Prediction of the Spread of the COVID-19 Pandemic with Google Searches: An Infodemiological Approach

Aim: In outbreaks, public concern is reflected in search behavior. Examining the health literacy of the population and predicting before the diagnosis of cases may benefit the outbreak management in countries. This study aims to evaluate the association of search behavior with the number of new confirmed cases in the affected countries by the Coronavirus disease 2019 (COVID-19) pandemic. This retrospective study is based on monitoring search behavior with an infodemiology and infoveillance approach. Materials and Method: Google TrendsTM was used to investigate the Internet search behavior related to COVID-19 for 10 countries from February 15, 2020, to November 10, 2020. Spearman’s rank correlation and time-lag correlation were used to determine the correlation with a delay of -30 days to +30 days between public interest and new daily-confirmed cases. Results: The level of COVID-19-related interest peaked about 33 days before the first peak in the number of cases. The correlation gradually decreased in seven countries towards the peak of cases. Spearman's rank correlations between Google searches and the number of new confirmed cases showed a negative correlation in Argentina, Brazil, India, and the United Kingdom (p