Building Linked Big Data for Stroke in Korea: Linkage of Stroke Registry and National Health Insurance Claims Data

• DOI: 10.3346/jkms.2018.33.e343
• Publication Date: 2018-12-14T06:22:24Z
• Authors: Tae Jung Kim; Ji Sung Lee; Ji-Woo Kim; Mi Sun Oh; Heejung Mo; Chan-Hyuk Lee; Han-Young Jeong; Keun-Hwa Jung; Jae-Sung Lim; Sang-Bae Ko; Kyung-Ho Yu; Byung-Chul Lee; Byung-Woo Yoon
• Keywords: Stroke; Record Linkage; Interquartile range; Medical record

Background: Linkage of public healthcare data is useful in stroke research because patients may visit different sectors the health system before, during, and after stroke.Therefore, we aimed to establish high-quality big on Korea by linking acute registry national claim databases.Methods: Acute (n = 65,311) with suitable for linkage were included Clinical Research Center Stroke (CRCS) during 2006-2014.We linked CRCS databases Health Insurance Review Assessment Service (HIRA).Linkage was performed using 6 common variables: birth date, gender, provider identification, receiving year number, statement serial number benefit statement.For matched records, accuracy evaluated differences between hospital visiting date commencement insurance care HIRA.Results: Of 65,311 cases, 64,634 HIRA cases (match rate, 99.0%).The proportion true matches 94.4% 61,017) data.Among (mean age 66.4 years; men 58.4%), median National Institutes Scale score 3 (interquartile range 1-7).When comparing baseline characteristics false matches, no substantial difference observed any variable. Conclusion:We could claims without personal identifiers.We plan conduct improve database.