Revealing regional disparities in the transmission potential of SARS-CoV-2 from interventions in Southeast Asia

Abstract

SARS-CoV-2 is a new pathogen responsible for the coronavirus disease 2019 (COVID-19) outbreak. Southeast Asia was the first region to be affected outside China, and although COVID-19 cases have been reported in all countries of Southeast Asia, both the policies and epidemic trajectories differ substantially, potentially due to marked differences in social distancing measures that have been implemented by governments in the region. This paper studies the across-country relationships between social distancing and each population's response to policy, the subsequent effects of these responses to the transmissibility and epidemic trajectories of SARS-CoV-2. The analysis couples COVID-19 case counts with real-time mobility data across Southeast Asia to estimate the effects of host population response to social distancing policy and the subsequent effects on the transmissibility and epidemic trajectories of SARS-CoV-2. A novel inference strategy for the time-varying reproduction number is developed to allow explicit inference of the effects of social distancing on the transmissibility of SARS-CoV-2 through a regression structure. This framework replicates the observed epidemic trajectories across most Southeast Asian countries, provides estimates of the effects of social distancing on the transmissibility of disease and can simulate epidemic histories conditional on changes in the degree of intervention scenarios and compliance within Southeast Asia.

Keywords: COVID-19; SARS-CoV-2; time-varying reproduction number.

Figure 1.

COVID-19 cumulative incidence per 10 000 across Southeast Asian countries from March 2020 to May 2020. (Online version in colour.)

Figure 2.

SARS-CoV-2 cases across time and time-varying reproductive number model fits across Southeast Asia from March 2020 to May 2020. Red lines represent the fitted case counts with lighter colours representing greater degrees of intervention. Blue points represent observed case counts. Highlighted areas represent time points where social distancing policies were not implemented. (Online version in colour.)

Figure 3.

Estimated time-varying reproduction number for SARS-CoV-2 across time across Southeast Asian countries from March 2020 to May 2020. Red lines represent the fitted time-varying reproduction number with the shaded area representing the implied density estimate. Blue points represent observed time-varying reproduction number. Highlighted areas represent time points where social distancing policies were not implemented. (Online version in colour.)

Figure 4.

Estimated coefficient size and 95% credible intervals of changes from baseline in residences on the time-varying reproduction number across Southeast Asian countries. Points represent the posterior mean estimate, black lines represent the 95% credible intervals. Black lines which cross the threshold represented in blue signifies that the 95% credible interval contains 0. (Online version in colour.)