J Korean Gerontol Nurs.  2023 Feb;25(1):44-55. 10.17079/jkgn.2023.25.1.44.

The identification of high-risk groups for long-term care insurance: A retrospective study using national health insurance service database

• Song MK
• Park YW
• Han EJ

• KMID: 2540625
• DOI: http://doi.org/10.17079/jkgn.2023.25.1.44

Abstract

Purpose
This study aimed to identify high-risk groups for Long-Term Care Insurance (LTCI) in older people not approved for LTCI and to examine the characteristics of each high-risk group.
Methods
This study was a retrospective study using the National Health Insurance Service database and included 7,724,101 older A decision- tree model was used to predict the high-risk groups for LTCI. The dependent variable was defined as LTCI eligibility. As independent variables, 78 variables consisting of personal factors, environmental factors, health status, and physical and cognitive abilities were used.
Results
The prediction model to identify high-risk groups for LTCI was developed as the decision-tree model consisting of 19 end nodes with 10 risk factors. Eleven groups were identified as high-risk groups. The results showed the model could predict about 72% of the older people at high risk for LTCI needs using the NHIS database without the assessment of LTCI eligibility.
Conclusion
The findings might be useful for the development of evidence-based preventative services and can contribute to preemptively discovering those who need preventive services in older adults.

Keyword

Long-term care; Primary prevention; Algorithms; Decision trees