In response to the COVID-19 pandemic, countries have sought to control transmission of SARS-CoV-2
by restricting population movement through social distancing interventions, reducing the number of
contacts.
Mobility data represent an important proxy measure of social distancing. Here, we develop a
framework to infer the relationship between mobility and the key measure of population-level disease
transmission, the reproduction number (R). The framework is applied to 53 countries with sustained
SARS-CoV-2 transmission based on two distinct country-specific automated measures of human
mobility, Apple and Google mobility data.
For both datasets, the relationship between mobility and transmission was consistent within and
across countries and explained more than 85% of the variance in the observed variation in
transmissibility. We quantified country-specific mobility thresholds defined as the reduction in
mobility necessary to expect a decline in new infections (R<1).
While social contacts were sufficiently reduced in France, Spain and the United Kingdom to control
COVID-19 as of the 10th of May, we find that enhanced control measures are still warranted for the
majority of countries. We found encouraging early evidence of some decoupling of transmission and
mobility in 10 countries, a key indicator of successful easing of social-distancing restrictions.
Easing social-distancing restrictions should be considered very carefully, as small increases in contact
rates are likely to risk resurgence even where COVID-19 is apparently under control. Overall, strong
population-wide social-distancing measures are effective to control COVID-19; however gradual
easing of restrictions must be accompanied by alternative interventions, such as efficient contacttracing, to ensure control.