Visualizing the Network Structure of COVID-19 in Singapore

Area of Science:

• Epidemiology
• Network Science
• Public Health

Background:

• Infectious diseases, including coronavirus disease 2019 (COVID-19), often spread through established social networks.
• Current network models primarily focus on micro-level interactions, leaving a gap in understanding higher-level disease transmission patterns (meso-structure).
• Meso-structure describes disease spread across aggregated groups like organizations, locales, and events.

Purpose of the Study:

• To investigate and visualize the meso-structure of disease transmission using real-world contact tracing data.
• To identify the role of aggregated clusters in driving epidemic dynamics.
• To assess the heterogeneity in disease spread originating from different clusters.

Main Methods:

• Utilized publicly available contact tracing data from Singapore.
• Employed visualization techniques to map disease spread at the meso-level.
• Analyzed the number of infection chains generated by individual infection clusters.

Main Results:

• A single, highly central infection cluster was identified as the source of approximately seven to eight infection chains.
• This central cluster accounted for about 60% of all non-imported cases during the study period.
• No other infection cluster generated more than two infection chains, indicating significant heterogeneity in transmission.

Conclusions:

• The study demonstrates that the meso-structure of social networks, specifically the pattern of disease spread among clusters, is critically important for epidemic dynamics.
• Heterogeneity in cluster-driven transmission suggests targeted interventions at central clusters could be highly effective.
• Understanding and visualizing meso-structure is essential for effective infectious disease control strategies.