Estimation of country-level basic reproductive ratios for novel Coronavirus (COVID-19) using synthetic contact matrices

Joe Hilton; Matt J Keeling

The predictions we have made are limited by two main elements. The first is the accuracy of the estimated contact matrices; although there are known issues (as discussed in [3] ) they remain our best estimate of agestructured contacts to date. Unfortunately, not all countries have an associated mixing matrix; many counties in Africa do not have the underlying demographic data to support the generation of the associated mixing matrix. Secondly, in inferring the age-dependent susceptibility, we are effectively generating a matrix R i,j which determines the distribution of secondary confirmed cases in terms of current confirmed cases. We are therefore assuming that either younger individuals are unlikely to infected or if infected they are generally asymptomatic and play a minor role in onward transmission. If such asymptomatic infections transmit equally to symptomatic cases, then the scaling of the reproductive ratio is expected to be closer to Figure 1 . Finally, it is worth stressing that these projections only inform about early phase of the outbreak in the absence of controls; the rapid and effective use of non-pharmaceutical interventions (contact-tracing, self-isolation and movement controls) can substantially reduce the reproductive ratio.