Email updates

Keep up to date with the latest news and content from BMC Public Health and BioMed Central.

Open Access Research article

Consistency between education reported in health survey and recorded in death certificate

Young-Ho Khang1*, Hye Ryun Kim2 and John W Lynch3

Author Affiliations

1 Department of Preventive Medicine, University of Ulsan College of Medicine, Seoul, Korea

2 Korea Institute for Health and Social Affairs, Seoul, Korea

3 Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, Canada

For all author emails, please log on.

BMC Public Health 2007, 7:294  doi:10.1186/1471-2458-7-294

Published: 18 October 2007



Education level is one indicator of socioeconomic position which, in several countries including South Korea, is provided though death certificate data. Its validity determines the usefulness of death certificate data for exploring the association between socioeconomic position and mortality. This study was to compare education recorded on the death certificate with that reported before death in a nationally representative cohort of participants in the National Health and Nutrition Examination Survey (NHANES).


The 1998/2001 NHANES data contained unique 13-digit personal identification numbers that were individually linked to death certificate data from the Korean National Statistical Office. Duration of mortality follow-up was 7.1 years. The data from 513 deaths were used to determine sensitivity and specificity of education in death certificate and estimate agreement rates of education level between NHANES data and death certificate data. Odds ratios for agreement in education were also estimated. Covariates considered in the analyses were gender, age, duration between NHANES and death, and cause of death.


The proportion of deaths without recorded education in death certificate was very low (0.2%). A total of 29.4% discordant pairs were found. Sensitivity and specificity for college or higher education were 0.84 (95% confidence interval 0.71–0.97) and 0.99 (0.98–1.00). However, sensitivity was poor for middle school education. The overall agreement rate was 70.7% (66.8%–74.6%) when education was categorized into five groups and increased up to 88.9% (86.2%–91.6%) when three education categories were used. The magnitude of validity and reliability for education did not generally vary with age, duration between health survey and death, and cause of death. However, a significantly smaller likelihood of agreement was found for middle and elementary school education after adjusting for covariates.


Low percentage of missing information on education in South Korean death certificate data could provide a great potential to monitor mortality inequalities. A more collapsed categorization in education would be recommended when a more definitive conclusion on educational mortality inequality is required.