Inferring temporal trends of multiple pathogens, variants, subtypes or serotypes from routine surveillance data

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Infectious disease surveillance often uses composite indicators (e.g., influenza-like illness) that combine multiple pathogens.
• Composite indicators may obscure individual pathogen dynamics and be affected by testing variations.
• Accurate temporal trend estimation is vital for disease monitoring and intervention assessment.

Purpose of the Study:

• To develop a general statistical framework for inferring individual pathogen temporal trends from composite surveillance data.
• To provide a robust method for disentangling pathogen-specific dynamics within mixed surveillance signals.

Main Methods:

• Developed a novel statistical framework for multi-pathogen trend inference.
• Applied the framework to diverse surveillance systems (influenza, SARS-CoV-2, dengue) across multiple countries.
• Validated the methodology across various epidemic scenarios and reporting resolutions (weekly, daily).

Main Results:

• Successfully inferred distinct temporal trends for individual pathogens from composite data.
• Demonstrated robustness across different pathogens, locations, and epidemic patterns.
• Showcased applicability to real-world surveillance data with varying complexities.

Conclusions:

• The developed framework accurately estimates individual pathogen trends from routine surveillance data.
• This methodology enhances the ability to monitor specific infectious diseases and evaluate public health interventions.
• Applicable to a wide range of pathogens and surveillance systems globally.