Google Flu Trends: Spatial Correlation with Influenza Emergency Department Visits

Joseph Klembczyk, Mehdi Jalalpour, Scott Levin, Raynard Washington, Jesse M. Pines, Richard Rothman, Andrea Dugas

Abstract


Google Flu Trends (GFT) is an internet search query-based application that has been proven to add value to influenza surveillance and forecasting tools. Previous validation studies have focused on national or regional predictions. While these results have been promising, GFT has yet to be extensively validated at the city level. The AHRQ has provided weekly data for influenza-related emergency room visits across 19 cities. Correlation coefficients with city-level GFT range from .67 to .93 with a median of .84. Characterizing the effectiveness of GFT at the local level is crucial to its integration into new surveillance and prediction tools.


Full Text:

PDF


DOI: https://doi.org/10.5210/ojphi.v7i1.5753



Online Journal of Public Health Informatics * ISSN 1947-2579 * http://ojphi.org