Moderator
Posts: 461 from 2003/7/21
Quote:takemehomegrandma wrote:
Here is an interesting article on the subject, based on data from Iceland.
Iceland is a unique country in that they have conducted such extensive testing of the entire population, not only of people with suspected symptoms and hospitalized patients like most counties does. This gives a much better and more complete picture of the outbreak and opens up for a deeper understanding about where we are heading.
https://medium.com/@ali_razavian/covid-19-from-a-data-scientists-perspective-95bd4e84843bHis model seems plausible.
No, it's not. There are 2 assumptions that are simply false: he uses mortality rate of 1.5% (which is the current WHO estimate) with the dataset of the currently tested positive. No, the 1.5% number is the result of applyng a similar data processing to the currently tested positive, death rate with respect to the people tested positive is much higher (15% in Belgium, 13% in Italy and UK, 12% in France, 10% in Spain, 5.5% in the USA). If you apply the same adjustement twice, you are not adjusting. You are getting completely wrong numbers. Even the second adjustement (96% of the cases gets unnoticed) is likely wrong: current consensus in Italy, for instance, is that there are 8 to 10 cases that go unnoticed for every person tested positive, which means that the number he should use is closer to 90% than 96% (which again makes some difference). And again the mortality of the CoVid appears to be somewhere between 1 and 2%, which is what the WHO says.
Heck, he should at least have taken a look at the data he was starting from (Iceland): mortality there is 0.5%, likely because positives gets controlled and treated very early and around 15% of the population has been tested. Experience is that if elder people wait until oxigen saturation is below 90% and they actually need oxigen before accessing the hospital, death is a much more likely outcome, so they should check saturation (but a lot of the times they do not know for sure to be infected before accessing the hospital).
Bottom line: it will likely take more time before there is no danger anymore, and death toll might continue to increase.
Kind regards,
Andrea