Correlations between Web Searches and COVID-19 Epidemiological Indicators in Brazil

• Autor(a) principal: Locatelli,Marcelo Sartori
• Data de Publicação: 2022
• Outros Autores: Cunha,Evandro L. T. P, Guiginski,Janaína, Franco,Ramon A. S, Bernardes,Tereza, Alzamora,Pedro Loures, Silva,Daniel Victor F. da, Ganem,Marcelo Augusto S, Santos,Thiago H. M, Carvalho,Anne I. R, Souza,Leandro M. V, Paixão,Gabriela P. F, Chaves,Elisa França, Santos,Guilherme Bezerra dos, Santos,Rafael Vinícius dos, Freitas,Amanda Cupertino de, Flores,Matheus G, Biezuner,Rachel F, Cardoso,Rodolfo Lins, Fonseca,Rodrigo Machado, Silva,Ana Paula Couto da, Meira Jr,Wagner
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Brazilian Archives of Biology and Technology
• Texto Completo: http://old.scielo.br/scielo.php?script=sci_arttext&pid=S1516-89132022000100314
• topic: Google Trends; infodemiology; epidemiological predictions; digital health

Abstract: COVID-19 rapidly spread across the world in an unprecedented outbreak with a massive number of infected and fatalities. The pandemic was heavily discussed and searched on the internet, which generated big amounts of data related to it. This led to the possibility of attempting to forecast coronavirus indicators using the internet data. For this study, Google Trends statistics for 124 selected search terms related to pandemic were used in an attempt to find which keywords had the best Spearman correlations with a lag, as well as a forecasting model. It was found that keywords related to coronavirus testing among some others, such as “I have contracted covid”, had high correlations (≥0.7) with few weeks of lag (≤4 weeks). Besides that, the ARIMAX model using those keywords had promising results in predicting the increase or decrease of epidemiological indicators, although it was not able to predict their exact values. Thus, we found that Google Trends data may be useful for predicting outbreaks of coronavirus a few weeks before they happen, and may be used as an auxiliary tool in monitoring and forecasting the disease in Brazil.