AIDS-related mortality in Pará Province, Brazilian Amazon region: spatial and temporal analysis

Coordenação de Aperfeiçoamente de Pessoal de Nível Superior - CAPES (https://www.gov.br/capes/ pt-br) through the Edital PROCAD-AM, grant number 88881.200527/2018-01.Federal University of Pará. Institute of Health Sciences. Faculty of Nursing. Belém, PA, Brazil.Federal University of Pará. Institute of Health Sciences. Faculty of Nursing. Belém, PA, Brazil.University of Évora. Department of Mathematics. Évora, Portugal.University of São Paulo at Ribeirão Preto. College of Nursing. Department of Maternal-Infant Nursing and Public Health. Ribeirão Preto, SP, Brazil.Ministério da Saúde. Secretaria de Vigilância em Saúde e Ambiente. Instituto Evandro Chagas. Ananindeua, PA, Brasil.Federal University of Pará. Institute of Health Sciences. Faculty of Nursing. Belém, PA, Brazil.Federal University of Pará. Institute of Health Sciences. Faculty of Nursing. Belém, PA, Brazil.Federal University of Pará. Institute of Health Sciences. Faculty of Nursing. Belém, PA, Brazil.Despite considerable therapeutic advances in the care of people living with human immunodeficiency virus (HIV) and with the acquired immunodeficiency syndrome (AIDS) and an overall reduction of 47% in the AIDS mortality rate in the last decade, the AIDS-mortality rates remains high. The social determinants of health (SDH) have a direct influence on the dynamics of this phenomenon. However, changes in SDH caused by the implemented policies against HIV have been poorly investigated. Moreover, the Brazilian rainforest has had the highest and continuously increasing AIDS mortality rate in Brazil since the 1980s. In this study, AIDS mortality in a province of the Brazilian rainforest was examined by using temporal and spatial analyses. Methods. In this ecological study, data from 2007 to 2018 were extracted from the Mortality Information System provided by the State Department of Public Health of Pará. For the temporal analysis, the integrated autoregressive model of moving average (ARIMA) and locally weighted polynomial regression (STLF) were used to forecast AIDS mortality from 2019 to 2022. For the spatial analysis, spatial autocorrelation and geographically weighted regression (GWR) analyses were employed. Results. The samples consisted of 6,498 notifications for AIDS-related deaths. From 2007 to 2013, the AIDS mortality rates showed an upward trend, followed by a stabilization until 2018 and an upward forecasted trend from 2019 to 2022. High mortality rates and high-high clusters were found in economic pole municipalities. Furthermore, AIDS mortality risk was directly associated with per capita income and demographic density, except in the southwestern region of Pará, which exhibited an inverse association with population density. Conclusion. Although the policies against HIV may have contributed to the stabilization of AIDS mortality rates from 2013 in Pará, the upward forecasted trend until 2022 raises an alert and concern to health authorities to provide reinforcement of the policies. The geographic variability of AIDS mortality promoted by SDH provides subsidies to health authorities to implement SDH-focused strategies for AIDS mortality reduction