Bayesian spatio-temporal modeling of severe acute respiratory syndrome in Brazil: A comparative analysis across pre-, during, and post-COVID-19 eras

Abstract

This paper presents an investigation into the spatio-temporal dynamics of Severe Acute Respiratory Syndrome (SARS) across the diverse health regions of Brazil from 2016 to 2024. Leveraging extensive datasets that include SARS cases, climate data, hospitalization records, and COVID-19 vaccination information, our study employs a Bayesian spatio-temporal generalized linear model to capture the intricate dependencies inherent in the dataset. The analysis reveals significant variations in the incidence of SARS cases over time, particularly during and between the distinct eras of pre-COVID-19, during, and post-COVID-19. Our modeling approach accommodates explanatory variables such as humidity, temperature, and COVID-19 vaccine doses, providing a comprehensive understanding of the factors influencing SARS dynamics. Our modeling revealed unique temporal trends in SARS cases for each region, resembling neighborhood patterns. Low temperature and high humidity were linked to decreased cases, while in the COVID-19 era, temperature and vaccination coverage played significant roles. The findings contribute valuable insights into the spatial and temporal patterns of SARS in Brazil, offering a foundation for targeted public health interventions and preparedness strategies.