Spatiotemporal Analysis and Epidemiology of Kawasaki Disease in Western New York

Abstract

Kawasaki disease (KD) is one of the leading causes of acquired heart disease in children in developed nations. Epidemiologic evidence suggests that KD is related to an infectious agent; however, the cause remains unknown. Yearly incidence in Japan has been steadily increasing, but few long-term databases of KD cases from North America have been reviewed. We reviewed the epidemiology of local cases over a 16-year period to study incidence with time and temporal and geographic clustering of cases in a representative cohort in North America. The yearly incidence in cases per population <5 years old per 100,000 was 20.2 and 15.9, using International Classification of Disease, ninth revision and detailed chart review, respectively. Using International Classification of Disease, ninth revision alone overestimates our incidence by 27%. We show a distinct seasonality of cases with winter predominance. Applying Kulldorff's spatial scan statistic revealed no significant clustering of cases with either purely spatial or space-time analyses. On purely nonconstrained temporal SaTScan analysis, there was a significant clustering of cases in a 67- to 68-week period in 2000-2001. Our analysis reveals an apparent outbreak of KD in our region in 2000-2001. In contrast to Japan, for the last 14 years, the incidence in our region has been stable.