Near Real-Time Measures of Interregional Mobility Restrictions on COVID-19 Transmission

The global SARS-CoV-2 pandemic prompted nations to implement mobility limitations to curb virus spread. In Chile, targeted interregional measures were employed to mitigate the social and economic costs. Here, we employ a novel real-time methodology to assess the impact of such mobility restrictions on epidemic control. Leveraging telecom-derived eXtended Detail Records (XDR) and official COVID-19 epidemiological data, we estimate interregional mobility and disease prevalence. Employing Bayesian adjustments, we compare different mobility restriction scenarios: business-as-usual (BAU), initial measures, and total lockdown. Mobility reductions significantly curtailed cases and risk across regions. Even modest mobility declines under total lockdowns considerably lowered imported cases. The high-risk Santiago Region, a national source of infections to other regions, demonstrated lowered risk due to mobility restrictions. Our approach facilitates rapid regional insights for informed policy responses. Author summary The implementation of mobility restrictions is a coarse approach to limiting pathogen transmission during the early stages of an epidemic. Disentangling the effects of the reduced population mixing and the effects of new seeding events can be challenging to perform in real time, making any evaluation of the efficacy of such interventions a post hoc exercise. Here we present an approach that uses aggregated data from individual mobile phone locations within the national network of antennas to statistically correct for changes in human mobility and estimate the changes in COVID-19 transmission across regions in Chile during the first wave of the pandemic. We estimate a risk score that can be implemented nearly in real time to estimate the effects of mobility restrictions on viral transmission and highlight that important urban centres that seed other regions, like the Santiago Region, are at low risk of experiencing increased viral transmission due to new seeding events from other regions.