Our analysis on age-specific mortality and immunity patterns of SARS-CoV-2 has now been published @nature. We look at COVID-19 death data in 45 countries and 22 serostudies to investigate the consistency of infection & fatality patterns across countries https://www.nature.com/articles/s41586-020-2918-0

We find the age-distribution of reported COVID-19 deaths in younger individuals (<65) to be very consistent across countries but quite variable for deaths in older people.

Considering only deaths in <65s we see notable differences in IFR estimates as suggested by different seroprevalence studies. In some instances, applying IFR estimates from the lower end of the range to mortality data in hard-hit countries, would imply >100% attack rates!

We combine all the serostudies and age-specific mortality data to produce ensemble estimates of IFR and infected population proportions for each country, though it is crucial to consider the full range of uncertainty in these estimates as suggested by individual serostudies.

We estimate that high levels of transmission are likely to have occurred in some South American countries at the national-level but again, please consider the full range of uncertainty in these estimates suggested by individual serostudies when interpreting the results.

With huge thanks to all the team @hsalje, @SCauchemez, Gabriel Ribeiro Dos Santos, @fdlwang, @datcummings, @andrewazman, Juliette Paireau & Arnaud Fontanet!