Cargando…

Predicting Prolonged Length of ICU Stay through Machine Learning

This study aimed to construct machine learning (ML) models for predicting prolonged length of stay (pLOS) in intensive care units (ICU) among general ICU patients. A multicenter database called eICU (Collaborative Research Database) was used for model derivation and internal validation, and the Medi...

Descripción completa

Detalles Bibliográficos
Autores principales: Wu, Jingyi, Lin, Yu, Li, Pengfei, Hu, Yonghua, Zhang, Luxia, Kong, Guilan
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8700580/
https://www.ncbi.nlm.nih.gov/pubmed/34943479
http://dx.doi.org/10.3390/diagnostics11122242
_version_ 1784620791384506368
author Wu, Jingyi
Lin, Yu
Li, Pengfei
Hu, Yonghua
Zhang, Luxia
Kong, Guilan
author_facet Wu, Jingyi
Lin, Yu
Li, Pengfei
Hu, Yonghua
Zhang, Luxia
Kong, Guilan
author_sort Wu, Jingyi
collection PubMed
description This study aimed to construct machine learning (ML) models for predicting prolonged length of stay (pLOS) in intensive care units (ICU) among general ICU patients. A multicenter database called eICU (Collaborative Research Database) was used for model derivation and internal validation, and the Medical Information Mart for Intensive Care (MIMIC) III database was used for external validation. We used four different ML methods (random forest, support vector machine, deep learning, and gradient boosting decision tree (GBDT)) to develop prediction models. The prediction performance of the four models were compared with the customized simplified acute physiology score (SAPS) II. The area under the receiver operation characteristic curve (AUROC), area under the precision-recall curve (AUPRC), estimated calibration index (ECI), and Brier score were used to measure performance. In internal validation, the GBDT model achieved the best overall performance (Brier score, 0.164), discrimination (AUROC, 0.742; AUPRC, 0.537), and calibration (ECI, 8.224). In external validation, the GBDT model also achieved the best overall performance (Brier score, 0.166), discrimination (AUROC, 0.747; AUPRC, 0.536), and calibration (ECI, 8.294). External validation showed that the calibration curve of the GBDT model was an optimal fit, and four ML models outperformed the customized SAPS II model. The GBDT-based pLOS-ICU prediction model had the best prediction performance among the five models on both internal and external datasets. Furthermore, it has the potential to assist ICU physicians to identify patients with pLOS-ICU risk and provide appropriate clinical interventions to improve patient outcomes.
format Online
Article
Text
id pubmed-8700580
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-87005802021-12-24 Predicting Prolonged Length of ICU Stay through Machine Learning Wu, Jingyi Lin, Yu Li, Pengfei Hu, Yonghua Zhang, Luxia Kong, Guilan Diagnostics (Basel) Article This study aimed to construct machine learning (ML) models for predicting prolonged length of stay (pLOS) in intensive care units (ICU) among general ICU patients. A multicenter database called eICU (Collaborative Research Database) was used for model derivation and internal validation, and the Medical Information Mart for Intensive Care (MIMIC) III database was used for external validation. We used four different ML methods (random forest, support vector machine, deep learning, and gradient boosting decision tree (GBDT)) to develop prediction models. The prediction performance of the four models were compared with the customized simplified acute physiology score (SAPS) II. The area under the receiver operation characteristic curve (AUROC), area under the precision-recall curve (AUPRC), estimated calibration index (ECI), and Brier score were used to measure performance. In internal validation, the GBDT model achieved the best overall performance (Brier score, 0.164), discrimination (AUROC, 0.742; AUPRC, 0.537), and calibration (ECI, 8.224). In external validation, the GBDT model also achieved the best overall performance (Brier score, 0.166), discrimination (AUROC, 0.747; AUPRC, 0.536), and calibration (ECI, 8.294). External validation showed that the calibration curve of the GBDT model was an optimal fit, and four ML models outperformed the customized SAPS II model. The GBDT-based pLOS-ICU prediction model had the best prediction performance among the five models on both internal and external datasets. Furthermore, it has the potential to assist ICU physicians to identify patients with pLOS-ICU risk and provide appropriate clinical interventions to improve patient outcomes. MDPI 2021-11-30 /pmc/articles/PMC8700580/ /pubmed/34943479 http://dx.doi.org/10.3390/diagnostics11122242 Text en © 2021 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Wu, Jingyi
Lin, Yu
Li, Pengfei
Hu, Yonghua
Zhang, Luxia
Kong, Guilan
Predicting Prolonged Length of ICU Stay through Machine Learning
title Predicting Prolonged Length of ICU Stay through Machine Learning
title_full Predicting Prolonged Length of ICU Stay through Machine Learning
title_fullStr Predicting Prolonged Length of ICU Stay through Machine Learning
title_full_unstemmed Predicting Prolonged Length of ICU Stay through Machine Learning
title_short Predicting Prolonged Length of ICU Stay through Machine Learning
title_sort predicting prolonged length of icu stay through machine learning
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8700580/
https://www.ncbi.nlm.nih.gov/pubmed/34943479
http://dx.doi.org/10.3390/diagnostics11122242
work_keys_str_mv AT wujingyi predictingprolongedlengthoficustaythroughmachinelearning
AT linyu predictingprolongedlengthoficustaythroughmachinelearning
AT lipengfei predictingprolongedlengthoficustaythroughmachinelearning
AT huyonghua predictingprolongedlengthoficustaythroughmachinelearning
AT zhangluxia predictingprolongedlengthoficustaythroughmachinelearning
AT kongguilan predictingprolongedlengthoficustaythroughmachinelearning