Web queries as a source for syndromic surveillance

Area of Science:

• Public Health
• Epidemiology
• Digital Surveillance

Background:

• Syndromic surveillance utilizes diverse data sources for disease outbreak detection and prediction.
• Traditional methods include laboratory verification and sentinel general practitioner reporting.

Purpose of the Study:

• To evaluate the efficacy of web search queries for estimating influenza epidemic timing and intensity.
• To compare the predictive power of web query data against established surveillance systems.

Main Methods:

• Analysis of influenza-related search queries from a Swedish medical website over two seasons.
• Development of partial least squares regression models to estimate influenza cases and influenza-like illness proportions.
• Comparison with laboratory-verified cases and general practitioner data.

Main Results:

• Certain influenza web queries mirrored patterns of traditional surveillance systems.
• Web query data demonstrated equal power in estimating the societal burden of influenza.
• Identified web queries as a valuable source for reaching individuals not yet seeking medical care.

Conclusions:

• Web search queries offer an accurate, cost-effective, and labor-efficient data source for syndromic surveillance.
• This approach provides unique insights into early-stage illness and public health trends.
• Digital data holds significant potential for enhancing real-time public health monitoring and response.