmission or death in adult ICU patients?
Does an XGBoost machine learning model improve the prediction of ICU readmission or death in adult ICU patients?
A lightweight gradient boosting model trained on standardized OMOP-CDM data accurately predicts ICU readmission, outperforming existing models and facilitating future external validation.
May aid ML-based ICU discharge prediction on MIMIC-OMOP; leaves open prospective validation and clinical utility.
Readmission in the intensive care unit (ICU) is associated with poor clinical outcomes and high costs. Traditional scoring methods to help clinicians deciding whether a patient is ready for discharge have failed to meet expectations, paving the way for machine learning based approaches. Freely available datasets such as MIMIC-III have served as benchmarking media to compare such tools. We used the OMOP-CDM version of MIMIC-III (MIMIC-OMOP) to train and evaluate a lightweight tree boosting method to predict readmission in ICU at different time points after discharge (3, 7 and 30 days), outperforming existing solutions with an AUROC of 0.805 for 3-days readmission.
No takes yet. Share an insight, caveat, or question.
Dinh-Phong Nguyen (2019) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: