Validation of self-reported morbidities in the Korean Atomic Bomb Survivor Cohort
- Author:
Ansun JEONG
1
;
Somin JEON
;
Seong-geun MOON
;
Mi Kyung KIM
;
Inah KIM
;
Yu-Mi KIM
;
Boyoung PARK
Author Information
- Publication Type:Methods
- From:Epidemiology and Health 2024;46(1):e2024058-
- CountryRepublic of Korea
- Language:0
-
Abstract:
OBJECTIVES:This study aimed to evaluate the agreement of disease status collected through a survey of the Korean Atomic Bomb Survivor Cohort (K-ABC), compared with medical claim records from the Korean National Health Insurance Service (NHIS) database and the Korean Central Cancer Registry (KCCR).
METHODS:Data on the lifetime physician-diagnosed morbidities of 1,215 K-ABC participants were collected through an interviewer-administered questionnaire between 2020 and 2022. Survey data were linked to the NHIS and KCCR databases. Eleven diseases were included for validation. We evaluated the following indicators: sensitivity, specificity, positive predictive value, negative predictive value, accuracy, the area under the curve, and the kappa coefficient.
RESULTS:The mean±standard deviation age was 62.1±18.7 years, and 42.6% of the participants were aged ≥70 years. Hypertension and cataracts showed the highest prevalence rates (33.8 and 28.8%, respectively). Hypertension, diabetes, and cancer demonstrated high sensitivity (>0.8) and specificity (>0.9), whereas diabetes, cancer, myocardial infarction, angina pectoris, and asthma exhibited high accuracy (>0.9). In contrast, arthritis, allergic rhinitis, and asthma showed low sensitivity (<0.4) and kappa values (<0.3). In the participants aged ≥70 years, the kappa value was ≥0.4 for all diseases except arthritis, allergic rhinitis, and asthma.
CONCLUSIONS:The results from this initial analysis showed relatively high agreement between the survey and NHIS/KCCR databases, especially for hypertension, diabetes, and cancer. Our findings suggest that the information on morbidities collected through the questionnaires in this cohort was valid for both younger and older individuals.
