Geographic Information System-based association between the sewage network, geographical location of intermediate hosts, and autochthonous cases for the estimation of risk areas of schistosomiasis infection in Ourinhos, São Paulo, Brazil

Abstract INTRODUCTION: Ourinhos is a municipality located between the Pardo and Paranapanema rivers, and it has been characterized by the endemic transmission of schistosomiasis since 1952. We used geospatial analysis to identify areas prone to human schistosomiasis infections in Ourinhos. We studied the association between the sewage network, co-occurrence of Biomphalaria snails (identified as intermediate hosts [IHs] of Schistosoma mansoni), and autochthonous cases. METHODS: Gi spatial statistics, Ripley's K12-function, and kernel density estimation were used to evaluate the association between schistosomiasis data reported during 2007-2016 and the occurrence of IHs during 2015-2017. These data were superimposed on the municipality sewage network data. RESULTS: We used 20 points with reported IH; they were colonized predominantly by Biomphalaria glabrata, followed by B. tenagophila and B. straminea. Based on Gi statistics, a significant cluster of autochthonous cases was superimposed on the Christoni and Água da Veada water bodies, with distances of approximately 300 m and 2200 m from the points where B. glabrata and B. straminea were present, respectively. CONCLUSIONS: The residence geographical location of autochthonous cases allied with the spatial analysis of IHs and the coverage of the sewage network provide important information for the detection of human-infection areas. Our results demonstrated that the tools used for direct surveillance, control, and elimination of schistosomiasis are appropriate.

Spatial pattern of mortality from breast and cervical cancer in the city of São Paulo / Padrão espacial da mortalidade por câncer de mama e colo do útero na cidade de São Paulo

ABSTRACT OBJECTIVE To verify the spatial pattern of mortality from breast and cervical cancer in areas of primary health care, considering socioeconomic conditions. METHODS This is an ecological study, from January 2000 to December 2016. The study area is the municipality of São Paulo, Brazil, and its 456 coverage areas of primary health units. Information on deaths of women aged 20 years or over were geocoded according to residence address. We calculated mortality rates, standardized by age, and smoothed by the local empirical Bayesian method, and grouped into three or two years to reduce the random fluctuation of the data. In addition, bivariate global and local Moran indexes were calculated to verify the existence of spatial agglomeration of standardized mortality rates with a domain of socioeconomic condition, elaborated based on the Índice Paulista de Vulnerabilidade Social (IPVS - São Paulo Index of Social Vulnerability). RESULTS The success rate of geocoding was 98.9%. Mortality from breast cancer, without stratification by time, showed a pattern with higher rates located in central regions with better socioeconomic conditions. It showed a decrease at the end of the period and a change in spatial pattern, with increased mortality in peripheral regions. On the other hand, mortality from cervical cancer remained with the highest rates in peripheral regions with worse socioeconomic conditions, despite being reduced over time. CONCLUSION The spatial pattern of mortality from the studied cancers, over time, suggests association with the best socioeconomic conditions of the municipality, either as protection (cervical) or risk (breast). This knowledge may direct resources to prevent and promote health in the territories.

RESUMO OBJETIVOS Verificar o padrão espacial da mortalidade pelos cânceres de mama e do colo do útero, em áreas da atenção primária à saúde, levando em consideração as condições socioeconômicas. MÉTODOS O estudo é ecológico, de janeiro de 2000 a dezembro de 2016. A área de estudo é o município de São Paulo, Brasil, e suas 456 áreas de abrangência das unidades básicas de saúde. As informações sobre óbitos de mulheres com 20 anos ou mais de idade foram geocodificadas segundo endereço de residência. Foram calculadas as taxas de mortalidade, padronizadas por idade, e suavizadas pelo método bayesiano empírico local, além de agrupadas em três ou dois anos para reduzir a flutuação aleatória dos dados. Além disso, foram calculados os índices de Moran global e local bivariados, para verificar a existência de aglomeração espacial das taxas de mortalidade padronizadas com um domínio de condição socioeconômica, elaborado a partir do Índice Paulista de Vulnerabilidade Social. RESULTADOS A taxa de sucesso da geocodificação foi de 98,9%. A mortalidade por câncer de mama, sem estratificação por tempo, apresentou um padrão com maiores taxas localizadas nas regiões centrais e com melhores condições socioeconômicas. Apresentou queda ao final do período e mudança de padrão espacial, com aumento da mortalidade nas regiões periféricas. Já a mortalidade por câncer do colo do útero manteve-se com as maiores taxas nas regiões periféricas e com piores condições socioeconômicas, apesar de apresentar redução ao longo do tempo. CONCLUSÃO O padrão espacial da mortalidade pelos cânceres do estudo, ao longo do tempo, sugere associação com as melhores condições socioeconômicas do município, seja como proteção (colo) ou risco (mama). Esse conhecimento pode direcionar recursos para a prevenção e a promoção da saúde nos territórios.