Hye-Ryun Kim1, Young-Ho Khang. 1. Korea Institute for Health and Social Affairs, Department of Preventive Medicine, University of Ulsan College of Medicine.
Abstract
OBJECTIVES: This study was done to evaluate the reliability of education and occupational class between using the health survey and the death certificate data. METHODS: The 1998 National Health and Nutrition Examination Survey (NHANES) was conducted on a cross-sectional probability sample of South Korean households, and it contained unique 13-digit personal identification numbers that were linked to the data on mortality from the Korean National Statistical Office. The data from 263 deaths were used to estimate the agreement rates and the Kappa indices of the education and occupational class between using the NHANES data and the death certificate data. RESULTS: The simple and weighted Kappa indices for education were 0.60 (95% CI=0.53-0.68) and 0.73 (95% CI=0.67-0.79) respectively, if the educational level was grouped into five categories: no-formal-education, elementary-school, middle-school, high-school and college or over. The overall agreement rate was 71.9% for these educational groups. The magnitude of reliability, as measured by the overall agreement rates and Kappa indices, tended to increase with a decrease in the educational class. The number of non-educated people with using the death certificate data was smaller than that with using the NHANES data. For the occupational class (manual workers, non-manual workers and others), the Kappa index was 0.40 (95% CI=0.30-0.51), which was relatively lower than that for the educational class. Compared with the NHANES, the number of non-manual workers for the deceased who were aged 30-64 tended to be increased (8 to 12) when using the death certificate data, whereas the number of manual workers tended to be decreased (59 to 41). CONCLUSIONS: The socioeconomic inequalities in the mortality rates that were based on the previous unlinked studies in South Korea were not due to a numerator/denominator bias. The mortality rates for the manual workers and the no-education groups might have been underestimated.
OBJECTIVES: This study was done to evaluate the reliability of education and occupational class between using the health survey and the death certificate data. METHODS: The 1998 National Health and Nutrition Examination Survey (NHANES) was conducted on a cross-sectional probability sample of South Korean households, and it contained unique 13-digit personal identification numbers that were linked to the data on mortality from the Korean National Statistical Office. The data from 263 deaths were used to estimate the agreement rates and the Kappa indices of the education and occupational class between using the NHANES data and the death certificate data. RESULTS: The simple and weighted Kappa indices for education were 0.60 (95% CI=0.53-0.68) and 0.73 (95% CI=0.67-0.79) respectively, if the educational level was grouped into five categories: no-formal-education, elementary-school, middle-school, high-school and college or over. The overall agreement rate was 71.9% for these educational groups. The magnitude of reliability, as measured by the overall agreement rates and Kappa indices, tended to increase with a decrease in the educational class. The number of non-educated people with using the death certificate data was smaller than that with using the NHANES data. For the occupational class (manual workers, non-manual workers and others), the Kappa index was 0.40 (95% CI=0.30-0.51), which was relatively lower than that for the educational class. Compared with the NHANES, the number of non-manual workers for the deceased who were aged 30-64 tended to be increased (8 to 12) when using the death certificate data, whereas the number of manual workers tended to be decreased (59 to 41). CONCLUSIONS: The socioeconomic inequalities in the mortality rates that were based on the previous unlinked studies in South Korea were not due to a numerator/denominator bias. The mortality rates for the manual workers and the no-education groups might have been underestimated.