Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health records

Background: Asthma is a heterogeneous disease with a diverse array of phenotypes that differ in inflammatory characteristics and severity. Identifying and classifying phenotypes in the real world could provide a foundation to improve and personalize asthma management. Leveraging machine learning in...

Full description

Saved in:
Bibliographic Details
Main Authors: Mehmet Furkan Bağcı, MSc, Toan Do, MD, Samantha R. Spierling Bagsic, PhD, Rahul F. Gomez, MD, Judy H. Jun, MD, Anna L. Ritko, MA, MPhil, Sally E. Wenzel, MD, Truong Nguyen, PhD, Yusuf Öztürk, PhD, Brian D. Modena, MD, MSc
Format: Article
Language:English
Published: Elsevier 2025-08-01
Series:Journal of Allergy and Clinical Immunology: Global
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2772829325000748
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849700490133110784
author Mehmet Furkan Bağcı, MSc
Toan Do, MD
Samantha R. Spierling Bagsic, PhD
Rahul F. Gomez, MD
Judy H. Jun, MD
Anna L. Ritko, MA, MPhil
Sally E. Wenzel, MD
Truong Nguyen, PhD
Yusuf Öztürk, PhD
Brian D. Modena, MD, MSc
author_facet Mehmet Furkan Bağcı, MSc
Toan Do, MD
Samantha R. Spierling Bagsic, PhD
Rahul F. Gomez, MD
Judy H. Jun, MD
Anna L. Ritko, MA, MPhil
Sally E. Wenzel, MD
Truong Nguyen, PhD
Yusuf Öztürk, PhD
Brian D. Modena, MD, MSc
author_sort Mehmet Furkan Bağcı, MSc
collection DOAJ
description Background: Asthma is a heterogeneous disease with a diverse array of phenotypes that differ in inflammatory characteristics and severity. Identifying and classifying phenotypes in the real world could provide a foundation to improve and personalize asthma management. Leveraging machine learning in analyzing electronic health records (EHRs) provides an opportunity to identify real-world asthma phenotypes. Objective: We utilized machine-learning techniques applied to EHRs to detect and predict real-world severe asthma (SA) phenotypes and improve the precision of asthma severity diagnoses. Methods: Data from 31,795 asthma patients were extracted from a health care system’s EHR, with 1,112 patients meeting inclusion criteria for analysis. Principal component analysis (PCA) and a Gaussian mixture model classified patients into subject clusters (SCs). Asthma severity was assessed using two predictive models, one based on the American Thoracic Society (ATS) definition and the other a supervised model trained on 50 randomly selected patients whose disease severity was predetermined by 2 independent physicians. Results: Three principal components (PCs) emerged, reflecting lung function (PC1), blood inflammatory markers (PC2), and systemic corticosteroid receipt (PC3). PCA identified 5 distinct asthma phenotypes with significant clinical, physiologic, and inflammatory differences. A supervised model, trained on 50 randomly selected patients, predicted SA with 92% precision and 85% accuracy. SC3 was classified as an inflammatory, SA phenotype, making it highly suitable for biologic therapy. Conclusion: Integrating machine learning with EHRs successfully classified and identified real-world asthma phenotypes, demonstrating the potential of this approach to identify SA for appropriate management and/or clinical studies.
format Article
id doaj-art-130964f68fef464c9f800eb794182b8f
institution DOAJ
issn 2772-8293
language English
publishDate 2025-08-01
publisher Elsevier
record_format Article
series Journal of Allergy and Clinical Immunology: Global
spelling doaj-art-130964f68fef464c9f800eb794182b8f2025-08-20T03:18:15ZengElsevierJournal of Allergy and Clinical Immunology: Global2772-82932025-08-014310047310.1016/j.jacig.2025.100473Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health recordsMehmet Furkan Bağcı, MSc0Toan Do, MD1Samantha R. Spierling Bagsic, PhD2Rahul F. Gomez, MD3Judy H. Jun, MD4Anna L. Ritko, MA, MPhil5Sally E. Wenzel, MD6Truong Nguyen, PhD7Yusuf Öztürk, PhD8Brian D. Modena, MD, MSc9Department of Electrical and Computer Engineering, University of California San Diego, La Jolla, Calif; Department of Electrical and Computer Engineering, San Diego State University, San Diego, CalifDepartment of Allergy & Immunology, University of California San Diego School of Medicine, San Diego, CalifDepartment of Research Development, Scripps Health, San Diego, CalifDepartment of Knowledge Management, Scripps Health, San Diego, CalifDepartment of Knowledge Management, Scripps