Identifying the main drivers of transmission in the early phase of the COVID-19 pandemic in Portugal

Abstract

In this study, we employed a modeling approach to describe how changes in age-specific epidemiological characteristics, such as behaviour, i.e. contact patterns, susceptibility and infectivity, influence the basic reproduction number R0\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_0$$\end{document}, while accounting for heterogeneity in transmission. We computed sensitivity measures related to R0\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_0$$\end{document}, that describe the relative contribution of each age group towards overall transmission. Additionally, we proposed a new indicator that provides the expected relative change in the number of new infections, given a public health intervention. Studying the outbreak of COVID-19 in Portugal during March 2020, our results show that the main drivers of transmission were individuals 30-59 years old. Furthermore, by studying the impact of imposed changes in susceptibility and infectivity, our results demonstrate that a 10% decrease in susceptibility for the 30-39 years old results in a incidence reduction after 3 generations of approximately 17% in this age group and 4-6% reduction as an indirect effect in the remaining age groups. The presented methodology provides tools to inform the allocation strategy of mitigation measures in an outbreak of an infectious disease. Its inherent versatility enables the easy incorporation of data specific to various populations, facilitating a comparative analysis of epidemic control effects across different countries.