Key result
The XGBoost model achieved the highest predictive performance for in-hospital mortality among ICU patients with heart failure (AUC 0.824; 95% CI 0.7766-0.8708).
Why the study?
Heart failure mortality prediction is critical for individualized care, but most prediction models lack interpretability and have not reached clinical practice.
Does an interpretable XGBoost machine learning model accurately predict in-hospital mortality in ICU patients with heart failure?
Cohort (n=2,798)
Does an interpretable XGBoost machine learning model accurately predict in-hospital mortality in ICU patients with heart failure?
Effect estimate: AUC 0.824 (95% CI 0.7766-0.8708)
An interpretable XGBoost machine learning model accurately predicts in-hospital mortality in ICU patients with heart failure, with blood urea nitrogen identified as the most important predictor.
Supports interpretable ICU HF mortality models; leaves open whether they improve outcomes or reach practice.
BACKGROUND Heart failure (HF) is a common disease and a major public health problem. The HF mortality prediction is critical for developing individualized prevention and treatment plans. However, due to their lack of interpretability, most HF mortality prediction models have not yet reached clinical practice. OBJECTIVE We aimed to develop an interpretable model to predict the risk mortality for HF patients in intensive care unit (ICU) and use the Shapley additive explanation (SHAP) method to explain the extreme gradient boosting (XGBoost) model and explore prognostic factors for HF. METHODS In this retrospective cohort study, we achieved model development and performance comparison on the eICU Collaborative Research Database (eICU-CRD). We extracted data during the first 24 hours of each ICU admission and the data set was randomly divided, with 70% used for model training and 30% for model validation. The prediction performance of the XGBoost model was compared with three other machine learning models by the areas under the receiver operating characteristic curve (AUROC). Moreover, we used the Shapley additive explanation SHAP method to explain the XGBoost model. RESULTS A total of 2,798 eligible patients with HF were included in the final cohort for this study. The observed in-hospital mortality of patients with HF was 9.97%. Comparatively, the XGBoost model had the highest predictive performance among four models (AUC 0.824, 95% Confidence Interval (CI) 0.7766 to 0.8708), while support vector machine (SVM) had the poorest generalization ability (AUC 0.701, 95% CI 0.6433 to 0.7582). The decision curve showed that the net benefit of the XGBoost model surpassed those of other machine learning models at 10%~28% threshold probabilities. The SHAP method reveals the top 20 predictors of HF according to the importance ranking and the average of the blood urea nitrogen was recognized as the most important predictor variable. CONCLUSIONS The interpretable predictive model helps physicians more accurately predict the mortality risk in ICU Patients with HF. This will help physicians to provide better treatment plan and optimal resource allocation for their patients. In addition, the interpretable framework can increase the transparency of the model and facilitates physicians to understand the reliability of the predictive model.
No takes yet. Share an insight, caveat, or question.
Li et al. (2022) conducted a cohort in Heart failure (n=2,798). XGBoost model vs. Support vector machine and other machine learning models was evaluated on In-hospital mortality prediction (AUROC) (AUC 0.824, 95% CI 0.7766-0.8708). The XGBoost model achieved the highest predictive performance for in-hospital mortality among ICU patients with heart failure (AUC 0.824; 95% CI 0.7766-0.8708).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: