Coronavirus (COVID-19) Discussion Thread

[SvennoJ]

JRPGfan said:

"But it's still a scary thought that if you happen to get tested positive in Brazil, UK, Belgium and Sweden, you're already up to a 1 in 5 chance of not surviving." - SvennoJ

Thats because alot of places they dont test "healthy" looking individuals without any signs of being infected.
Alot of people have this virus, and arnt even aware of it. These type of people just assume they dont have it, so why would they bother to go get tested?

The people that go get tested, are usually the ones that already have symptoms (ei. their sickness is turning out worse than most).
or people that have bad symptoms... like alot of people will just assume "I can pull through this, its not so bad" and they dont go get tested.

Ones that get tested, are often people that have such horrible symptoms they go to the hospital, and explain their symptoms to the hospitals, and then get tested there.

Even those tents were you can just drive in and get tested, usually require you to have a doctor giveing you the "okay go get tested".
These doctors usually only give them to people with symptoms too.

So basically, we mainly test those who already are showing signs of bad outcomes.
The remaining ones with better outcomes, are just walking around ignorantly bliss, not knowing they are infected and infecting others.

"ignorance really is bliss" with this virus.
You want to go take a anti body test, find out you have antibodies, unaware that you were ever even infected.
Thats the best possible outcome to a antibody test imo.

"Thanks, yep at least 0.7% I can agree with. Closer to 1.0% would be my guess, although the older estimate of 1.2% is also still in range." - SvennoJ

Yeah I think John Campbell is over optimistic, about it being only ~0.7%.
Given that most places excess mortality are higher than reported covid19 deaths.

It's hard to get to case fatality rates with all the different ways testing happens. I just used 20 days as an estimate.

It's 18.5 days on average from onset to death.
It's 5.5 days on average from infection to onset (incubation period)

People actually showing symptoms don't go to get tested until well after the incubation period, likely days after onset.
Tests can take days to come back and even longer to get included in the stats.
However, many tests are also directed / reserved for front line workers regardless of symptoms, contacts of confirmed cases and targeted testing of for example nursing homes.

Thus you have a mix of catching people during their incubation period to catching those that already had it for weeks.

Someone with bad symptoms, gets tested a week into onset, test results gets added 10 days after onset, patient dies 8 days after their results were tallied.
Or a front line worker, gets tested more efficiently and caught early, 3 days after infection, dies 20 days after their results were tallied.

Looking 20 days back was probably a bit far so my CFR estimates are still a bit on the high side for countries where there is still a lot of growth. (Wouldn't make much of a difference for those on the way down for the past weeks)

If I adjust my top 5 for 15 day period to compare with:

3. Brazil 20.5% -> 16.5%
5. UK 20.1% -> 18.5%
17. Belgium 18.7% -> 18.2%
24. Sweden 18.0% -> 16.7%
7. France 16.8% -> 16.5%

Anyway, all it shows is which countries miss the most cases.

If the IFR in the UK is indeed 0.94% (based on that early report of 510K possible deaths when 81% gets infected), the UK is only catching 1 in 20 infections. Which of course assumes again that the spread of the infection is uniform, which it probably isn't.

Now in Germany, (still 5.0% CFR comparing deaths to reported cases 15 days ago), that would mean Germany is 'only' missing 4 out of 5 infections assuming everything else is equal. (which it won't be)

Btw tests per million are pretty much equal between UK and Germany, who gets tested and where the virus spreads have a big influence.


It's impossible to make any generalizing statements of the fatality rate or how many more infections there are were. But it's a good bet it's at least 0.7% in western countries with a higher average age. (compared to developing countries) We'll need a lot more truly random anti body test screenings to get a decent idea of how many cases could have been missed in other areas.




Update for Ontario, 413 new cases, which raises the 3 day avg week over week comparison to 114%, still getting worse. We've had week over week growth for 4 days now. Might it still be the effect of Mother's day or simply catching up on tests since we're now down to 5K tests pending (while before it was as high as 17K tests pending)

CFR for Ontario (deaths / total cases 15 days ago) is 10.6%. So if actual fatality rate is under 1% we're still missing 10 out of every 11 cases.

The weather is offering a good opportunity to test the effect of UV light. UV index yesterday and today was at 9, severe. Blue skies, dry, burning sun, plenty people out. Will next week be better (UV light hampering transmission) or worse (more people out spreading infection). Or will any change be too small to draw any conclusions (likely)

Last edited by SvennoJ - on 21 May 2020