Análise espaço-temporal da leptospirose no município do Rio de Janeiro (1995-1999)

The leptospirosis, one of the most prevalent zoonosis in the world, is endemic in Brazil. In urban areas it occurs, characteristically, after periods of high pluviosity with flooding. The city of Rio de Janeiro presented, in 1996, one of the largest urban epidemics, just after big storms that happened in February, with 1792 notified cases and 51 deaths. The aim of this dissertation is to describe the space-time pattern of leptospirosis, and, at the same time, to evaluate a techniques for detection of clusters of cases and delimitation of risk surfaces, contributing to develop methods for environmental surveillance with a territorial approach. In the first paper, an exploratory technique of spatial analysis was used, based on the ratio between two layers, allowing the visualisation of the risk surface of the disease. The interpolation method for the creation of the layers was a intensity Kernel of the points representing the place of residence of the cases, in the numerator, and a Kernel of the population of the census tracts in the denominator. Through this method it was possible to delimit “hot” areas, and to compare them, visually, with the maps of the demographic density, of slums and the polygons that delimit areas subject to floods. Comparing the epidemic year (1996) with the other ones, it was noticed that the largest incidence rates were not observed in supposed “higher risk” areas. In the second paper we present a space-time analysis of leptospirosis cases in the city of Rio de Janeiro, between 1995 and 1999, using a method for detection and localisation of space-time clusters, based on a scan statistics, available in the software SaTScan. Clusters were detected only in the epidemic year of 1996. The main cluster was localised in Jacarepagua and the others were spread throughout the entire city, expressing the intensity of the epidemic on this year. Differences in the profile of the cases in the two moments – epidemic and endemic – were significant (qui-square test) only for the variables severity of disease and gender. Endemic cases were more severe, and the absolute number of women was significantly larger in the epidemic period.