Abstract

SARS-CoV-2, the virus that causes coronavirus disease (COVID-19), was first identified in Wuhan, China, in December 2019 and has since spread across the world and was responsible for the pandemic that has resulted in global health crises. As the population of patients recovering from COVID-19 grows, establishing an understanding of the health issues surrounding them is critical.

Objectives: The general objective is to analyze the incidence of HIV infections among females in Brazil.

Design: This study is characterized as ecological. Sample: The study population will be composed of females aged 15 years or older, living with HIV/AIDS, notified to the Notifiable Diseases Information System (SINAN), disposable by DATASUS website.

Measurements: For temporal analysis, the annual incidence rate will be considered by region of the country. The series were analyzed using an R package for causal inference using Bayesian structural time series models. This R package implements an approach to estimate the causal effect of a designed intervention in a time series, this model is then used to attempt to predict the counterfactual, that is, how the response metric would have evolved after the intervention if the intervention had never occurred. As with all non-experimental approaches to causal inference, valid conclusions require strong assumptions, in which case from the R package CausalImpact, it is assumed that there is a set of control time series that were not affected by the intervention.

Results: In the period from 2016 to 2023, 80,282 cases of HIV/AIDS were reported in females, with the highest percentage reported in 2016 (15.26%), followed by the years 2017 (14.43%), 2018 (14.12% ), 2019 (14.01%), 2022(13.04%), 2021 (12.49%), 2020 (10.88%) and 2023 (5.77%). Consolidated data up to 06/30/2023 was used, with the DATASUS page being updated on 11/30/2023. In all regions of Brazil, incidence rates reduced from 2016 to 2017, followed by an increase from 2017 to 2018, except in the Central-West, South and Southeast regions, which showed a reduction. From 2018 to 2019, there was an increase in rates only in the North and Central-West regions, followed by a reduction in rates in all regions in 2020 and an increase in 2021. The country shows a reduction in the rate over the years, with an increase in 2021, as most states showed a drop in HIV infection rates between 2016 and 2020, followed by an increase in 2021. Series analysis using an R package for causal inference using Bayesian structural time series models implemented an approach to estimate the causal effect of a designed intervention on a time series, the package builds a Bayesian structural time series model, this model is then used to try to predict. In the post-intervention period, the post-pandemic period, the response variable had an average value, in the absence of it, in a hypothetical world without COVID-19, we would have expected a higher average response. Subtracting this prediction from the observed response produces an estimate of the causal effect that the intervention had on response. This means that the reducing effect observed during the intervention period is statistically significant. The probability of obtaining this effect by chance is very small (one-sided Bayesian probability of tail area p = 0.001).

Conclusions: With the information collected, it will be possible to compare the incidence of HIV infections compared to the projection for the period and set specific goals for diagnosis and treatment for the population, thus effectively using rapid and extensive testing for users, seeking early diagnosis.
Keywords: Women; HIV; COVID-19.