A retrospective analysis of COVID-19 non-pharmaceutical interventions for Mexico and Peru: a modeling study

Abstract

We model the observed dynamics of COVID-19 in Mexico and Peru and explore the impact of hypothetical non-pharmaceutical interventions applied on key days of civic, religious, or political nature that increased contacts and transmission events. Using as a baseline the observed epidemic curve, we apply hypothetical reductions in the contact rates during the first year of the pandemic: i) near the beginning, ii) at the beginning of the second outbreak, and iii) end of the year. The effects of the interventions are different for Mexico and Peru but underlie the fact that strong early interventions do reduce the prevalence and, in general, allow for an epidemic evolution of relatively lower prevalence than interventions applied once the epidemic is underway. We provide evidence that key calendar days are good approximations of times when contact rates change and, therefore, are efficient periods for effective interventions particularly in places with low testing and lack of contact tracing. This has helped us to recreate different outbreaks of the COVID-19 disease dynamics in Mexico and Peru and explore the impact of hypothetical interventions that reduce the contact rate.