Impact of a nighttime curfew on overnight mobility

Abstract

Background: Among non-pharmaceutical interventions, individual movement restrictions have been among the most impactful methods for controlling COVID-19 case growth. While nighttime curfews to control COVID-19 case growth have been implemented in certain regions and cities, few studies have examined their impacts on mobility or COVID-19 incidence. In the second wave of COVID-19, Canada's two largest and adjacent provinces implemented lockdown restrictions with (Quebec) and without (Ontario) a nighttime curfew, providing a natural experiment to study the association between curfews and mobility. Methods: This study spanned from December 1, 2020 to January 23, 2021 and included the populations of Ontario (including Toronto) and Quebec (including Montreal). The intervention of interest was a nighttime curfew implemented across Quebec on January 9, 2021. Unadjusted and adjusted difference-in-differences models (DID) were used to measure the incremental impact of the curfew on nighttime mobility in Quebec as compared to Ontario. Results: The implementation of the curfew was associated with an immediate reduction in nighttime mobility. The adjusted DID analysis indicated that Quebec experienced a 31% relative reduction in nighttime mobility (95%CI: -36% to -25%) compared to Ontario, and that Montreal experienced a 39% relative reduction compared to Toronto (95%CI: -43, -34). Discussion: However, this natural experiment among two neighbouring provinces provides useful evidence that curfews lead to an immediate and substantial decrease nighttime mobility, particularly in these provinces' largest urban areas hardest hit by COVID-19.

Impact of non-pharmaceutical interventions, weather, vaccination, and variants on COVID-19 transmission across departments in France

“…could not be evaluated separately and were therefore collapsed into broader categories (lockdowns, reopenings and moderate restrictions). Interestingly, overnight curfews considerably reduced SARS-CoV-2 transmission, corroborating results obtained in French Guiana (up to 35% reduction in transmission rates) [ 27 ] and in Quebec, Canada (similar reductions in human mobility) [ 28 ]. Moreover, we found that curfews starting earlier in the evening (6/7 pm) had a larger impact on transmission than curfews starting later (8/9 pm).…”

Adaptive data-driven age and patch mixing in contact networks with recurrent mobility

“…The mobility matrix between patches was derived using private data on geolocation service usage among a sample of approximately 2% of mobile devices in Ontario [14] during January–December 2020. Appendix A.3 details the specific methods and assumptions used; to summarize: Each devices was assigned an approximate home location (152.9 m 152.4 m) based on the most common location during overnight hours for each calendar month.…”

Adaptive data-driven age and patch mixing in contact networks with recurrent mobility

“…Following through Eqs. (11)(12), we obtain X h gag ′ a ′ y , representing the total contacts formed within home pools. Then, the total type y contacts formed between populations P ga and P g ′ a ′ across all relevant mixing pools is given by the sum:…”

“…The mobility matrix B gg ′ between patches was derived using private data on geolocation service usage among a sample of approximately 2% of mobile devices in Ontario [12] during January-December 2020. Appendix A.3 details the specific methods and assumptions used; to summarize: Each devices was assigned an approximate home location (152.9 m × 152.4 m) based on the most common location during overnight hours for each calendar month.…”

“…Following through Eqs. (11)(12), we obtain X h gag ′ a ′ y , representing the total contacts formed within home pools. Then, the total type y contacts formed between populations P ga and P g ′ a ′ across all relevant mixing pools is given by the sum: 6 If residents of different patches might have relatively different numbers of contacts, a scaling factor could be applied here.…”