Detection and Hierarchy of Aedes aegypti Clusters by the Scan Statistic with a Space-time Permutation Model

Abstract

The use of spatial statistics associated with human epidemiological studies has intensified in the last two decades. In this sense, this article uses scan statistic for detections of spatio-temporal clusters of Aedes aegypti outbreaks at the urban area of Joinville, Santa Catarina. The control of this insect is very important for many cities in Brazil, which has presented some difficult in order to restring infestations, due to the survival of these insects and also dangerous factors related to the society. The aim of this paper is to apply an statistic spatio-temporal model in order to identify clusters, defining its location, size and order. The systematization of the database covers the dairy records of Ae. aegypti outbreaks, of the period from January 2009 to September 2018, led by the Office of Epidemiological Surveillance of Santa Catarina. The outbreaks which were located with the use of the cadastral data of Joinville. For the statistical analysis, the space-time probability model was used, based on the retrospective approach. In the result of this research it was possible to showed the potential of this model in order to detect historical and active clusters. The retrospective approach has identified 26 clusters; the only one active was located at the Boa Vista neighborhood, which shows that this region needs intervention measures. The development of the method was efficient for knowledge of historical clusters and, mainly, for definition of early outbreaks. This characterizes Scan statistics as an important tool in order to guide the sanitary activities for insect control.