Abstract
IMPORTANCE: Infections are proposed to be triggering factors for Kawasaki disease (KD), although its etiological factors remain unknown. Recent reports have indicated a 4- to 6-week lag between SARS-CoV-2 infection and multisystem inflammatory syndrome in children with a similar presentation to that of KD. OBJECTIVE: To investigate the temporal correlation between KD and viral infections, focusing on respiratory viruses. DESIGN, SETTING, AND PARTICIPANTS: This cohort study was conducted among individuals aged 0 to 19 years diagnosed with KD between January 2010 and September 2020 from the Korean National Health Insurance Service. Data on infectious disease outbreaks from 2016 to 2019 were collected from the Korea Disease Control and Prevention Agency, Korean Influenza and Respiratory Virus Monitoring System, Korea Enteroviruses Surveillance System, and the Enteric Pathogens Active Surveillance Network in South Korea. Data were analyzed from December 2020 to October 2021. MAIN OUTCOMES AND MEASURES: National databases for infectious diseases were used for a time-series analysis of the correlation between viral infections and KD. The temporal correlation between infectious disease outbreaks and KD outbreaks was evaluated using the Granger causality test (G-test), which is a useful tool to estimate correlations between 2 time series of diseases based on time lags. RESULTS: Overall, 53 424 individuals with KD were identified, including 22 510 (42.1%) females and 30 914 (57.9%) males and 44 276 individuals (82.9%) younger than 5 years. Intravenous immunoglobulin-resistant KD was identified in 9042 individuals (16.9%), and coronary artery abnormalities were identified in 384 individuals (0.7%). Of 14 infectious diseases included in the analyses, rhinovirus infection outbreaks were identified as significantly correlated at 1 to 3 months before KD outbreaks in South Korea (r = 0.3; 1 month: P < .001; 2 months: P < .001; 3 months: P < .001). Outbreaks of respiratory syncytial virus infection were identified as significantly correlated with KD outbreaks by 2 months (r = 0.5; 2 months: P < .001). Additionally, varicella outbreaks were identified as significantly correlated at 2 and 3 months before KD outbreaks (r = 0.7; 2 months: P < .001; 3 months: P < .001). CONCLUSIONS AND RELEVANCE: In this cohort study with a time series analysis of children and youth in South Korea with KD, respiratory infections caused by rhinovirus and respiratory syncytial virus and varicella outbreaks were significantly correlated with KD at 1 to 3 months before KD outbreaks.