Quantifying the relationship between lockdowns, mobility, and effective reproduction number (Rt) during the COVID-19 pandemic in the Greater Toronto Area

Abstract

Background

The effectiveness of lockdowns in mitigating the spread of COVID-19 has been the subject of intense debate. Data on the relationship between public health restrictions, mobility, and pandemic growth has so far been conflicting.

Objective

We assessed the relationship between public health restriction tiers, mobility, and COVID-19 spread in five contiguous public health units (PHUs) in the Greater Toronto Area (GTA) in Ontario, Canada.

Methods

Weekly effective reproduction number (R _t) was calculated based on daily cases in each of the five GTA public health units between March 1, 2020, and March 19, 2021. A global mobility index (GMI) for each PHU was calculated using Google Mobility data. Segmented regressions were used to assess changes in the behaviour of R _t over time. We calculated Pearson correlation coefficients between GMI and R _t for each PHU and mobility regression coefficients for each mobility variable, accounting for time lag of 0, 7, and 14 days.

Results

In all PHUs except Toronto, the most rapid decline in R _t occurred in the first 2 weeks of the first province-wide lockdown, and this was followed by a slight trend to increased R _t as restrictions decreased. This trend reversed in all PHUs between September 6th and October 10th after which R _t decreased slightly over time without respect to public health restriction tier. GMI began to increase in the first wave even before restrictions were decreased. This secular trend to increased mobility continued into the summer, driven by increased mobility to recreational spaces. The decline in GMI as restrictions were reintroduced coincides with decreasing mobility to parks after September. During the first wave, the correlation coefficients between global mobility and R _t were significant ( p < 0.01) in all PHUs 14 days after lockdown, indicating moderate to high correlation between decreased mobility and decreased viral reproduction rates, and reflecting that the incubation period brings in a time-lag effect of human mobility on R _t. In the second wave, this relationship was attenuated, and was only significant in Toronto and Durham at 14 days after lockdown.

Conclusions

The association between mobility and COVID-19 spread was stronger in the first wave than the second wave. Public health restriction tiers did not alter the existing secular trend toward decreasing R _t over time.