80 - IDENTIFICATION OF EPIDEMIOLOGIC FEATURES OF KAWASAKI DISEASE USING BIG DATA ANALYSIS FROM THE NATIONAL DATABASE OF HEALTH INSURANCE CLAIMS AND SPECIFIC HEALTH CHECKUPS IN JAPAN

Background: In Japan, a national survey of Kawasaki Disease (NSKD) has been conducted every two years since 1970, and KD epidemiological characteristics have gradually become clear. However, the possibility of extracting clinical and epidemiological profiles similar to those of NSKD from medical databases remains unexplored. The aim of the study was to investigate KD incidence and acute treatment in Japan using medical information databases and to examine its consistency with NSKD data.

Methods: Two databases were used: the National Database of Health Insurance Claims and Specific Health Checkups of Japan (NDB) and the NSKD. Currently, the NDB has been implemented in 99% of healthcare facilities in Japan and contains detailed medical information, diagnosis, treatment, and other aspects related to KD. We used the NDB data from January 2013 to December 2020 for children aged < 18 years who were diagnosed with KD. The following patients were excluded: (1) those with suspected KD; (2) those who did not receive inpatient treatment; and (3) those who did not receive intravenous immunoglobulin (≥ 700 mg/kg) or cyclooxygenase inhibitors. Using drug, disease, and procedure records in the NDB, we extracted information on similar variables from the NSKD. We compared the clinical information extracted from the NDB between January 2013 and December 2020 with that of patients registered in the NSKD database.

Results: During the study period, 137,242 patients with KD were identified from the NDB. Sex ratio, age at diagnosis, chronological change in KD onset, and clinical ecological distribution were identical in both databases. Furthermore, the proportion of patients receiving intravenous immunoglobulin, prednisolone, methylprednisolone, infliximab, cyclosporine A, urinastatin, and plasma exchange as initial or additional treatments was similar in both databases. However, the proportions of incomplete KD, coronary artery disease, myocardial infarction, and valvular disease due to KD obtained from the NDB disease information differed from the NSKD data. The proportion of patients with KD receiving warfarin, clopidogrel, or ticlopidine was similar to that of patients with abnormal coronary arteries at 30 days of illness.

Conclusion: As with NSKD, epidemiologic characteristics and treatment outcomes can be extracted for the NDB. Further research is needed to determine the diagnostic certainty and clinical outcomes.