An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients

Abstract Background Intradialytic hypotension (IDH) and intradialytic hypertension (IDHTN) are major complications of maintenance hemodialysis (MHD) that significantly impact patient morbidity and mortality. Effective, explainable prediction of IDH and IDHTN can improve their management. Methods Thi...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhuoyu Li, Siying Hao, Shujun Shi, Lin Li, Ziwei Tao
Format: Article
Language:English
Published: BMC 2025-07-01
Series:BMC Nephrology
Subjects:
Online Access:https://doi.org/10.1186/s12882-025-04270-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849335057888575488
author Zhuoyu Li
Siying Hao
Shujun Shi
Lin Li
Ziwei Tao
author_facet Zhuoyu Li
Siying Hao
Shujun Shi
Lin Li
Ziwei Tao
author_sort Zhuoyu Li
collection DOAJ
description Abstract Background Intradialytic hypotension (IDH) and intradialytic hypertension (IDHTN) are major complications of maintenance hemodialysis (MHD) that significantly impact patient morbidity and mortality. Effective, explainable prediction of IDH and IDHTN can improve their management. Methods This study introduces a dual-model system for predicting IDH and IDHTN, using SHAP (SHapley Additive exPlanations) to enhance explainability. We analyzed data from maintenance dialysis patients at the Second Hospital of Lanzhou University, covering treatments from February 2019 to August 2023. Two models were developed: Model A, with a small set of easily obtainable features, and Model B, with a comprehensive set of multidimensional features. Results The study cohort included 193 patients and 45,825 dialysis samples, with an average age of 54 years and 66.32% male. Model A used 12 features, while Model B used 51. Models were trained using XGBoost, Random Forest, logistic regression, and KNN. Random Forest achieved the highest AUROC of 0.7160 in Model A. XGBoost reached an AUROC of 0.7412 in Model B. SHAP analysis identified key predictors such as pre-dialysis blood pressure, lactate dehydrogenase, and age. Older patients (>60 years) were at higher risk for hypotension. A larger gradient between plasma sodium and dialysate sodium was associated with increased IDH risk and required more aggressive ultrafiltration. Adjusting the sodium gradient through dialysate sodium concentration may help manage IDHTN risk. Conclusions This study demonstrates that explainable AI models can predict IDH and IDHTN risks accurately before treatment, potentially reducing severe adverse events and improving patient outcomes. Clinical trial number Not applicable.
format Article
id doaj-art-0c2fd13e5bb646c2a84872d9115119f8
institution Kabale University
issn 1471-2369
language English
publishDate 2025-07-01
publisher BMC
record_format Article
series BMC Nephrology
spelling doaj-art-0c2fd13e5bb646c2a84872d9115119f82025-08-20T03:45:24ZengBMCBMC Nephrology1471-23692025-07-0126111410.1186/s12882-025-04270-5An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patientsZhuoyu Li0Siying Hao1Shujun Shi2Lin Li3Ziwei Tao4School of Information Science and Engineering, Lanzhou UniversitySchool of Information Science and Engineering, Lanzhou UniversityRenal Division, Department of Medicine, Second Hospital of Lanzhou UniversityCyberspace Administration Office, Lanzhou UniversitySchool of Computer Science and Technology, Guangxi University of Science and TechnologyAbstract Background Intradialytic hypotension (IDH) and intradialytic hypertension (IDHTN) are major complications of maintenance hemodialysis (MHD) that significantly impact patient morbidity and mortality. Effective, explainable prediction of IDH and IDHTN can improve their management. Methods This study introduces a dual-model system for predicting IDH and IDHTN, using SHAP (SHapley Additive exPlanations) to enhance explainability. We analyzed data from maintenance dialysis patients at the Second Hospital of Lanzhou University, covering treatments from February 2019 to August 2023. Two models were developed: Model A, with a small set of easily obtainable features, and Model B, with a comprehensive set of multidimensional features. Results The study cohort included 193 patients and 45,825 dialysis samples, with an average age of 54 years and 66.32% male. Model A used 12 features, while Model B used 51. Models were trained using XGBoost, Random Forest, logistic regression, and KNN. Random Forest achieved the highest AUROC of 0.7160 in Model A. XGBoost reached an AUROC of 0.7412 in Model B. SHAP analysis identified key predictors such as pre-dialysis blood pressure, lactate dehydrogenase, and age. Older patients (>60 years) were at higher risk for hypotension. A larger gradient between plasma sodium and dialysate sodium was associated with increased IDH risk and required more aggressive ultrafiltration. Adjusting the sodium gradient through dialysate sodium concentration may help manage IDHTN risk. Conclusions This study demonstrates that explainable AI models can predict IDH and IDHTN risks accurately before treatment, potentially reducing severe adverse events and improving patient outcomes. Clinical trial number Not applicable.https://doi.org/10.1186/s12882-025-04270-5HemodialysisIntradialytic hypertensionIntradialytic hypotensionMachine learningShap
spellingShingle Zhuoyu Li
Siying Hao
Shujun Shi
Lin Li
Ziwei Tao
An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients
BMC Nephrology
Hemodialysis
Intradialytic hypertension
Intradialytic hypotension
Machine learning
Shap
title An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients
title_full An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients
title_fullStr An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients
title_full_unstemmed An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients
title_short An explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients
title_sort explainable machine learning model for early warning of hypertensive and hypotensive anomalies in maintenance hemodialysis patients
topic Hemodialysis
Intradialytic hypertension
Intradialytic hypotension
Machine learning
Shap
url https://doi.org/10.1186/s12882-025-04270-5
work_keys_str_mv AT zhuoyuli anexplainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT siyinghao anexplainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT shujunshi anexplainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT linli anexplainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT ziweitao anexplainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT zhuoyuli explainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT siyinghao explainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT shujunshi explainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT linli explainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients
AT ziweitao explainablemachinelearningmodelforearlywarningofhypertensiveandhypotensiveanomaliesinmaintenancehemodialysispatients