SciELO - Scientific Electronic Library Online

vol.21Impact of autocorrelation in the time analysis of HIV mortality coefficients in Brazil author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Revista Brasileira de Epidemiologia

Print version ISSN 1415-790XOn-line version ISSN 1980-5497

Rev. bras. epidemiol. vol.21  São Paulo  2018  Epub Dec 03, 2018 


Reply to comment on the article “HIV/AIDS Mortality in Brazil, 2000-2015: Are there reasons for concern?

Mark Drew Crosland GuimarãesI 

Mariângela CarneiroII 

Daisy Maria Xavier de AbreuI 

Elisabeth Barboza FrançaI 

IEpidemiology and Health Assessment Research Group, Department of Preventive and Social Medicine, School of Medicine, Universidade Federal de Minas Gerais - Belo Horizonte (MG), Brazil.

IIDepartment of Parasitology, Institute of Biological Sciences, Universidade Federal de Minas Gerais - Belo Horizonte (MG), Brazil.

Garcia and Traebert1 commented our article published in Revista Brasileira de Epidemiologia.2 The correspondents questioned the application of the linear regression method in data from Brazil, the Southern and Southeastern regions, and a series of six other states. However, they are mistaken in their reading of the methodology used in our article and incoherently interpreted the results, whose central focus was to compare HIV/AIDS mortality data at two points in time, 2000 and 2014/2015, according to two distinct sources: Department of STD, AIDS, and Viral Hepatitis (Departamento de DST, Aids e Hepatites Virais - DDAHV) and GBD. We clarify that we used linear regression, a classic and easy-to-understand method, only in Brazilian regions, according to data published by the DDAHV, to describe the information about historical series of HIV/AIDS mortality in the country. This analysis does not have the purpose of predicting the future, let alone comparing data with the GBD. About the past, there is no mention of angular coefficient significance. That is, it continues to be an analysis that describes the situation in the period in relation to the decrease in absolute value of the rate, and not in its percentage, as suggested by the correspondents’ methodology. We did not make regressions for the states, and the relative percentage variations for the two sources were calculated based on data observed and published by both in 2000 and 2014/2015, respectively. We did not estimate these coefficients, given that the complexity of each methodology requires the same pattern for comparability. Moreover, as pointed out in the limitations, the interpretation of data from sources so heterogeneous should always be made with caution. Thus, we stand by our analysis and interpretation considering the proposed objective of the article. We thank the correspondents and understand that the proposals made are relevant, and a new study assessing and comparing both data sources can be developed.


1. Garcia LP, Traebert JL. Carta ao Editor: impacto da autocorrelação na análise temporal dos coeficientes de mortalidade pelo HIV no Brasil. Rev Bras Epidemiol. 2018; 21: e180020 ]

2. Guimarães MDC, Carneiro M, Abreu DMX, França E. Mortalidade por HIV/Aids no Brasil, 2000-2015: Motivos para preocupação? Rev Bras Epidemiol. 2017; 20 (Supl. 1): 182-90. ]

Creative Commons License Este é um artigo publicado em acesso aberto sob uma licença Creative Commons