To develop a machine learning model for predicting in-hospital mortality risk among ICU patients with cardiovascular disease and comorbid diabetes mellitus.
Key Findings:
Patients with diabetes and cardiovascular disease have a significantly higher risk of in-hospital mortality.
Machine learning models demonstrated superior accuracy in predicting mortality risk compared to traditional methods.
The SHAP method effectively clarified the impact of individual variables on mortality risk.
Interpretation:
The study highlights the potential of machine learning in enhancing risk assessment for critically ill patients with complex comorbidities, providing a valuable tool for clinicians.
Limitations:
The study is limited to data from two specific ICU databases, which may affect generalizability.
Potential biases in data collection and missing values could impact model accuracy.
Conclusion:
The developed machine learning model serves as a promising tool for individualized risk stratification in ICU patients with cardiovascular disease and diabetes, aiding clinical decision-making.
