Who infects whom in an epidemic?

Area of Science:

• Epidemiology
• Network Science
• Data Science

Background:

• Epidemiological surveillance typically analyzes infected populations by demographic groups (e.g., age, location).
• However, the dynamic interactions and spread patterns *between* these groups are often underutilized in epidemic analysis.
• Understanding group-specific epidemic dynamics is vital for effective public health strategies.

Purpose of the Study:

• To develop and validate a tool for analyzing epidemic heterogeneity across distinct population groups.
• To identify synchronicity and causality in disease transmission patterns among different demographic segments.
• To reveal which population groups initiate outbreaks and how they influence others.

Main Methods:

• Representing time-series data of infected individuals by group as a weighted graph to measure similarity.
• Applying community detection algorithms to partition the graph into synchronized groups.
• Utilizing sample cross-correlation to determine causal interactions between communities.

Main Results:

• The methodology successfully identified age groups exhibiting synchronized epidemic behavior.
• Causality analysis revealed which age groups influenced subsequent outbreaks in other groups.
• The tool was validated using COVID-19 and influenza A(H1N1) datasets.

Conclusions:

• The proposed tool effectively quantifies epidemic spread dynamics and inter-group relationships.
• Identifying synchronized and causally linked population groups enhances understanding of epidemic evolution.
• This approach provides actionable insights for public health decision-makers to target interventions effectively.