Impact of weekday and weekend mobility and public policies on COVID-19 incidence and deaths across 76 large municipalities in Colombia: statistical analysis and simulation

Abstract

Background: Despite widespread restrictions on residents' mobility to limit the COVID-19 pandemic, controlled impact evaluations on such restrictions are rare. While Colombia imposed a National Lockdown, exceptions and additions created variations across municipalities and over time. METHODS: We analyzed how weekend and weekday mobility affected COVID-19 cases and deaths. Using GRANDATA from the United Nations Development Program (UNDP) we examined movement in 76 Colombian municipalities, representing 60% of Colombia's population, from March 2, 2020 through October 31, 2020. We combined the mobility data with Colombia's National Epidemiological Surveillance System (SIVIGILA) and other databases and simulated impacts on COVID-19 burden. RESULTS: During the study period, Colombians stayed at home more on weekends compared to weekdays. In highly dense municipalities, people moved less than in less dense municipalities. Overall, decreased movement was associated with significant reductions in COVID-19 cases and deaths two weeks later. If mobility had been reduced from the median to the threshold of the best quartile, we estimate that Colombia would have averted 17,145 cases and 1,209 deaths over 34.9 weeks, reductions of 1.63% and 3.91%, respectively. The effects of weekend mobility reductions (with 95% confidence intervals) were 6.40 (1.99-9.97) and 4.94 (1.33-19.72) times those of overall reductions for cases and deaths, respectively.

Conclusions: We believe this is the first evaluation of day-of-the week mobility on COVID-19. Weekend behavior was likely riskier than weekday behavior due to larger gatherings and less social distancing or protective measures. Reducing or shifting such activities outdoors would reduce COVID-19 cases and deaths.

Keywords: COVID-19; Case; Colombia; Death; Lockdown; Mobility; Municipality; Population density; SARS-CoV-2; Weekend.

Fig. 1

Mobility change over time compared to new COVID-19 cases nationally. Notes: The x-axis, denoting time, corresponds to the 34.9 complete weeks in the data set. The labels denote key dates in Colombia regarding exceptions to staying in mandatory isolation. Time progresses from left to right of the figure, with more exceptions declared allowing citizens to reopen businesses or travel for essential needs. The start of mandatory isolation occurred on March 25, 2020, with 34 exceptions permitting select groups to leave isolation for essential services. Other key dates included May 11, 2020, with 46 exceptions to mandatory isolation to engage in physical activity in municipalities with a low number of cases. On June 1, 2020, the government declared 43 exceptions to the stay-at-home measures that included, for example, reopening commerce places such as hair salons. On August 1, 2020, mandatory isolation for all departments in Colombia changed to selective isolation status, ending mandatory stay-at-home measures. Source: Authors’ calculations based on GRANDATA and SIVIGILA

Fig. 2

Impact of density on weekly mobility during varying COVID mobility restrictions. Notes: The right- most vertical bar (to the right of the white bar) shows the municipality’s overall mean. Source: Authors’ calculations based primarily on GRANDATA and information from DANE, the National Statistics Department of Colombia