Development and validation of an ensemble learning risk model for sepsis after abdominal surgery

Introduction Although their importance has gained attention, the clinical applications of methods for screening patients at high risk of sepsis after abdominal surgery have been restricted. Therefore, we aimed to develop and validate models for screening patients at high risk of sepsis after abdomin...

Full description

Saved in:
Bibliographic Details
Main Authors: Xin Shu, Yujie Li, Yiziting Zhu, Zhiyong Yang, Xiang Liu, Xiaoyan Hu, Chunyong Yang, Lei Zhao, Tao Zhu, Yuwen Chen, Bin Yi
Format: Article
Language:English
Published: Termedia Publishing House 2024-06-01
Series:Archives of Medical Science
Subjects:
Online Access:https://www.archivesofmedicalscience.com/Development-and-validation-of-an-ensemble-learning-risk-model-for-sepsis-after-abdominal,189505,0,2.html
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850134952152137728
author Xin Shu
Yujie Li
Yiziting Zhu
Zhiyong Yang
Xiang Liu
Xiaoyan Hu
Chunyong Yang
Lei Zhao
Tao Zhu
Yuwen Chen
Bin Yi
author_facet Xin Shu
Yujie Li
Yiziting Zhu
Zhiyong Yang
Xiang Liu
Xiaoyan Hu
Chunyong Yang
Lei Zhao
Tao Zhu
Yuwen Chen
Bin Yi
author_sort Xin Shu
collection DOAJ
description Introduction Although their importance has gained attention, the clinical applications of methods for screening patients at high risk of sepsis after abdominal surgery have been restricted. Therefore, we aimed to develop and validate models for screening patients at high risk of sepsis after abdominal surgery based on machine learning with routine variables. Material and methods The whole dataset was composed of three representative academic hospitals in China and the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. Routine clinical variables were implemented for model development. The Boruta algorithm was applied for feature selection. Afterwards, ensemble learning and eight other conventional algorithms were used for model fitting and validation based on all features and selected features. The area under the receiver operating characteristic curves (ROC AUC), sensitivity, specificity, F1 score, accuracy, net reclassification index (NRI), integrated discrimination improvement (IDI), decision curve analysis (DCA), and calibration curves were used for model evaluation. Results A total of 955 patients undergoing abdominal surgery were finally analyzed (sepsis: 285, non-sepsis: 670). After feature selection, the ensemble learning model constructed by integrating k-Nearest Neighbor (KNN) and Support Vector Machine (SVM) yielded the ROC AUC of 0.892 (0.841–0.944) and accuracy of 85.0% on the test data, and the ROC AUC of 0.782 (0.727–0.838) and accuracy of 68.1% on the validation data, which performed best. Albumin, ASA score, neutrophil-lymphocyte ratio, age, and glucose were the top features associated with postoperative sepsis by KNN and SVM. Conclusions We developed a new and potential generalizable model to preoperatively screen patients at high risk of sepsis after abdominal surgery, with the advantages of a representative training cohort and routine variables.
format Article
id doaj-art-aafac5d6c8d549c486fb981245de6a11
institution OA Journals
issn 1734-1922
1896-9151
language English
publishDate 2024-06-01
publisher Termedia Publishing House
record_format Article
series Archives of Medical Science
spelling doaj-art-aafac5d6c8d549c486fb981245de6a112025-08-20T02:31:34ZengTermedia Publishing HouseArchives of Medical Science1734-19221896-91512024-06-0121113815210.5114/aoms/189505189505Development and validation of an ensemble learning risk model for sepsis after abdominal surgeryXin Shu0Yujie Li1Yiziting Zhu2Zhiyong Yang3Xiang Liu4Xiaoyan Hu5Chunyong Yang6Lei Zhao7Tao Zhu8Yuwen Chen9Bin Yi10Department of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaDepartment of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaDepartment of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaDepartment of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaDepartment of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaDepartment of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaDepartment of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaDepartment of Anesthesiology, Xuan Wu Hospital, Capital Medical University, ChinaDepartment of Anesthesiology, West China Hospital of Sichuan University, ChinaChongqing Institute of Green and Intelligent Technology, Chinese Academy of Science, ChinaDepartment of Anesthesiology, Southwest Hospital, Third Military Medical University, ChinaIntroduction Although their importance has gained attention, the clinical applications of methods for screening patients at high risk of sepsis after abdominal surgery have been restricted. Therefore, we aimed to develop and validate models for screening patients at high risk of sepsis after abdominal surgery based on machine learning with routine variables. Material and methods The whole dataset was composed of three representative academic hospitals in China and the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. Routine clinical variables were implemented for model development. The Boruta algorithm was applied for feature selection. Afterwards, ensemble learning and eight other conventional algorithms were used for model fitting and validation based on all features and selected features. The area under the receiver operating characteristic curves (ROC AUC), sensitivity, specificity, F1 score, accuracy, net reclassification index (NRI), integrated discrimination improvement (IDI), decision curve analysis (DCA), and calibration curves were used for model evaluation. Results A total of 955 patients undergoing abdominal surgery were finally analyzed (sepsis: 285, non-sepsis: 670). After feature selection, the ensemble learning model constructed by integrating k-Nearest Neighbor (KNN) and Support Vector Machine (SVM) yielded the ROC AUC of 0.892 (0.841–0.944) and accuracy of 85.0% on the test data, and the ROC AUC of 0.782 (0.727–0.838) and accuracy of 68.1% on the validation data, which performed best. Albumin, ASA score, neutrophil-lymphocyte ratio, age, and glucose were the top features associated with postoperative sepsis by KNN and SVM. Conclusions We developed a new and potential generalizable model to preoperatively screen patients at high risk of sepsis after abdominal surgery, with the advantages of a representative training cohort and routine variables.https://www.archivesofmedicalscience.com/Development-and-validation-of-an-ensemble-learning-risk-model-for-sepsis-after-abdominal,189505,0,2.htmlsepsismachine learningpostoperative complicationsperioperative periodrisk assessment
spellingShingle Xin Shu
Yujie Li
Yiziting Zhu
Zhiyong Yang
Xiang Liu
Xiaoyan Hu
Chunyong Yang
Lei Zhao
Tao Zhu
Yuwen Chen
Bin Yi
Development and validation of an ensemble learning risk model for sepsis after abdominal surgery
Archives of Medical Science
sepsis
machine learning
postoperative complications
perioperative period
risk assessment
title Development and validation of an ensemble learning risk model for sepsis after abdominal surgery
title_full Development and validation of an ensemble learning risk model for sepsis after abdominal surgery
title_fullStr Development and validation of an ensemble learning risk model for sepsis after abdominal surgery
title_full_unstemmed Development and validation of an ensemble learning risk model for sepsis after abdominal surgery
title_short Development and validation of an ensemble learning risk model for sepsis after abdominal surgery
title_sort development and validation of an ensemble learning risk model for sepsis after abdominal surgery
topic sepsis
machine learning
postoperative complications
perioperative period
risk assessment
url https://www.archivesofmedicalscience.com/Development-and-validation-of-an-ensemble-learning-risk-model-for-sepsis-after-abdominal,189505,0,2.html
work_keys_str_mv AT xinshu developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT yujieli developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT yizitingzhu developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT zhiyongyang developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT xiangliu developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT xiaoyanhu developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT chunyongyang developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT leizhao developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT taozhu developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT yuwenchen developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery
AT binyi developmentandvalidationofanensemblelearningriskmodelforsepsisafterabdominalsurgery