An explainable online frailty prediction model for community-dwelling older adults based on machine learning algorithms: a cross-sectional study based on retrospective health data - PubMed
5 hours ago
- #Frailty Prediction
- #Machine Learning
- #Older Adults Health
- An explainable online frailty prediction model was developed for community-dwelling older adults using machine learning algorithms.
- The study is a cross-sectional analysis based on retrospective health data from 1,156 older adults in 31 community health centers in Nanjing, collected between January and October 2024.
- Key predictors of frailty identified include age, education, medicine use, vegetable consumption, cognitive status, number of diseases, hemoglobin, total cholesterol, and neutrophil-to-lymphocyte ratio.
- Among six machine learning models tested, the CatBoost model showed the highest performance with an AUROC of 0.886 in training and 0.831 in testing sets.
- A web-based risk calculator was created using the SHAP method for model interpretation, aimed at helping general practitioners identify high-risk individuals for timely interventions to slow frailty progression.