Evaluating the effects of population mobility on transmission during the COVID-19 pandemic

Abstract Changes to community mobility during the COVID-19 pandemic, together with unprecedented availability of surveillance and mobility data, provide a unique opportunity to quantify the effect of mobility on respiratory infection transmission. We present a renewal model that incorporates epidemiological mechanisms influencing SARS-CoV-2 transmission along with a non-mechanistic scaling process. We compare models with and without mobility using epidemic data from 122 countries and quantify the influence of mobility through reductions in non-mechanistic scaling. In several countries, this non-mechanistic process closely tracks mobility trends and varies substantially less when mobility is incorporated. This is especially evident in Western Europe and North America, but not in several countries of West Africa and the Arabian Peninsula. Google and Facebook data produce similar findings, and parks is frequently assigned the lowest weight in the multi-location Google analysis. Mobility trends are strongly aligned with short-term transmission dynamics in many high-income settings, although between-country heterogeneity remains.