Google scholars developed a model that links flu queries from their search engine to actual influenza incidence rates.
In Missouri, and nationally, Google predicts flu symptoms peaked in mid-Oct. and are on the decline, at least for now.
The Oct. spike was reportedly due mostly to the outbreak of H1N1 -- and that's why there were lower intensities reported in Florida where the high number of retirees weren't impacted as much, since H1N1 targets the younger crowd.
Another spike may be reported at the beginning of 2010, similar to 2008 when flu intensity peaked in Feb.
So how accurate is Google's model? The results of a study Google's researchers published in Nature shows the results very closely align with actual data collected by the Centers for Disease Control (R^2>0.9).
Related Links:
Google FluTrends
FluTrend Method in Nature
Posts
0 comments:
Post a Comment