Development and validation of predictive models for the hemodialysis adequacy index (Kt/V) using multiple linear regression and machine learning algorithms
8 affiliations
- 1Department of Nephrology, The Affiliated Hospital of Inner Mongolia Medical University, Hohhot, China
- 2Department of Nephrology, Hohhot First Hospital, Hohhot, China
- 3Department of Information Engineering, Ordos Vocational College, Ordos, China
- 4Hemodialysis Room, Togtoh County Hospital, Hohhot, China
- 5Department of Nephrology, Inner Mongolia Autonomous Region Hospital of Traditional Chinese Medicine, Hohhot, China
- 6The First Department of Specialty Medicine, Inner Mongolia Corps Hospital of the Chinese People’s Armed Police Force, Hohhot, China
- 7Department of Internal Medicine, The First Hospital of Inner Mongolia Autonomous Region Prison Administration, Hohhot, China
- 8Department of Nephrology, Inner Mongolia People’s Hospital, Hohhot, China
Abstract
Background and objective: Kt/V is a widely used indicator for assessing hemodialysis (HD) adequacy, but conventional measurement methods require blood sampling and are impractical for large-scale or real-time use. This study aimed to develop and validate predictive models for Kt/V using multiple linear regression (MLR) and machine learning algorithms, enabling clinicians to estimate dialysis adequacy before a session begins using simple, non-invasive parameters.
Methods: A total of 1,017 eligible patients were enrolled from seven HD centers in Hohhot City, China, and divided into training (n = 677), internal (n = 220), and external validation (n = 120) sets. From 26 candidate variables, eight clinically meaningful derived features were constructed, and a multi-algorithm fusion strategy integrating F test, Mutual Information, Random Forest importance, XGBoost importance, LASSO coefficients, Permutation Importance, Recursive Feature Elimination, and Boruta was employed for feature selection. Five final features (body surface area, pre-dialysis weight, standardized blood flow, sex, and circulating blood volume) were selected. Eight point-prediction models (MLR, XGBoost, LightGBM, Random Forest, Gradient Boosting, KNN, SVM, and BPNN), a simple average ensemble of the three top-performing models, and two probabilistic prediction methods (LightGBM Quantile Regression and NGBoost) were developed. Hyperparameter tuning was performed using GridSearchCV with 5-fold cross-validation. Model performance was evaluated using R2, RMSE, and MAE on internal and external validation sets.
Results: On external validation, the SVM model achieved the highest R2 (0.5931) and lowest RMSE (0.1633) and MAE (0.1214) among individual models, followed by Gradient Boosting (R2 = 0.5748, RMSE = 0.1670, MAE = 0.1276) and XGBoost (R2 = 0.5681, RMSE = 0.1683, MAE = 0.1288). The ensemble model achieved the best overall performance (R2 = 0.5911, RMSE = 0.1637, MAE = 0.1240). Calibration assessment revealed all models were well-calibrated without requiring post-hoc adjustment. For probabilistic prediction, NGBoost provided well-calibrated 80% prediction intervals (PICP = 82.27%) with narrower intervals than LightGBM Quantile Regression. SHAP analysis identified BSA, standardized blood flow, and gender as the most influential predictors.
Conclusion: Machine learning models, particularly SVM, the ensemble model, and NGBoost, can provide accurate, interpretable, and uncertainty-aware predictions of Kt/V using non-invasive, readily available parameters. The development of a local standalone executable application further facilitates clinical translation, enabling pre-dialysis prediction and real-time prescription adjustment. Prospective validation in diverse populations is warranted to confirm generalizability.
Related
- Construction and validation of an identification model for frailty in maintenance hemodialysis patients using multiple machine learning algorithms: a cross-sectional study of retrospectively collected data
- Development and validation of a nomogram-based predictive model for infectious diseases of the spine utilizing peripheral blood laboratory parameters
- Development and Validation of a Predictive Model for 6-Month All-Cause Mortality in Pulmonary Embolism.
