• The Korean Journal of Applied Statistics
• /
• v.30 no.3
• /
• pp.391-401
• /
• 2017

This study develops an indicator to manage regional variations of approval rates for long-term care (LTC) service. We used LTC insurance data for grade assessment that include 433,155 applicants from 227 LTC centers across Korea in 2015. The approval rate for each center was defined as the proportion of the numbers of approved applicants out of all applicants. We assumed that the approval rates depended on the characteristics of applicants. We estimated the 'standard' approval rates from a multiple linear regression analysis using the characteristics of applicants as independent variables. The difference between the observed and the standard rates was then defined as an indicator for deviation. A center having a large difference could be considered as a center with a potential error in grade assessment. We also examined if the characteristics of investigators affected the approval rates. We found that the socio-demographic characteristics of applicants and reapplication rate for LTC grade were independent factors affecting the approval rates. Centers having the management indicator values falling outside the middle 95% of the distribution were identified as centers with an error in grading. We expect that this study will contribute to enhancing reliability and equity in LTC grading.