JMIR Publications

ABSTRACT

Background:

Recognizing the pivotal role of human mobility in the spread of infectious diseases, social distancing policies are promptly initiated during emerging epidemics. However, significant gaps remain in understanding how mobility influences geographical spread of infectious diseases and at what scale to design predictive models and implement control policies.

Objective:

Addressing these questions is crucial for characterizing the key mechanisms of geographical diffusion and improving the reliability of models for outbreak response.

Methods:

We analyze high-resolution mobility data from mobile app usage from SafeGraph, mapping daily connectivity between US counties to grasp spatial clustering and temporal stability. Integrating this into transmission models, we replicate SARS-CoV-2’s first wave invasion, assessing mobility’s spatio-temporal impact on disease predictions.

Results:

Temporal stability is observed in intercounty connectivity annually, unaffected by early pandemic mobility restrictions in April 2020. Spatially, 104 US mobility-based clusters show high internal mobility but sparse connections externally. This suggests stable, highly connected intercounty mobility at sub-state levels. While static mobility data captures infection dynamics effectively, county-scale data is crucial for spatial disease diffusion prediction.

Conclusions:

Intercounty mobility remained largely unaffected beyond Spring 2020 lockdowns, explaining COVID-19’s broad initial US outbreak. Geographically dispersed outbreaks strain national health resources, requiring complex metapopulation models. Our findings inform such model designs to balance high disease predictability with low data requirements. These insights are pivotal for strategic planning and resource allocation during future emerging epidemics, enhancing the robustness and responsiveness of epidemic modeling for public health interventions.