Citat:
Ursprungligen postat av
OUFCompulsive
True, but the entire meaning of taking these 2000 tests are to measure against each other. It’s the same reason the vårdcentralen tests are not worthless despite their inherent bias, even if they aren’t truly representative of the population.
You’ll see maybe one false-negative in that amount of tests because such a low true-positive prevalence is always weighted towards false-positives. False-negatives will barely have any effect, if at all. There are simply too many true-negatives.
I don’t think the blood samples are that biased and I would argue they are far more representative of the population than samples from Vårdcentralen. Once again, it would be nice if they would stop fucking around and order in a large random-sample study so they can get their shit together. It would help in these arguments, too.
The percentage will always bounce around when dealing with such low prevalence, as I’ve brought up previously. The IFR for widespread infection is uniformly over 1% around the world, and the same will be found in Stockholm and Göteborg. Low-scale infections will lead to a lower IFR, so I wouldn’t be surprised if several regions in Sweden report less than 0.7% but extrapolating to the country as a whole I would imagine somewhere around 1% is correct, while remaining higher in the cities.
1%? 100 000 smittade? 5% IFR helt plötsligt?
Du menar alltså att vi i Sverige skulle ha lyckats hitta 50% av alla smittade genom testning? Vi har över 50000 fall nu och 33000 för en månad sedan. Med tanke på hur lite Sverige testat så vore det helt världsunikt. Att vi ens skulle lyckats hitta 33% av fallen är bara det
mycket osannolikt. Snarare max 10%. Du får nog fundera ett varv till på om dina siffror är ens i närheten av rimliga.
Minst 350 000 hade antagligen redan haft smittan i Sthlm, Västra Götaland och Skåne i slutet på Maj och idag lär det vara en bra bit över 600 000. Vilket även verkar vara en rimligare siffra utifrån antalet bekräftade fall och dödsfall.