보험가입유형에 따른 심뇌혈관질환 발생률 및 의료이용 예측모형 개발

The incidence of cardio-cerebrovascular diseases, and the resulting socioeconomic burden, is on the rise due to aging population and worsening lifestyle habits. We aimed to estimate the occurrence of future cardio-cere brovascular diseases and their medical utilization at the individual level, and to identify key risk factors. To achieve this, we used health checkup and hospital utilization records from the Korean National Health Insurance Service Cohort DB and borrowed classification criteria from private health insurers to stratify the analysis sample into two categories based on past and current health status to fit models for each sample group. The onset of cardio-cerebrovascular diseases and mortality were modeled using penalized logistic regression models, while medical utilization predictions, such as medical expenses and outpatient days, employed a two-part model integrating classification and regression frameworks. The model with high performance was fitted to the simplified underwriting group, identifying smoking and blood pressure as major risk factors. We expect that the models developed in this study will quantitatively evaluate the societal costs incurred by individual occurrences of cardio-cerebrovascular diseases and associated medical utilizations, contributing to the formulation of national health policies and the development of private insurance products.