Related Experiment Video
Updated: Feb 10, 2026
Enema of Traditional Chinese Medicine for Patients with Severe Acute Pancreatitis
Published on: January 27, 2023
Explainable machine learning model to predict refeeding syndrome in patients with severe acute pancreatitis
Cui Wu1, Shuangshuang Jing2,3, Dinghui Guo2
1Department of Nursing, Xuzhou Tongshan District Hospital of Traditional Chinese Medicine, Xuzhou, Jiangsu, China.
Insights
This study developed a Gradient Boosting Machine model to predict refeeding syndrome (RFS) risk in severe acute pancreatitis (SAP) patients. The model identifies high-risk individuals for timely enteral nutrition intervention.
Area of Science:
- Medical Science
- Clinical Research
- Health Informatics
Background:
- Refeeding syndrome (RFS) poses a significant risk to patients with severe acute pancreatitis (SAP).
- Early identification of patients at risk for RFS is crucial for timely intervention and improved outcomes.
- Current methods for RFS risk assessment in SAP patients require enhancement for precision and timeliness.
Purpose of the Study:
- To construct and validate a predictive risk model for refeeding syndrome (RFS) in patients with severe acute pancreatitis (SAP).
- To identify high-risk individuals susceptible to RFS before overt electrolyte abnormalities manifest.
- To provide decision support for optimizing the timing of enteral nutrition initiation and personalized interventions.
Main Methods:
- A retrospective cohort study involving SAP patients from September 2018 to September 2025.
- Utilized least absolute shrinkage and selection operator (LASSO) regression for feature selection and six machine learning (ML) algorithms for model development.
- Evaluated model performance using receiver operating characteristic (ROC) curves, calibration curves, decision curves, and SHapley Additive exPlanations (SHAP) analysis.
Main Results:
- Seven predictive features were identified, with serum potassium, serum sodium, and serum calcium being key indicators.
- The Gradient Boosting Machine (GBM) model demonstrated strong predictive performance (AUC training: 0.851, AUC testing: 0.762).
- SHAP analysis confirmed feature importance, including gastrointestinal decompression, BUN, DM history, and diuretic use.
Conclusions:
- The developed GBM model effectively predicts the risk of refeeding syndrome in SAP patients.
- This tool can aid clinicians in making informed decisions regarding enteral nutrition initiation.
- Personalized interventions based on predicted RFS risk can improve patient management and outcomes.
Objective:
To construct and validate a risk prediction model for refeeding syndrome (RFS) in patients with severe acute pancreatitis (SAP), identify high-risk individuals before overt electrolyte abnormalities occur, and provide decision support for the timing of enteral nutrition initiation and early personalized intervention.
Methods:
A retrospective cohort study was conducted on SAP patients admitted to Xuzhou Medical University Affiliated Hospital (XYFY) between September 2018 and September 2025. Patients were divided into RFS and Non-RFS groups based on the development of RFS after initiating enteral nutrition. Clinical data differences between groups were compared, least absolute shrinkage and selection operator (LASSO) regression was used for feature selection, and six machine learning (ML) algorithms were applied to build prediction models. Model performance was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curves. SHapley Additive exPlanations (SHAP) analysis was performed to interpret the contribution of key features.
Results:
Seven predictive features were identified for model construction. The gradient boosting machine (GBM) model exhibited good generalization ability, with area under the curve (AUC) values of 0.851 (95% CI: 0.809-0.894) in the training set and 0.762 (95% CI: 0.672-0.852) in the testing set. Calibration curves confirmed consistency between predicted probabilities and actual outcomes, while decision curves demonstrated favorable net benefits across different clinical decision thresholds. SHAP analysis ranked feature importance as follows: serum potassium (K), serum sodium (Na), serum calcium (Ca), gastrointestinal decompression, blood urea nitrogen (BUN), diabetes mellitus (DM) history, and diuretic use.
Conclusion:
The GBM model effectively predicts RFS risk in SAP patients after initiating enteral nutrition.
Related Concept Videos
01:27Acute Pancreatitis I: Introduction
Acute pancreatitis is characterized by rapid inflammation of the pancreas, often caused by factors like gallstone blockage or excessive alcohol consumption. Chronic pancreatitis, on the other hand, is a slow, progressive inflammation that may result from long-term alcohol abuse, obstructions in the pancreatic duct, or genetic factors.
The causes of acute pancreatitis include:
01:30Acute Coronary Syndrome I: Introduction
01:26Acute Coronary Syndrome V: Nursing Management
01:30Acute Pancreatitis II: Clinical Manifestations and Management
01:30Acute Coronary Syndrome III: Diagnostic Studies
01:28Acute Coronary Syndrome IV: Interprofessional Care