Like many, we've modelled potential deaths per country using two different case fatality rates (Hubei vs ex-Hubei) and then reversed the model to see what it implies about how many have had the #coronavirus. Suggests 6% for the UK vs 15% in Spain - note this data lags reality

This is based on the assumption that only 1 in 40 cases are “confirmed cases”. Ie 97.5% of people with the virus don’t show symptoms or get tested. That 1 in 40 could be in 1 in 10 or 1 in 100 which obviously changes the numbers significantly

This expert guesses 1 in 10 or 1 in 20 ..the thread he cites might take this from Imperial College model. It cites 15% prevalence in Spain, but on 28 March already

https://twitter.com/trvrb/status/1249414308355649536?s=21 https://twitter.com/trvrb/status/1249414308355649536

This study in California suggests only 1 in 50 are confirmed cases
https://twitter.com/scottgottliebmd/status/1251266310891352066?s=21 https://twitter.com/scottgottliebmd/status/1251266310891352066

A lot of criticism of that Santa Clara study in this thread and the replies https://twitter.com/wfithian/status/1252692357788479488?s=21 https://twitter.com/wfithian/status/1252692357788479488