Health, San Diego, CalifDepartment of Internal Medicine, Scripps Health, San Diego, CalifUniversity of Pittsburgh, Pittsburgh, PaDepartment of Electrical and Computer Engineering, University of California San Diego, La Jolla, CalifDepartment of Electrical and Computer Engineering, San Diego State University, San Diego, Calif; Corresponding author: Yusuf Öztürk, PhD, Smart Health Institute, San Diego State University, 5500 Campanile Dr, San Diego, CA 92182-1309.Modena Allergy + Asthma, La Jolla, CalifBackground: Asthma is a heterogeneous disease with a diverse array of phenotypes that differ in inflammatory characteristics and severity. Identifying and classifying phenotypes in the real world could provide a foundation to improve and personalize asthma management. Leveraging machine learning in analyzing electronic health records (EHRs) provides an opportunity to identify real-world asthma phenotypes. Objective: We utilized machine-learning techniques applied to EHRs to detect and predict real-world severe asthma (SA) phenotypes and improve the precision of asthma severity diagnoses. Methods: Data from 31,795 asthma patients were extracted from a health care system’s EHR, with 1,112 patients meeting inclusion criteria for analysis. Principal component analysis (PCA) and a Gaussian mixture model classified patients into subject clusters (SCs). Asthma severity was assessed using two predictive models, one based on the American Thoracic Society (ATS) definition and the other a supervised model trained on 50 randomly selected patients whose disease severity was predetermined by 2 independent physicians. Results: Three principal components (PCs) emerged, reflecting lung function (PC1), blood inflammatory markers (PC2), and systemic corticosteroid receipt (PC3). PCA identified 5 distinct asthma phenotypes with significant clinical, physiologic, and inflammatory differences. A supervised model, trained on 50 randomly selected patients, predicted SA with 92% precision and 85% accuracy. SC3 was classified as an inflammatory, SA phenotype, making it highly suitable for biologic therapy. Conclusion: Integrating machine learning with EHRs successfully classified and identified real-world asthma phenotypes, demonstrating the potential of this approach to identify SA for appropriate management and/or clinical studies.http://www.sciencedirect.com/science/article/pii/S2772829325000748Asthmamachine learningelectronic health recordspredictive modeling
spellingShingle Mehmet Furkan Bağcı, MSc
Toan Do, MD
Samantha R. Spierling Bagsic, PhD
Rahul F. Gomez, MD
Judy H. Jun, MD
Anna L. Ritko, MA, MPhil
Sally E. Wenzel, MD
Truong Nguyen, PhD
Yusuf Öztürk, PhD
Brian D. Modena, MD, MSc
Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health records
Journal of Allergy and Clinical Immunology: Global
Asthma
machine learning
electronic health records
predictive modeling
title Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health records
title_full Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health records
title_fullStr Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health records
title_full_unstemmed Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health records
title_short Detection and prediction of real-world severe asthma phenotypes by application of machine learning to electronic health records
title_sort detection and prediction of real world severe asthma phenotypes by application of machine learning to electronic health records
topic Asthma
machine learning
electronic health records
predictive modeling
url http://www.sciencedirect.com/science/article/pii/S2772829325000748
work_keys_str_mv AT mehmetfurkanbagcımsc detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT toandomd detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT samantharspierlingbagsicphd detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT rahulfgomezmd detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT judyhjunmd detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT annalritkomamphil detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT sallyewenzelmd detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT truongnguyenphd detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT yusufozturkphd detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords
AT briandmodenamdmsc detectionandpredictionofrealworldsevereasthmaphenotypesbyapplicationofmachinelearningtoelectronichealthrecords