How machine learning on real world clinical data improves adverse event recording for endoscopy

Abstract Endoscopic interventions are essential for diagnosing and treating gastrointestinal conditions. Accurate and comprehensive documentation is crucial for enhancing patient safety and optimizing clinical outcomes; however, adverse events remain underreported. This study evaluates a machine lea...

Full description

Saved in:
Bibliographic Details
Main Authors: Stefan Wittlinger, Isabella C. Wiest, Mahboubeh Jannesari Ladani, Jakob Nikolas Kather, Matthias P. Ebert, Fabian Siegel, Sebastian Belle
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:npj Digital Medicine
Online Access:https://doi.org/10.1038/s41746-025-01826-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849342083391815680
author Stefan Wittlinger
Isabella C. Wiest
Mahboubeh Jannesari Ladani
Jakob Nikolas Kather
Matthias P. Ebert
Fabian Siegel
Sebastian Belle
author_facet Stefan Wittlinger
Isabella C. Wiest
Mahboubeh Jannesari Ladani
Jakob Nikolas Kather
Matthias P. Ebert
Fabian Siegel
Sebastian Belle
author_sort Stefan Wittlinger
collection DOAJ
description Abstract Endoscopic interventions are essential for diagnosing and treating gastrointestinal conditions. Accurate and comprehensive documentation is crucial for enhancing patient safety and optimizing clinical outcomes; however, adverse events remain underreported. This study evaluates a machine learning-based approach for systematically detecting endoscopic adverse events from real-world clinical metadata, including structured hospital data such as ICD-codes and procedure timings. Using a random forest classifier detecting adverse events perforation, bleeding, and readmission, we analysed 2490 inpatient cases, achieving significant improvements over baseline prediction accuracy. The model achieved AUC-ROC/AUC-PR values of 0.9/0.69 for perforation, 0.84/0.64 for bleeding, and 0.96/0.9 for readmissions. Results highlight the importance of multiple metadata features for robust predictions. This semi-automated method offers a privacy-preserving tool for identifying documentation discrepancies and enhancing quality control. By integrating metadata analysis, this approach supports better clinical decision-making, quality improvement initiatives, and resource allocation while reducing the risk of missed adverse events in endoscopy.
format Article
id doaj-art-845f83beb32045dc88147f533e32529f
institution Kabale University
issn 2398-6352
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series npj Digital Medicine
spelling doaj-art-845f83beb32045dc88147f533e32529f2025-08-20T03:43:30ZengNature Portfolionpj Digital Medicine2398-63522025-07-018111010.1038/s41746-025-01826-5How machine learning on real world clinical data improves adverse event recording for endoscopyStefan Wittlinger0Isabella C. Wiest1Mahboubeh Jannesari Ladani2Jakob Nikolas Kather3Matthias P. Ebert4Fabian Siegel5Sebastian Belle6Department of Medicine II, University Medical Center Mannheim, Medical Faculty Mannheim, Heidelberg UniversityDepartment of Medicine II, University Medical Center Mannheim, Medical Faculty Mannheim, Heidelberg UniversityDepartment of Biomedical Informatics, Mannheim Institute for intelligent Systems in Medicine (MIISM), Medical Faculty Mannheim, Heidelberg UniversityElse Kroener Fresenius Center for Digital Health, Faculty of Medicine and University Hospital Carl Gustav Carus, TUD Dresden University of TechnologyDepartment of Medicine II, University Medical Center Mannheim, Medical Faculty Mannheim, Heidelberg UniversityDepartment of Biomedical Informatics, Mannheim Institute for intelligent Systems in Medicine (MIISM), Medical Faculty Mannheim, Heidelberg UniversityDepartment of Medicine II, University Medical Center Mannheim, Medical Faculty Mannheim, Heidelberg UniversityAbstract Endoscopic interventions are essential for diagnosing and treating gastrointestinal conditions. Accurate and comprehensive documentation is crucial for enhancing patient safety and optimizing clinical outcomes; however, adverse events remain underreported. This study evaluates a machine learning-based approach for systematically detecting endoscopic adverse events from real-world clinical metadata, including structured hospital data such as ICD-codes and procedure timings. Using a random forest classifier detecting adverse events perforation, bleeding, and readmission, we analysed 2490 inpatient cases, achieving significant improvements over baseline prediction accuracy. The model achieved AUC-ROC/AUC-PR values of 0.9/0.69 for perforation, 0.84/0.64 for bleeding, and 0.96/0.9 for readmissions. Results highlight the importance of multiple metadata features for robust predictions. This semi-automated method offers a privacy-preserving tool for identifying documentation discrepancies and enhancing quality control. By integrating metadata analysis, this approach supports better clinical decision-making, quality improvement initiatives, and resource allocation while reducing the risk of missed adverse events in endoscopy.https://doi.org/10.1038/s41746-025-01826-5
spellingShingle Stefan Wittlinger
Isabella C. Wiest
Mahboubeh Jannesari Ladani
Jakob Nikolas Kather
Matthias P. Ebert
Fabian Siegel
Sebastian Belle
How machine learning on real world clinical data improves adverse event recording for endoscopy
npj Digital Medicine
title How machine learning on real world clinical data improves adverse event recording for endoscopy
title_full How machine learning on real world clinical data improves adverse event recording for endoscopy
title_fullStr How machine learning on real world clinical data improves adverse event recording for endoscopy
title_full_unstemmed How machine learning on real world clinical data improves adverse event recording for endoscopy
title_short How machine learning on real world clinical data improves adverse event recording for endoscopy
title_sort how machine learning on real world clinical data improves adverse event recording for endoscopy
url https://doi.org/10.1038/s41746-025-01826-5
work_keys_str_mv AT stefanwittlinger howmachinelearningonrealworldclinicaldataimprovesadverseeventrecordingforendoscopy
AT isabellacwiest howmachinelearningonrealworldclinicaldataimprovesadverseeventrecordingforendoscopy
AT mahboubehjannesariladani howmachinelearningonrealworldclinicaldataimprovesadverseeventrecordingforendoscopy
AT jakobnikolaskather howmachinelearningonrealworldclinicaldataimprovesadverseeventrecordingforendoscopy
AT matthiaspebert howmachinelearningonrealworldclinicaldataimprovesadverseeventrecordingforendoscopy
AT fabiansiegel howmachinelearningonrealworldclinicaldataimprovesadverseeventrecordingforendoscopy
AT sebastianbelle howmachinelearningonrealworldclinicaldataimprovesadverseeventrecordingforendoscopy