Análise da taxa de hospitalização devido a doenças de transmissão hídrica e alimentar no Brasil em 2021

Abstract

Foodborne and Waterborne Diseases (FWBD) represent a persistent challenge to public health, with significant impacts on morbidity, mortality, and the burden on healthcare systems. The ingestion of contaminated food or water—by biological, chemical, or physical agents—can trigger outbreaks of varying clinical severity, often requiring hospitalization. In this context, the present study aimed to analyze the factors associated with hospitalization rates resulting from WFBD outbreaks in Brazil, focusing on epidemiological data from the year 2021. This is a descriptive study with a quantitative approach, based on secondary data extracted from the Notifiable Diseases Information System (SINAN), accessed via Tabwin/DATASUS. Statistical analysis was performed using R software, applying a zero-and-one inflated beta regression model, suitable for proportional variables with a concentration of extreme values (0 and 1), such as hospitalization rates. The results showed that bacterial outbreaks were the most prevalent, with Escherichia coli and Salmonella spp. being the most frequent pathogens. Most outbreaks occurred in households, followed by restaurants and events. Açaí was the main food associated with protozoan outbreaks, while water was strongly linked to bacterial and viral outbreaks. Statistical modeling indicated that outbreaks classified as “dispersed cases” had lower average hospitalization rates, suggesting lower clinical severity. Conversely, symptoms such as diarrhea and vomiting were significantly associated with hospitalization, while fever showed an inverse association. The absence of information regarding improper food handling was linked to a higher likelihood of outbreaks without hospitalization, possibly indicating underreporting or lower severity. The analysis of model parameters (μ, σ, ν, and τ) allowed the identification of factors associated with the mean hospitalization rate, its dispersion, the absence of hospitalizations, and total hospitalization of cases, respectively. The model demonstrated good statistical fit, with a higher log-likelihood than the null model, a pseudo-R² of 28.4%, and favorable AIC and deviance values. These findings reinforce the importance of epidemiological surveillance, data quality, and the implementation of rigorous preventive measures—especially in domestic environments—to reduce the severity of outbreaks and the need for hospitalization. This study contributes to the understanding of the determinants of hospitalization due to FWBD and provides insights for improving public health policies related to food safety and collective health in Brazil.