A snapshot of a pandemic: The interplay between social isolation and COVID-19 dynamics in Brazil

Abstract

In response to the coronavirus pandemic, governments implemented social distancing, attempting to block the virus spread within territories. While it is well accepted that social isolation plays a role in epidemic control, the precise connections between mobility data indicators and epidemic dynamics are still a challenge. In this work, we investigate the dependency between a social isolation index and epidemiological metrics for several Brazilian cities. Classic statistical methods are employed to support the findings. As a first, initially surprising, result, we illustrate how there seems to be no apparent functional relationship between social isolation data and later effects on disease incidence. However, further investigations identified two regimes of successful employment of social isolation: as a preventive measure or as a remedy, albeit remedy measures require greater social isolation and bring higher burden to health systems. Additionally, we exhibit cases of successful strategies involving lockdowns and an indicator-based mobility restriction plan.

Keywords: COVID-19; human mobility; mobile geolocation; social isolation; spatial-temporal patterns.

Figure 1

Map of Brazilian regions and states abbreviations There are 26 states in Brazil plus the DF, which holds the national capital, Brasília. They are: Acre (AC), Alagoas (AL), Amapá (AP), Amazonas (AM), Bahia (BA), Ceará (CE), Espírito Santo (ES), Goiás (GO), Maranhão (MA), Mato Grosso (MT), Mato Grosso do Sul (MS), Minas Gerais (MG), Pará (PA), Paraíba (PB), Paraná (PR), Pernambuco (PE), Piauí (PI), Rio de Janeiro (RJ), Rio Grande do Norte (RN), Rio Grande do Sul (RS), Rondônia (RO), Roraima (RR), Santa Catarina (SC), São Paulo (SP), Sergipe (SE), and Tocantins (TO).

Figure 2

Temporal evolution of the mean relative isolation index of 1 week ago (dashed line) and the 7-day moving average of incidence (solid line) for the main cities in Brazil Colors refer to the five geographic regions that divide the country.

Figure 3

Dispersion between the mean relative isolation index of 1 week ago and the 7-day moving average of incidence for the main cities in Brazil Colors refer to the five geographic regions that divide the country. The linear correlation coefficient (R) is presented in each plot.

Figure 4

Temporal evolution of the mean relative isolation index of 1 week ago (dashed line) and the 7-day moving average of nowcasted incidence (solid line prior to the peak and dotted line after the peak) for 32 selected cities in Brazil Colors refer to the five geographic regions that divide the country. The vertical lines delimit the three stages of the upward cases, which represent 0%–25%, 25%–50%, and 50%–100% of the accumulated cases until the peak (upward phase).

Figure 5

Dispersion between the mean relative isolation index of 1 week ago and the 7-day moving average of incidence for 32 cities during the upward phase. Colors refer to the stage. The line is a LOESS . The value of S is the skewness coefficient of the normalized LOESS curve.

Figure 6

Dispersion between the length of each stage and the skewness coefficient of the smooth curve, and dispersion of the length of the upward phase and the skewness coefficient of the smooth curve. The linear correlation coefficient (R) and its p value is present on the plot.

Figure 7

Dispersion between the incidence on the peak and the skewness coefficient of the smooth curve. The linear correlation coefficient (R) and its p value is present on the plot.

Figure 8

Box plots of HDI, incidence on the peak, median isolation on each stage, length of each stage, and length of the upward phase for the cities that employed isolation as a remedy (negative skewness coefficient of the smooth curve) and cities that employed it as prevention (positive skewness) p values refer to the Wilcoxon test comparing the two groups of cities.

Figure 9

Temporal evolution of the 7-day moving average of incidence (green), and $R_t$ (blue) In red, we have the lockdown period. The dashed horizontal line is drawn for $R_t=1$. The $R_t$ values were truncated at three for a better visualization.

Figure 10

São Paulo (DRS I: São Paulo), Campinas (DRS XII: Campinas), and Votuporanga (DRS XV: São José do Rio Preto) Temporal evolution of the 7-day moving average of incidence (green) and $R_t$ (blue). The colored vertical bands represent the phases of São Paulo Plan. The dashed horizontal line is drawn for $R_t=1$. The $R_t$ values were truncated at three for a better visualization.