Long-Term Survival Prediction Model for Elderly Community Members Using a Deep Learning Method

Abstract

In an aging society, maintaining healthy aging, preventing death, and enabling a continuation of economic activities are crucial. This study sought to develop a model for predicting survival times among community-dwelling older individuals using a deep learning method, and to identify the level of influence of various risk factors on the survival period, so that older individuals can manage their own health. This study used the Korean National Health Insurance Service claims data. We observed community-dwelling older people, aged 66 years, for 11 years and developed a survival time prediction model. Of the 189,697 individuals enrolled at baseline, 180,235 (95.0%) survived from 2009 to 2019, while 9462 (5.0%) died. Using deep-learning-based models (C statistics = 0.7011), we identified various factors impacting survival: Charlson’s comorbidity index; the frailty index; long-term care benefit grade; disability grade; income level; a combination of diabetes mellitus, hypertension, and dyslipidemia; sex; smoking status; and alcohol consumption habits. In particular, Charlson’s comorbidity index (SHAP value: 0.0445) and frailty index (SHAP value: 0.0443) were strong predictors of survival time. Prediction models may help researchers to identify potentially modifiable risk factors that may affect survival.

Keywords

• community-dwelling older individuals
• comorbidity
• deep learning
• frailty
• survival prediction model