Machine learning approaches for practical predicting outpatient near-future AECOPD based on nationwide electronic medical records
Kuang-Ming Liao,
Kuo-Chen Cheng,
Mei-I Sung,
Yu-Ting Shen,
Chong-Chi Chiu,
Chung-Feng Liu,
Shian-Chin Ko
Affiliations
Kuang-Ming Liao
Department of Internal Medicine, Chi Mei Medical Center, Chiali, Tainan 722013, Taiwan; Department of Nursing, Min-Hwei Junior College of Health Care Management, Tainan 73658, Taiwan
Kuo-Chen Cheng
Department of Pulmonary Medicine, Chi Mei Medical Center, Tainan 710402, Taiwan
Mei-I Sung
Department of Medical Research, Chi Mei Medical Center, Tainan 710402, Taiwan
Yu-Ting Shen
Department of Medical Research, Chi Mei Medical Center, Tainan 710402, Taiwan
Chong-Chi Chiu
Department of General Surgery, E-Da Cancer Hospital, I-Shou University, Kaohsiung 82445, Taiwan; School of Medicine, College of Medicine, I-Shou University, Kaohsiung 82445, Taiwan; Department of Medical Education and Research, E-Da Cancer Hospital, I-Shou University, Kaohsiung 82445, Taiwan
Chung-Feng Liu
Department of Medical Research, Chi Mei Medical Center, Tainan 710402, Taiwan; Corresponding author
Shian-Chin Ko
Department of Pulmonary Medicine, Chi Mei Medical Center, Tainan 710402, Taiwan; Corresponding author
Summary: In this research, we aimed to harness machine learning to predict the imminent risk of acute exacerbation in chronic obstructive pulmonary disease (AECOPD) patients. Utilizing retrospective data from electronic medical records of two Taiwanese hospitals, we identified 26 critical features. To predict 3- and 6-month AECOPD occurrences, we deployed five distinct machine learning algorithms alongside ensemble learning. The 3-month risk prediction was best realized by the XGBoost model, achieving an AUC of 0.795, whereas the XGBoost was superior for the 6-month prediction with an AUC of 0.813. We conducted an explainability analysis and found that the episode of AECOPD, mMRC score, CAT score, respiratory rate, and the use of inhaled corticosteroids were the most impactful features. Notably, our approach surpassed predictions that relied solely on CAT or mMRC scores. Accordingly, we designed an interactive prediction system that provides physicians with a practical tool to predict near-term AECOPD risk in outpatients.
