Lupus & antiphospholipid syndrome
Prediction of mortality risk in critically ill patients with systemic lupus erythematosus: a machine learning approach using the MIMIC-IV database
Authors' conclusion
The machine learning model provides a superior tool for predicting mortality risk in patients with SLE
Study context
- Population
- Patients with systemic lupus erythematosus (SLE) requiring hospital admission, MIMIC-IV database; external validation set from Fujian Provincial Hospital395 MIMIC-IV, 100 validation
- Comparison
- Stacking ensemble machine learning model vs logistic regression nomogram
- Outcome
- Mortality risk prediction (AUC and calibration)
Methodological detailsobservational, Least Absolute Shrinkage and Selection Operator (LASSO) variable selection, stacking ensemble machine learning, logistic regression nomogram, external validation
- Observational
- LASSO variable selection
- Stacking ensemble machine learning
- Logistic regression nomogram
- External validation
Performance
As reported in the abstract
Discrimination for mortality
- Both models demonstrated good discrimination, with AUCs above 0.8
- Machine learning model outperformed the nomogram in precision and specificity
Further findings
Variable selection and contributors
- Significant variables identified by LASSO regression
- 18
- Important factors (SHAP): urine output, age, weight and alanine aminotransferase
Original publication
Prediction of mortality risk in critically ill patients with systemic lupus erythematosus: a machine learning approach using the MIMIC-IV database. Lupus Sci Med. Published 13 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →