Predictive modeling of 30-day readmission risk of diabetes patients by logistic regression,artificial neural network,and EasyEnsemble

作     者：Xiayu Xiang Chuanyi Liu Yanchun Zhang Wei Xiang Binxing Fang Xiayu Xiang;Chuanyi Liu;Yanchun Zhang;Wei Xiang;Binxing Fang

作者机构：School of Cyberspace SecurityBeijing University of Posts and TelecommunicationsBeijingChina School of Computer Science and TechnologyHarbin Institute of Technology(Shenzhen)ShenzhenChina Institute for Sustainable Industries and Liveable CitiesVictoria UniversityMelbourneAustralia Key Laboratory of Tropical Translational Medicine of Ministry of EducationHainan Medical UniversityHaikouChina NHC Key Laboratory of Control of Tropical DiseasesHainan Medical UniversityHaikouChina Cyberspace Security Research CenterPeng Cheng LaboratoryShenzhenChina 

出 版 物：《Asian Pacific Journal of Tropical Medicine》 (亚太热带医药杂志（英文版）)

年 卷 期：2021年第14卷第9期

页      面：417-428页

核心收录：

学科分类：1004[医学-公共卫生与预防医学(可授医学、理学学位)] 1002[医学-临床医学] 10[医学] 

基　　金：supported in part by the Key Research and Development Program for Guangdong Province(No.2019B010136001) in part by Hainan Major Science and Technology Projects(No.ZDKJ2019010) in part by the National Key Research and Development Program of China(No.2016YFB0800803 and No.2018YFB1004005) in part by National Natural Science Foundation of China(No.81960565,No.81260139,No.81060073,No.81560275,No.61562021,No.30560161 and No.61872110) in part by Hainan Special Projects of Social Development(No.ZDYF2018103 and No.2015SF 39) in part by Hainan Association for Academic Excellence Youth Science and Technology Innovation Program(No.201515) 

主　　题：Electronic health records Hospital readmissions Feature analysis Predictive models Imbalanced learning Diabetes 

摘      要：Objective:To determine the most influential data features and to develop machine learning approaches that best predict hospital readmissions among patients with diabetes.Methods:In this retrospective cohort study,we surveyed patient statistics and performed feature analysis to identify the most influential data features associated with readmissions.Classification of all-cause,30-day readmission outcomes were modeled using logistic regression,artificial neural network,and Easy Ensemble.F1 statistic,sensitivity,and positive predictive value were used to evaluate the model performance.Results:We identified 14 most influential data features(4 numeric features and 10 categorical features)and evaluated 3 machine learning models with numerous sampling methods(oversampling,undersampling,and hybrid techniques).The deep learning model offered no improvement over traditional models(logistic regression and Easy Ensemble)for predicting readmission,whereas the other two algorithms led to much smaller differences between the training and testing datasets.Conclusions:Machine learning approaches to record electronic health data offer a promising method for improving readmission prediction in patients with diabetes.But more work is needed to construct datasets with more clinical variables beyond the standard risk factors and to fine-tune and optimize machine learning models.