Assessing spatial variability in observed infectious disease spread in a prospective time-space series

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Traditional disease surveillance often focuses on early outbreak detection, but doesn't always account for non-uniform spatial and temporal spread.
• Identifying and evaluating anomalies (excess or decline) in disease incidence during an outbreak is crucial for effective public health response.

Purpose of the Study:

• To propose and formulate a novel hypergeometric probability model for investigating real-time anomalies in infectious disease incidence spread.
• To assess whether disease incidence is growing or declining more rapidly in specific geographic regions compared to others during an ongoing outbreak.

Main Methods:

• Developed a hypergeometric probability model for daily monitoring of geographically dispersed populations.
• Incorporated a time-varying baseline risk model using regularly updated disease incidence data.
• Evaluated deviations from expected frequencies, accounting for sampling fluctuations and unequal population sizes across regions.

Main Results:

• The model quantifies the probability of observed deviations in disease incidence being due to random chance.
• Demonstrated the model's application using spatiotemporal surveillance data for dengue and COVID-19 in Taiwan.
• Efficient R packages are available for implementing the model's formulae for large datasets.

Conclusions:

• The proposed hypergeometric probability model advances the investigation of infectious disease incidence spread anomalies.
• This method provides a robust tool for real-time evaluation of disease dynamics during outbreaks.
• The model is applicable to various infectious diseases and geographical scales, enhancing public health surveillance capabilities.