Análise da distribuição espaço-temporal da leishmaniose visceral e perfil clínico-epidemiológico dos casos e óbitos, Belo Horizonte Minas Gerais, 1994 a 2009

Abstract

Since the decade of 1980, the visceral leishmaniasis (VL) has expanded to major urban centers of Brazil. Several factors contribute to this situation and among them the great mobility of populations, environmental changes and the increase in population density, which promote contact between humans, vectors and other animals. In this latter group the dog is the main reservoir of disease in the urban environment. Among the goals of the Program of Surveillance and Control of Visceral Leishmaniais (PCVL) are the reduction of morbidity and mortality and transmission of VL. Whereas the best definition of risk areas for VL can maximise the result of interventions for prevention and control, the objectives of this study were to analyze the spatio-temporal distribution and the relative risk of VL in Belo Horizonte, the clinical-epidemiological profile of cases and deaths. Other no less important objective was to verify the existence of association between the relative risk and canine infection, elevation, vegetation and demographic and socio-economic variables present in Health Vulnerability Index. The vegetation was characterized from the satellite images obtained by remote sensing. In the period studied, 1994-2009, 1293 human cases of VL were reported, including 167 deaths. All the studied variables have been geo-referenced according to the geographical units of analysis: sanitary districts and covered areas by 146 basic health units of Belo Horizonte (BH). This study has two epidemiological designs: ecological and historical cohort. The data of the Reportable Disease Information System Brazilian (SINANBrazilian Ministry of Health) were used in historic cohort study that enabled to identify signs and symptoms, present at the time of clinical suspicion of VL, significantly associated with death: weakness (OR 2.90), Leishmania-HIV co-infection (OR 2.36), age 60 years (OR 3.11), presence of other infectious (OR 3.22), hemorrhagic phenomena (OR 3.50) and jaundice (OR 10.08). Children younger than 5 years were most affected by the disease, while the higher lethality was observed in individuals of 30-39 years old (20.21%) and 60 years (24.42%). In ecological design, spatiotemporal data statistical modeling were used to evaluate the incidence and relative risk (RR) of VL. Generalized linear models have identified increasing trends of temporal evolution of the incidence of VL in all districts of BH with large variations between them. From 2004, the districts Barreiro and West have been characterized as the most iv recent areas of increased incidence of VL, what characterizes the difficulty of containing the spread of the disease. Bayesian modeling, with spatio-temporal dependency between neighbouring areas, allowed better visualization and understanding of the evolution of the incidence of VL. Another Bayesian approach allowed to identify the association between RR of VL, average income and dogs serum reagent/inhabitant ratio. This reason has not been described in the literature and, in this study, was used with success. It was concluded that there is heterogeneity of RR among the 146 areas of BH, 25 of which (17%) had higher values of RR (1,5-2,55) and are located in East, Northeast, Northwest, North and Venda Nova. By the obtained results, it was considered appropriate to the use of Health Vulnerability Index to assist in the planning of the PCVL, since its components of income and education showed significant association with the RR. The identification of factors associated with death can contribute to better understanding and reduction of lethality by VL. The SINAN is an important tool of PCVL, even regard the need to improve the underreporting, and quality and consistency of databases. Belo Horizonte is an urban centre with active transmission and expansion of VL, becoming a challenge to the surveillance and control program.