Accurate regional influenza epidemics tracking using Internet search data.

Area of Science:

• Epidemiology
• Public Health Informatics
• Computational Biology

Background:

• Accurate regional influenza tracking is crucial for public health decision-making during outbreaks.
• Internet search data offers potential for real-time influenza surveillance but faces challenges with data quality and structure at regional levels.
• Existing methods for regional influenza tracking using internet data often yield unsatisfactory performance.

Purpose of the Study:

• To develop and evaluate a novel method, ARGO2 (2-step Augmented Regression with GOogle data), for accurate, high-resolution, real-time regional influenza tracking.
• To assess the performance of ARGO2 by comparing it with existing internet-data-based regional influenza tracking methods.
• To demonstrate the reliability and robustness of ARGO2 for public health surveillance.

Main Methods:

• ARGO2 combines Google search data at national and regional resolutions with traditional influenza surveillance data from the Centers for Disease Control and Prevention (CDC).
• The method employs a 2-step augmented regression approach to integrate diverse data sources and resolutions.
• Performance evaluation was conducted across all US regions for the period of March 2009 to March 2018.

Main Results:

• ARGO2 demonstrated highly competitive performance across all US regions compared to existing internet-data-based methods.
• The proposed method achieved a 30% reduction in error compared to the best alternative method tested.
• ARGO2 proved reliable and robust in its tracking capabilities.

Conclusions:

• ARGO2 offers a powerful and flexible tool for accurate, real-time regional influenza tracking.
• The method's ability to integrate data from multiple sources and resolutions enhances its utility for public health.
• ARGO2 has potential applications for tracking other regional or local social, economic, or public health events.