Modelagem da vulnerabilidade à COVID-19 em Minas Gerais: influências socioeconômicas

Abstract

This thesis investigates the vulnerability of municipalities in Minas Gerais to the spread of COVID-19 due to socioeconomic factors. To this end, it integrates bibliometric approaches, spatial statistics, and multicriteria modeling, providing a comprehensive analysis of the distribution of the disease and its structural determinants. The study is structured along three interdependent axes. The first axis carried out a bibliometric analysis of scientific production on the socioeconomic impacts of COVID-19 at three geographic levels: global, national (Brazil), and regional (Minas Gerais). Articles from the Scopus and Web of Science databases were analyzed using the pubmedR and bibliometrix packages in the R environment. The results indicated a slowdown in global (-38.77%) and national (-55.91%) scientific production, contrasting with a significant growth in regional studies (+41.42%). The analysis revealed the centrality of countries such as the United States, the United Kingdom, and China in scientific collaboration and highlighted the predominance of research on vulnerable groups, the labor market, poverty, and mental health. This stage allowed us to identify gaps that guided the empirical investigation of the following axes. The second axis applied spatial analyses to investigate the relationship between socioeconomic factors and the spread of COVID-19 in Minas Gerais. Spatial regression models, including spatial error models (SEMs) and spatial lag models (SLMs), were constructed based on data from the Institute of Applied Economic Research (IPEA). The results showed that municipalities with greater longevity and better urban infrastructure had lower mortality rates. This stage revealed spatial patterns of autocorrelation that provided support for the formulation of strategies to mitigate risks and optimize the allocation of medical resources. In the third axis, the previous findings were integrated for multicriteria modeling of municipal vulnerability to COVID-19, considering socioeconomic, structural, and epidemiological variables. Using a multicriteria analysis model (MCA), the 853 municipalities in Minas Gerais were classified according to their risk of transmission. The results indicated that Belo Horizonte, Uberlândia and Contagem were more vulnerable due to their high population density, road connectivity and predominance of jobs in the trade and services sector. Urbanization was the most influential factor in the spread of the virus (22.92%), followed by road connectivity (17.54%). Cross-validation improved the predictive capacity of the analysis, reinforcing the need for interdisciplinary approaches in the formulation of public policies. By integrating these three axes, the thesis shows how the combination of different analytical approaches – from scientific production to spatial and multicriteria modeling – allows for an in-depth understanding of municipal vulnerability to COVID-19. The results reinforce the need for evidence-based policies and demonstrate the importance of integrating epidemiology, geoinformation and public management to address health emergencies.