Web browser as a tool for predicting the incidence of influenza

Background. Infodemiology is focused on the analysis of web content to predict health phenomena. Google Trends (GT) is a free and publicly available service that permits analyses of searches performed with the Google web search engine. With GT it is possible to specify how often certain keywords are searched for. Objectives. The purpose of the study was to determine the feasibility of using data on the frequency of searches with the Google search engine to predict influenza incidence. Material and methods. Using the GT service, data on the frequency of searches for the Polish equivalents of “flu”, “cold” and “fever” in the period of 2014–2016 in Poland were retrieved. Simultaneously, the epidemiological reports prepared by the National Institute of Public Health – National Institute of Hygiene (NIPH-NIH) were obtained for influenza incidence in the same period. Correlations between the variables were assessed using Spearman's rank-order correlation. Results. A statistically significant correlation was confirmed between the average daily search coefficients (ADSC) for all 3 keywords and weekly influenza incidence according to the NIPH-NIH data. The strongest correlation was found for the ADSC of the word “cold” (r = 0.77; p < 0.05). Conclusions. The frequency of searches implemented with the Google search engine may be used for predicting the incidence of influenza in the Polish population.