COVID-19 effective reproductive ratio determination: An application, and analysis of issues and influential factors

An essential indicator of COVID-19 transmission is the effective reproduction number ( R _t ), the number of cases which an infected individual is expected to infect at a particular point in time; curves of the evolution of R _t over time (transmission curves) reflect the impact of preventive measures and whether an epidemic is controlled. We have created a Shiny/R web application ( <ext-link xmlns:xlink="http://www.w3.org/1999/xlink" ext-link-type="uri" xlink:href="https://alfredob.shinyapps.io/estRO/">https://alfredob.shinyapps.io/estRO/</ext-link> ) with user-selectable features: open data sources with daily COVID-19 incidences from all countries and many regions, customizable preprocessing options (smoothing, proportional increment, backwards distribution of negative corrections, etc), different MonteCarlo-Markov-Chain estimates of the generation time or serial interval distributions and state-of-the-art R _t estimation frameworks (EpiEstim, R0). We have analyzed the impact of these factors in the obtained transmission curves. We also have obtained curves at the national and sub-national level and analyzed the impact of epidemic control strategies, superspreading events, socioeconomic factors and outbreaks.

We conclude that country wealth and, to a lesser extent, mitigation strategies, were associated with poorer epidemic control. Dataset quality was an important factor, and sometimes dictated the necessity of time series smoothing. We couldn’t find conclusive evidence regarding the impact of alleged superspreading events. In the reopening phase, outbreaks had an impact on transmission curves. This application could be used interactively as a tool both to obtain transmission estimates and to perform interactive sensitivity analysis.