Background/Purpose: Kawasaki disease (KD) is a systemic vasculitis associated with the development of coronary artery aneurysms predominantly affecting children less than 5 years of age. Prompt and accurate diagnosis remains challenging due to the unknown cause of this condition and the overlap of its clinical findings with other febrile illnesses in children. Viral infections have been proposed as a potential trigger for KD, particularly given the observed seasonal peaks in winter and spring across the continental United States. Previous studies have also suggested temporal associations between viral activity and KD incidence at certain centers. This study aims to characterize the seasonal distribution of KD and assess the temporal relationship between hospital viral positivity rates and KD incidence at a large tertiary center.

Methods: With IRB approval, we retrospectively reviewed patients diagnosed with KD at our institution from 2012 to 2019. We collected admission dates and demographic data. We obtained monthly viral positivity rates in hospitalized patients for 18 viruses (influenza A/B, respiratory syncytial virus, adenovirus, human metapneumovirus, rhinovirus, parainfluenza viruses 1-4, norovirus, sapovirus, rotavirus, adenovirus (including types F40/41), astrovirus, herpes simplex virus (cutaneous, CSF and blood), enterovirus (CSF and blood), and varicella-zoster virus (cutaneous)) from institutional epidemiology reports. Negative binomial regression models were used to assess associations between viral activity and KD incidence: (1) a univariable analysis assessing each virus independently; (2) a multivariable model for individual viral infection adjusting for month and year; and (3) a fully adjusted model including all viruses while controlling for month and year.

Results: From 2012-2019, 858 patients were diagnosed with KD at our institution. The median age was 2.6 years (IQR 1.2–4.5), with no significant difference in male-to-female ratio across the study period. A statistically significant difference in racial/ethnic distribution was observed in 2013 compared to 2012, 2014, 2015, and 2017 (Figure 1). KD cases were evenly distributed across winter (26%), spring (27%), and summer (26%) (Figure 2). For all regression models, no significant temporal association was found between viral positivity rates and KD incidence (Figure 3).

Conclusion: At our institution, KD cases were evenly distributed across winter, spring, and summer, a pattern that contrasts with prior studies reporting seasonal peaks. Additionally, we found no significant temporal associations between viral positivity rates and KD incidence. These negative findings suggest that common viruses may not be the primary triggers of KD and underscore the need for broader investigative approaches. Ongoing research is critical to elucidate the underlying causes of KD and to enhance timely diagnosis and treatment in young children.

Figure 1: Characteristics of patients with Kawasaki disease, by year of diagnosis

Figure 2: Number of Kawasaki disease cases, by season

Figure 3: Association of viral infections and numbers of Kawasaki disease cases diagnosed, negative binomial regression models

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/temporal-association-between-viral-positivity-rates-and-kawasaki-disease-incidence-at-a-large-tertiary-center/