Data Issues and Suggestions in the National Health Insurance Service-National Sample Cohort for Assessing the Long-term Health Effects of Air Pollution Focusing on Mortality

Abstract

Objectives: National-scale cohort studies provided evidence of the association between long-term exposure to air pollution and mortality. Recently, South Korean government created the National Health Insurance Service-National Sample Cohort (NHIS-NSC) which included about one million subjects in 2002 and followed up every year through 2013. We aimed to explore major data issues of NHIS-NSC and to provide plausible suggestions in data processing for cohort studies of long-term air pollution and mortality. Methods: We specified four data issues: tracking failure, address inconsistency, limited address information, and absence of risk factors. Our exploratory analyses clarified data characteristics related to four issues and led us to provide the best options out of several suggestions in data processing to aovid inaccurate risk estimates in subsequent health analyses. Results: Twelve percent of the cohort was lost to follow up for at least one year. We proposed to restrict the study population to cohort members with more than 50% of follow-ups. Although the NHIS-NSC report mentioned some address changes from workplaces to residences in 2005, our exploratory analysis provided little evidence of the change. We concluded that it is not suggestive to consider the change in data processing. The incomplete address information limited to the district level prevents us from estimating individual exposures to air pollution. However, we could compute population-represenative district-level averages by using exposure prediction models. Important individual-level risk factors such as smoking were not available in NHIS-NSC. Our suggestion was to use area-level information from external data or to restrict the study population to a subset who carried out national health examinations. Conclusions: Our extensive investigation of the NHIS-NSC data and constructive suggestions of data handling focusing on cohort studies of air pollution and mortality will help assess the association between long-term exposures to air pollution and health in South Korea. issues and suggestions in the national health insurance service-national sample cohort for assessing the long-term health effects of air pollution focusing on mortality. It is identical to the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) whichpermit sunrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.