Machine learning for mortality prediction using nutritional and inflammatory markers in critically ill patients
A García-Grimaldo1, C A Galindo-Martín2, N C Rodriguez-Moguel3
1Departamento de Nutrición Clínica, Instituto Nacional de Enfermedades Respiratorias, Mexico City, Mexico; Sección de Estudios de Posgrado e Investigación, Escuela Superior de Medicina, Instituto Politécnico Nacional. Mexico City, Mexico.
Clinical Nutrition ESPEN
|March 26, 2026
Summary
This study developed a machine learning model to predict survival in critically ill patients by integrating clinical, nutritional, and inflammatory indicators. The model shows promise for improving patient outcomes and can be implemented even in resource-limited settings.
Area of Science:
- Critical Care Medicine
- Machine Learning in Healthcare
- Nutritional Biomarkers
Background:
- Conventional intensive care unit (ICU) mortality scores lack nutritional status assessment.
- Malnutrition and muscle loss significantly impact outcomes for critically ill patients.
Purpose of the Study:
- To develop a machine learning model for predicting in-hospital mortality in critically ill patients with respiratory diseases.
- To integrate clinical and nutritional indicators for improved mortality prediction.
Main Methods:
- An artificial neural network (NN) was trained using age, SOFA score, CRP, BMI-adjusted CC, and MUAC.
- The cohort was divided into training (70%) and test (30%) sets.
- Class imbalance was addressed using cross-validation and random oversampling.
Main Results:
- The model achieved an AUC of 0.77 in the test set.
- Key predictors included age, BMI-adjusted calf circumference, and SOFA score.
- Integrating nutritional and inflammatory indicators improved predictive performance.
Conclusions:
- A feasible, clinically relevant NN-based model for predicting survival in critically ill patients was developed.
- The model integrates nutritional and inflammatory markers using routinely collected variables.
- This tool can be implemented in resource-limited settings and warrants further evaluation for nutritional support response.