Family tree database of the National Health Information Database in Korea / 한국역학회지

ABSTRACT

We constructed the family tree database (DB) by using a new family code system that can logically express interpersonal family relationships and by comparing and complementing health insurance eligibility data and resident register data of the National Health Information Database (NHID). In the family tree DB, Parents and grandparents are matched for more than 95% of those who were born between 2010 and 2017. Codes for inverse relationships and extended relationships are generated using sequences of the three-digit basic family codes. The family tree DB contains variables such as sex, birth year, family relations, and degree of kinship (maximum of 4) between subjects and family members. Using the family tree DB, we find that prevalence rates of hypertension, diabetes, ischemic heart disease, cerebrovascular disease, and cancer are higher for those with family history. The family tree DB may omit some relationships due to incomplete past data, and some family relations cannot be uniquely determined because the source data only contain relationships between head and members of the household. The family tree DB is a part of the NHID, and researchers can submit requests for data on the website at http//nhiss.nhis.or.kr. Requested data will be provided after approval from the data service review board. However, the family tree DB can be limitedly provided for studies with high public value in order to maximize personal information protection.