Correlation between flu and Wikipedia's pages visualization

Acta Biomed. 2021 Feb 8;92(1):e2021056. doi: 10.23750/abm.v92i1.9790.

Abstract

Introduction: This study aimed to assess if the frequency of the Italian general public searches for influenza, using the Wikipedia web-page, are aligned with Istituto Superiore di Sanità (ISS) influenza cases.

Materials and methods: The reported cases of flu were selected from October 2015 to May 2019. Wikipedia Trends was used to assess how many times a specific page was read by users; data were extracted as daily data and aggregated on a weekly basis. The following data were extracted: number of weekly views by users from the October 2015 to May 2019 of the pages: Influenza, Febbre and Tosse (Flu, Fever and Cough, in English). Cross-correlation results are obtained as product-moment correlations between the two times series.

Results: Regarding the database with weekly data, temporal correlation was observed between the bulletin of ISS and Wikipedia search trends. The strongest correlation was at a lag of 0 for number of cases and Flu (r=0.7571), Fever and Cough (r=0.7501). The strongest correlation was at a lag of -1 for Fever and Cough (r=0.7501). The strongest correlation was at a lag of 1 for number of cases and Flu (r=0.7559), Fever and Cough (r=0.7501).

Conclusions: A possible future application for programming and management interventions of Public Health is proposed.

MeSH terms

  • Databases, Factual
  • Humans
  • Influenza, Human* / epidemiology
  • Public Health