Doctor Google: Correlating internet search trends for epistaxis with metropolitan climates

Publication date: Available online 18 February 2019

Source: American Journal of Otolaryngology

Author(s): Aykut A. Unsal, Pariket M. Dubal, Julia A. Pfaff, Mark E. Friedel, Jean Anderson Eloy, Stilianos E. Kountakis

Abstract

Objective

Variation in weather patterns is often cited as a risk factor for epistaxis although robust studies investigating specific climate factors are lacking. As society is increasingly utilizing the Internet to learn more about their medical conditions, we explore whether Internet search activity related to epistaxis is influenced by fluctuations in climate.

Methods

Internet search activity for epistaxis-related search terms during 2012–2017 were extracted from Google Trends and localized to six highly populated cities in the US: New York, New York; Los Angeles, California; Chicago, Illinois; Houston, Texas; Philadelphia, Pennsylvania; and Atlanta, Georgia. Data were compared to local average monthly climate data from the National Centers for Environmental Information for the same time period.

Results

Spearmen correlations (r) were statistically strongest for dew point temperature (r_NewYork = −0.82; r_Philadelphia = −0.74; r_Chicago = −0.65; r_Atlanta = −0.49, r_LosAngeles = −0.3). This was followed closely by relative humidity (r_NewYork = −0.63; r_Philadelphia = −0.57; r_LosAngeles = −0.44; r_Atlanta = −0.42; r_Houston = −0.40) and average temperature (r_NewYork = −0.8; r_Philadelphia = −0.72; r_Chicago = −0.62; r_Atlanta = −0.45). Overall, correlations were most significant and predictable for cities with the greatest seasonal climate shifts (New York, Philadelphia, and Chicago). The weakest environmental factor was barometric pressure, which was found to be moderately positive in Atlanta (r_barometric = 0.31), Philadelphia (r_barometric = 0.30) and New York (r_barometric = 0.27).

Conclusions

Google Trends data for epistaxis-related search activity responds closely to climate patterns in most cities studied, thus underscoring the potential utility of Internet search activity data as a resource for epidemiologic study and for the identification of at risk populations.

http://bit.ly/2BDEfQ3