What does this research mean for the field?

A stacked ensemble machine learning model integrating 12 clinical variables can accurately predict unplanned continuous renal replacement therapy (CRRT) interruptions in critically ill burn patients. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

Develop and validate an interpretable stacked ensemble machine learning model to predict unplanned continuous renal replacement therapy interruptions in critically ill burn patients.

September 14, 2026Journal of Burn Care & ResearchOpen Access

Predicting unplanned CRRT interruptions in critically Ill burn patients using stacked ensemble machine learning framework

View Full Paper

Ask AI

Bookmark

Authors

TLTiantian Li HLHu Liu ZCZhigang Chu

Discussion

Member takes

Overview

Retrospective cohort study demonstrates high accuracy in predicting unplanned CRRT interruptions in critically ill burn patients, suggesting utility for early preventive clinical surveillance.

Key Points

Develop and validate an interpretable stacked ensemble machine learning model to predict unplanned continuous renal replacement therapy interruptions in critically ill burn patients.
Analyzed a retrospective cohort of 666 critically ill burn patients undergoing continuous renal replacement therapy, partitioned into a training cohort (n=560, 2015–2023) and a temporal validation cohort (n=106, 2024–2026).
Selected 12 independent predictors via multivariate logistic regression and evaluated 47 machine learning algorithms to build a stacked ensemble model integrating bayesglm, fda, knn, and naive_bayes base learners with two-level SHAP interpretability.
In the temporal validation cohort, the stacked ensemble model achieved an AUC of 0.963, accuracy of 0.943, sensitivity of 1.000, specificity of 0.896, and a Kappa of 0.887, with satisfactory calibration and clinical net benefit.
Two-level SHAP analysis identified blood flow rate, anticoagulation strategy, hematocrit, and filtration fraction as the most substantial contributors to predicting unplanned therapy interruption.

Cite This Study

Li et al. (2026) studied this question.

synapsesocial.com/papers/6aa7b3bf0926e14a848b2e5f https://doi.org/10.1093/jbcr/irag163

View Full Paper

Ask AI

Bookmark

September 14, 2026Journal of Burn Care & ResearchOpen Access

Predicting unplanned CRRT interruptions in critically Ill burn patients using stacked ensemble machine learning framework

View Full Paper

Ask AI

Bookmark

Authors

TLTiantian Li HLHu Liu ZCZhigang Chu

Discussion

Member takes

Overview

Retrospective cohort study demonstrates high accuracy in predicting unplanned CRRT interruptions in critically ill burn patients, suggesting utility for early preventive clinical surveillance.

Key Points

Develop and validate an interpretable stacked ensemble machine learning model to predict unplanned continuous renal replacement therapy interruptions in critically ill burn patients.
Analyzed a retrospective cohort of 666 critically ill burn patients undergoing continuous renal replacement therapy, partitioned into a training cohort (n=560, 2015–2023) and a temporal validation cohort (n=106, 2024–2026).
Selected 12 independent predictors via multivariate logistic regression and evaluated 47 machine learning algorithms to build a stacked ensemble model integrating bayesglm, fda, knn, and naive_bayes base learners with two-level SHAP interpretability.
In the temporal validation cohort, the stacked ensemble model achieved an AUC of 0.963, accuracy of 0.943, sensitivity of 1.000, specificity of 0.896, and a Kappa of 0.887, with satisfactory calibration and clinical net benefit.
Two-level SHAP analysis identified blood flow rate, anticoagulation strategy, hematocrit, and filtration fraction as the most substantial contributors to predicting unplanned therapy interruption.

Cite This Study

Li et al. (2026) studied this question.

synapsesocial.com/papers/6aa7b3bf0926e14a848b2e5f https://doi.org/10.1093/jbcr/irag163

View Full Paper

Ask AI

Bookmark