Predicting epidemics using search engine data: a comparative study on measles in the largest countries of Europe

Area of Science:

• Epidemiology
• Public Health Surveillance
• Digital Health

Background:

• Internet-based syndromic surveillance systems offer potential for early epidemic prediction.
• Existing literature lacks extensive research on forecasting measles using Internet data.

Purpose of the Study:

• To investigate the efficacy of Internet data, specifically Google Trends, for forecasting measles outbreaks.
• To analyze measles forecasting models using European Centre for Disease Prevention and Control (ECDC) data and Google Trends.

Main Methods:

• Utilized official measles data (2013-2018) from the European Centre for Disease Prevention and Control (ECDC).
• Acquired Google Trends data using Python scripts.
• Compared regression models for measles forecasting across five European countries.

Main Results:

• Google Trends effectively estimates and predicts measles cases regarding timing, volume, and spread.
• A strong correlation (R=0.779, p<0.01) was found between actual and predicted measles cases.
• Mean standard error was low (45.2 or 12.19%), though deviations occurred in countries with lower measles incidence.

Conclusions:

• Estimating measles cases via Google Trends yields acceptable results for robust outbreak prediction up to two months ahead.
• Python scripts integrated into surveillance systems can effectively track epidemics like measles.