Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgery

Abstract Background Unplanned postoperative intensive care unit admissions (UIAs) are rare events that cause significant challenges to perioperative workflow. We describe the development of a machine-learning derived model to predict UIAs using only widely used preoperative variables. Methods This w...

Full description

Saved in:
Bibliographic Details
Main Authors: Catherine Chiu, Matthias R. Braehler, Anne L. Donovan, Atul J. Butte, Romain Pirracchio, Andrew M. Bishara
Format: Article
Language:English
Published: BMC 2025-07-01
Series:BMC Anesthesiology
Subjects:
Online Access:https://doi.org/10.1186/s12871-025-03195-8
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849332112664035328
author Catherine Chiu
Matthias R. Braehler
Anne L. Donovan
Atul J. Butte
Romain Pirracchio
Andrew M. Bishara
author_facet Catherine Chiu
Matthias R. Braehler
Anne L. Donovan
Atul J. Butte
Romain Pirracchio
Andrew M. Bishara
author_sort Catherine Chiu
collection DOAJ
description Abstract Background Unplanned postoperative intensive care unit admissions (UIAs) are rare events that cause significant challenges to perioperative workflow. We describe the development of a machine-learning derived model to predict UIAs using only widely used preoperative variables. Methods This was a 3-year retrospective review of all adult surgeries under the General, Vascular, and Thoracic surgical services with anticipated length of greater than 180 minutes at a single institution. A UIA was defined as any post-operative patient recovering in the post-anesthesia care unit (PACU) requiring direct transfer to the intensive care unit (ICU) for higher level of care. We developed our prediction model with a gradient-boosting decision tree algorithm (XGBoost). The model incorporated sixteen generalizable predictor variables that were derived from the demographics and surgical booking details. Validation and evaluation were performed with 10-fold cross validation, and model performance was evaluated using the area under the receiver operating characteristic (ROC) curve, sensitivity, specificity, and likelihood ratio. Results A total of 4658 patients were included for analysis. The incidence of UIAs was 2.3%. With 10-fold cross validation, the area under the ROC curve was 0.80 (95% CI 0.74–0.86). Two decision thresholds were used, which achieved the best specificity of 94% (95% CI 92–96%), best positive likelihood ratio of 4.22 (95% CI 0.99–8.79), and best sensitivity of 82% (95% CI 58–100%). Conclusions Our machine learning-derived model is a reliable tool for the perioperative clinician to predict a rare outcome in high-risk patients using only preoperative variables. Future studies will include prospective validation of this model at other institutions and real-time incorporation for improvement in perioperative workflow.
format Article
id doaj-art-3c8d67cef985476fad7a9ff1f1ab2baa
institution Kabale University
issn 1471-2253
language English
publishDate 2025-07-01
publisher BMC
record_format Article
series BMC Anesthesiology
spelling doaj-art-3c8d67cef985476fad7a9ff1f1ab2baa2025-08-20T03:46:19ZengBMCBMC Anesthesiology1471-22532025-07-0125111010.1186/s12871-025-03195-8Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgeryCatherine Chiu0Matthias R. Braehler1Anne L. Donovan2Atul J. Butte3Romain Pirracchio4Andrew M. Bishara5Department of Anesthesia and Perioperative Care, University of CaliforniaDepartment of Anesthesia and Perioperative Care, University of CaliforniaDepartment of Anesthesia and Perioperative Care, University of CaliforniaBakar Computational Health Sciences Institute, University of California San FranciscoDepartment of Anesthesia and Perioperative Care, University of CaliforniaDepartment of Anesthesia and Perioperative Care, University of CaliforniaAbstract Background Unplanned postoperative intensive care unit admissions (UIAs) are rare events that cause significant challenges to perioperative workflow. We describe the development of a machine-learning derived model to predict UIAs using only widely used preoperative variables. Methods This was a 3-year retrospective review of all adult surgeries under the General, Vascular, and Thoracic surgical services with anticipated length of greater than 180 minutes at a single institution. A UIA was defined as any post-operative patient recovering in the post-anesthesia care unit (PACU) requiring direct transfer to the intensive care unit (ICU) for higher level of care. We developed our prediction model with a gradient-boosting decision tree algorithm (XGBoost). The model incorporated sixteen generalizable predictor variables that were derived from the demographics and surgical booking details. Validation and evaluation were performed with 10-fold cross validation, and model performance was evaluated using the area under the receiver operating characteristic (ROC) curve, sensitivity, specificity, and likelihood ratio. Results A total of 4658 patients were included for analysis. The incidence of UIAs was 2.3%. With 10-fold cross validation, the area under the ROC curve was 0.80 (95% CI 0.74–0.86). Two decision thresholds were used, which achieved the best specificity of 94% (95% CI 92–96%), best positive likelihood ratio of 4.22 (95% CI 0.99–8.79), and best sensitivity of 82% (95% CI 58–100%). Conclusions Our machine learning-derived model is a reliable tool for the perioperative clinician to predict a rare outcome in high-risk patients using only preoperative variables. Future studies will include prospective validation of this model at other institutions and real-time incorporation for improvement in perioperative workflow.https://doi.org/10.1186/s12871-025-03195-8Unanticipated ICU admissionMachine learningArtificial intelligencePredictive analytics
spellingShingle Catherine Chiu
Matthias R. Braehler
Anne L. Donovan
Atul J. Butte
Romain Pirracchio
Andrew M. Bishara
Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgery
BMC Anesthesiology
Unanticipated ICU admission
Machine learning
Artificial intelligence
Predictive analytics
title Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgery
title_full Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgery
title_fullStr Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgery
title_full_unstemmed Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgery
title_short Development of a machine learning-derived model to predict unplanned ICU admissions after major non-cardiac surgery
title_sort development of a machine learning derived model to predict unplanned icu admissions after major non cardiac surgery
topic Unanticipated ICU admission
Machine learning
Artificial intelligence
Predictive analytics
url https://doi.org/10.1186/s12871-025-03195-8
work_keys_str_mv AT catherinechiu developmentofamachinelearningderivedmodeltopredictunplannedicuadmissionsaftermajornoncardiacsurgery
AT matthiasrbraehler developmentofamachinelearningderivedmodeltopredictunplannedicuadmissionsaftermajornoncardiacsurgery
AT anneldonovan developmentofamachinelearningderivedmodeltopredictunplannedicuadmissionsaftermajornoncardiacsurgery
AT atuljbutte developmentofamachinelearningderivedmodeltopredictunplannedicuadmissionsaftermajornoncardiacsurgery
AT romainpirracchio developmentofamachinelearningderivedmodeltopredictunplannedicuadmissionsaftermajornoncardiacsurgery
AT andrewmbishara developmentofamachinelearningderivedmodeltopredictunplannedicuadmissionsaftermajornoncardiacsurgery