Validation of the ‘In-Hospital Death’ Classification in the Treatment Results Code of the Korean Health Insurance Review and Assessment Service Database
Year: 2025
Type: article
Study design: Observational Study
Abstract: Background: In Korea, mortality of maintenance hemodialysis (HD) patients is often estimated from the Health Insurance Review and Assessment Service (HIRA) database using operational definitions such as absence of claims for certain periods.However, some have suggested that the 'in-hospital death' designation within the treatment results code may serve as a direct indicator of mortality.We aimed to validate the accuracy of this classification code. Material and Methods:We analyzed patients aged >18 years on maintenance HD in the Korean National Health Insurance Service (NHIS) database between 2010 and 2019.Group A (reference) included patients confirmed deceased in the NHIS certificate database.Group B included patients labeled 'in-hospital death' by the treatment results code.Group C included patients presumed deceased by the operational definition (absence of claims for >150 days).Accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were analyzed.Results: A total of 118,386 patients (59.4% male, mean age 64.1 ± 13.7 years) were included.Among them, 51,146 patients (43.2%) were confirmed deceased (Group A).Group B comprised 39,447 patients with "in-hospital death," and Group C comprised 51,758 patients presumed deceased by the operational definition.Accuracy of Group B compared with Group A was 89.8%, significantly lower than 99.5% for Group C (p < 0.001). Conclusion:The 'in-hospital death' classification in the treatment results code demonstrated lower accuracy compared with the operational definition of mortality.Therefore, the treatment results code is not recommended to be used for mortality research in HIRA-based studies, and the 150-day operational definition remains the most appropriate surrogate for mortality.
DOI: 10.56774/jkda250012
Language: English
Subfield: Health Information Management
Field: Health Professions
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W4416731544