Fine-Scale Space-Time Cluster Detection of COVID-19 in Mainland China Using Retrospective Analysis

Area of Science:

• Epidemiology
• Public Health
• Geographic Information Systems (GIS)

Background:

• Understanding spatio-temporal disease patterns is crucial for identifying high-risk areas and potential causes.
• Retrospective analysis of disease incidence aids in public health interventions and resource allocation.

Purpose of the Study:

• To detect and analyze space-time clusters of COVID-19 in mainland China using family-level case data.
• To evaluate the impact of different maximum clustering radii on cluster detection.
• To assess the effectiveness of China's COVID-19 control strategies.

Main Methods:

• Retrospective analysis utilizing the space-time scan statistic.
• Detection of COVID-19 clusters based on family-level case onset dates.
• Varying the maximum clustering radius (e.g., 10 km, 100 km) to assess its effect on cluster identification.

Main Results:

• Forty-three clusters detected with a 100 km radius and 88 clusters with a 10 km radius (Dec 2019 - June 2020).
• Hubei province exhibited the highest number of clusters across scales; most clusters emerged in February.
• Well-developed regions with large populations and transport networks showed higher cluster prevalence.
• Onset date analysis potentially identified cluster start times seven days earlier than diagnosis date analysis.

Conclusions:

• Smaller clustering radii reveal finer-grained spatio-temporal disease patterns.
• China's control strategies were effective in limiting COVID-19 spread from Hubei.
• The methodology is applicable to other severely affected countries like the USA, India, and Brazil for precise clustering signals.