- A Scalable Predictive Modeling Approach to Duplicate Detection in Pharmacovigilance: Algorithm Development and Validation Study
- Development and interpretable validation of a machine-learning model for edentulism risk in US adults : A cross-sectional study integrating five algorithms and dietary data from NHANES 2009–2014
BibTeX
@article{li2026development,
title = {{Development and validation of predictive models for the hemodialysis adequacy index (Kt/V) using multiple linear regression and machine learning algorithms}},
author = {Le-le Li and Rui-feng Xu and Liang-liang Huang and Bing-ru Lu and Shu-ting Zhang and Ya Tuo and Hui-qing Liu and Rui Ren and Shu-fang Wang and Ming-xia Sun and Xiao-tong Zhang and Jian-rong Zhao and Lei Yu and Yan Meng},
journal = {Frontiers in Digital Health},
year = {2026},
month = oct,
doi = {10.3389/fdgth.2026.1771806},
url = {https://doi.org/10.3389/fdgth.2026.1771806},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Development and validation of predictive models for the hemodialysis adequacy index (Kt/V) using multiple linear regression and machine learning algorithms
AU - Le-le Li
AU - Rui-feng Xu
AU - Liang-liang Huang
AU - Bing-ru Lu
AU - Shu-ting Zhang
AU - Ya Tuo
AU - Hui-qing Liu
AU - Rui Ren
AU - Shu-fang Wang
AU - Ming-xia Sun
AU - Xiao-tong Zhang
AU - Jian-rong Zhao
AU - Lei Yu
AU - Yan Meng
PY - 2026
DA - 2026/10/12
JO - Frontiers in Digital Health
DO - 10.3389/fdgth.2026.1771806
UR - https://doi.org/10.3389/fdgth.2026.1771806
AB - Background and objective: Kt/V is a widely used indicator for assessing hemodialysis (HD) adequacy, but conventional measurement methods require blood sampling and are impractical for large-scale or real-time use. This study aimed to develop and validate predictive models for Kt/V using multiple linear regression (MLR) and machine learning algorithms, enabling clinicians to estimate dialysis adequacy before a session begins using simple, non-invasive parameters. Methods: A total of 1,017 eligible patients were enrolled from seven HD centers in Hohhot City, China, and divided into training (n = 677), internal (n = 220), and external validation (n = 120) sets. From 26 candidate variables, eight clinically meaningful derived features were constructed, and a multi-algorithm fusion strategy integrating F test, Mutual Information, Random Forest importance, XGBoost importance, LASSO coefficients, Permutation Importance, Recursive Feature Elimination, and Boruta was employed for feature selection. Five final features (body surface area, pre-dialysis weight, standardized blood flow, sex, and circulating blood volume) were selected. Eight point-prediction models (MLR, XGBoost, LightGBM, Random Forest, Gradient Boosting, KNN, SVM, and BPNN), a simple average ensemble of the three top-performing models, and two probabilistic prediction methods (LightGBM Quantile Regression and NGBoost) were developed. Hyperparameter tuning was performed using GridSearchCV with 5-fold cross-validation. Model performance was evaluated using R2, RMSE, and MAE on internal and external validation sets. Results: On external validation, the SVM model achieved the highest R2 (0.5931) and lowest RMSE (0.1633) and MAE (0.1214) among individual models, followed by Gradient Boosting (R2 = 0.5748, RMSE = 0.1670, MAE = 0.1276) and XGBoost (R2 = 0.5681, RMSE = 0.1683, MAE = 0.1288). The ensemble model achieved the best overall performance (R2 = 0.5911, RMSE = 0.1637, MAE = 0.1240). Calibration assessment revealed all models were well-calibrated without requiring post-hoc adjustment. For probabilistic prediction, NGBoost provided well-calibrated 80% prediction intervals (PICP = 82.27%) with narrower intervals than LightGBM Quantile Regression. SHAP analysis identified BSA, standardized blood flow, and gender as the most influential predictors. Conclusion: Machine learning models, particularly SVM, the ensemble model, and NGBoost, can provide accurate, interpretable, and uncertainty-aware predictions of Kt/V using non-invasive, readily available parameters. The development of a local standalone executable application further facilitates clinical translation, enabling pre-dialysis prediction and real-time prescription adjustment. Prospective validation in diverse populations is warranted to confirm generalizability.
SN - 2673-253X
ER -