Assessing seasonal dynamics of Guillain-Barré syndrome with search engine query data

In previous studies, data deriving from Google Trends showed promising correlation with disease incidence trends assessed with public health control systems. The aim of this work is to use search engine query data to investigate seasonal dynamics in Guillain-Barré syndrome (GBS) in the USA.

Average Google monthly search volumes for GBS from 2008 to 2017 were analysed for the USA overall and on regional base with generalized estimating equation models. Association with monthly historical temperature variations was tested.

Monthly search volume for GBS displayed the greatest positive anomaly for October, clustering with September and November. Region-wide analysis confirmed this pattern and showed secondary spring (Feb/Apr) subpeaks in Pacific and Midwest. Association of GBS search volume with month-to-month temperature variations showed J-shaped relationship, with the highest peak occurring in months with greatest temperature falls, and subpeak in months with sharpest temperature rises.

This study represents the first approach in investigating digital epidemiology of GBS and establishing possible links with traditional epidemiology. Cold season GBS peak has been observed by some traditional studies; hypothetical pathogenic relationship with infectious antecedents is supported from finding GBS peaks clustering with greatest temperature change. Further studies are needed to compare these findings to traditional public health approaches.