Análise espacial dos casos de violência infanto-juvenil em regiões de faixa de fronteira do Brasil

• Ano de defesa: 2020
• Autor(a) principal: Miura, Aline Suelen
• Orientador(a): Nihei, Oscar Kenji
• Banca de defesa: Andrade, Luciano de , Arcoverde, Marcos Augusto Moraes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual do Oeste do Paraná; Foz do Iguaçu
• Programa de Pós-Graduação: Programa de Pós-Graduação em Saúde Pública em Região de Fronteira
• Departamento: Centro de Educação Letras e Saúde
• País: Brasil
• Palavras-chave em Português: Violência; Análise espacial; Maus-tratos infantis; Mapeamento gráfico; Epidemiologia
• Palavras-chave em Inglês: Violence; Spatial analysis; Child abuse; Geographic mapping; Epidemiology
• Palavras-chave em Espanhol: Violencia; Análisis espacial; Maltrato a los niños; Mapeo geográfico; Epidemiologia
• Área do conhecimento CNPq: SAUDE COLETIVA::EPIDEMIOLOGIA
• Link de acesso: http://tede.unioeste.br/handle/tede/5071

The research aimed to analyze the spatial distribution of violence among children and adolescents from zero to 18 years of age in the 588 municipalities of the Brazilian Border Strip (FFB). Ecological, descriptive, cross-sectional study, with secondary data from 2009 to 2017, using spatial analysis techniques. The data were obtained from the e-SIC. Population, socioeconomic and demographic information were made available by the IBGE and Atlas of Human Development in Brazil. The FFB and its three arcs (north, central and south) were analyzed. The total of the notifications were analyzed temporally through the programs Minitab 18 and Stata 15 and analyzed spatially through the programs QGIS 3.4.8 and GeoDa 1.14.0, the global Moran Index was used for the analysis of spatial autocorrelation, LISA analyzes and Getis-Ord to identify clusters, and multivariate regression analysis OLS, SLM and SEM to identify the socioeconomic and demographic indicators associated with municipal violence reporting rates. As a result, there were 43,509 notifications of violence, of which 10037 belonged to the north arc, 8710 to the central arc, and 24762 to the south, with physical and sexual being the most reported. For temporal trend, the greatest ascendancy for physics in the FFB and central and south arches, and in the north the sexual. The highest rates of annual averages per 100,000 children and adolescents in the FFB were for physical violence (46.7), followed by psychological (29.2), sexual (20.1), and considering the typology and highest annual average, it was in the south arc with the physical (49.3), the north with the sexual (38.6), again the south with the psychological (35.3). In the spatial exploratory analysis at FFB, 147 municipalities had violence rates > 865.1; the northern arc 17 municipalities with rates > 601.8; central 25 had a score > 575.6; and south 104 of these had rates > 909.7 per 100 thousand children and adolescents. The analyzes revealed a positive spatial autocorrelation of the FFB violence notification rate (I= 0.117753, p= 0.003), and three arcs: north (I= 0.109556, p=0.047000), central (I= 0.3337191, p= 0.00200) and south (I = 0.0649426, p= 0.028000), indicating the presence of spatial dependence among the municipalities in relation to this dependent variable. For the LISA analysis, in relation to FFB, 12 municipalities were found in the formation of 7 Alto-Alto clusters, and 46 in 10 Baixo-Baixo clusters. In the Getis-Ord analysis, 33 municipalities were found in the formation of 13 Alto clusters, 56 municipalities in 15 Baixo clusters. In the multivariate analysis for FFB, the best model obtained was the SLM, where the variable PEA18M (I= -0.046, p= 0.046) demonstrated a significant negative association with the violence notification rate, PESOURB (p= 0.021) and T_SUPER25M (p=0.001) pointed to a significant positive association. To the north, the multivariate analysis did not present a valid model in the OLS analysis. At the center, the variables that showed a significant positive association were T_SUPER25M (p<0.001) and T_DENS (p<0.001). In the south, T_ANALF18M (p= 0.031) demonstrated a significant negative association with the reporting rate of violence, and PESOURB (p<0.001) a significant positive association. It is concluded that the distribution of cases of violence against children and adolescents in the FFB and arches does not occur randomly, with spatial dependence on their disposition and being influenced by the degree of occupation, education, employment / income, demographic area and housing.