Any reliable source to say exactly what the fatality rate is?

[yarncakes]

I've been watching London Protest on Trafalgar Square and one said 99.94% survival rate, another saying 0.4% death rate. Which one is it? I've tried to research but they all just ask the same question that nobody seems to know the answer!

It's really age dependent so it's difficult to pin down an exact number

I don't think it's possible to know because you can never know just how many people have been infected - especially in the early weeks/months when testing wasn't widespread, and then all the very mild and asymptomatic cases that were not counted in any official records.

The figures I saw the other day for the UK as a whole looked to be just a tiny bit over 1% (I like 1.1 or 1.2%). A 'survival rate' is a difficult thing to measure (I work in infectious diseases) because there has to be a cut off...e.g. someone survives if they are still alive 6 weeks after onset of infection for example (I don't actually know that that's the definition for COVID, but as an example). But we know that some people die 3 months later, so they could be counted as 'surviving' but actually they don't.

Similarly, PHE counts a 'death' as within 28 days of a positive test, but we know people die later than 28 days, so that won't capture 100% of deaths either.

But the current PHE figures show about 1% of confirmed cases die in the UK. Those figures will not be the same globally where there isn't testing capacity or where health systems aren't even able to count deaths because people live to far from a clinic or can't afford to even come in.

Here is the PHE COVID tracker: coronavirus.data.gov.uk/

Yes, rather than focus on overall IFR, it's more useful to look at the age-mortality curve. Variation in overall IFR is largely due to population age distribution and who gets infected. So for example countries with a population of lower average age than others should have a lower overall IFR.

That said, if you want a guideline figure for Western Europe, US, Canada etc then there is a fair amount of consensus around an overall IFR figure of around 0.6% with a range of 0.5 to 1%. But as said, probably not that meaningful when it varies so much by age.

As another example, the large serological study done recently in the UK gave an IFR figure of 0.9%
www.imperial.ac.uk/news/201893/largest-study-home-coronavirus-antibody-testing/

For an age based breakdown of likely IFR, see the figures here for an example:
mobile.twitter.com/GidMK/status/1300644364117356545

Of course it's not just an outcome of you will either die or you will be fine, a minority will become seriously ill but recover and some may experience longer term illness, even from relatively mild or moderate cases. How many and how long term we don't yet know but it's a reason to remain cautious.

But the current PHE figures show about 1% of confirmed cases die in the UK.

Using deaths and confirmed cases from the PHE dashboard, we are at almost 10% deaths/cases BUT that is extremely misleading, because the number of confirmed cases is much smaller than the actual number of cases, as millions of infections have happened in the UK without being officially tested.

The IFR - Infection Fatality Rate - is the % of all those who are infected who die. We know that many cases of CV go undetected/undiagnosed; serology estimates for the UK suggest that more than 4m of us may have had it. Which would put the IFR at 1% or lower - but again, that is misleading, as it will differ very much depending (largely) on age.

A recent meta analysis of deaths in 6 affected areas ( Belgium, Geneva, Indiana, New York, Spain, and Sweden) suggested that

"The estimated IFR is close to zero for children and younger adults but rises exponentially with age, reaching about 0.3 percent for ages 50-59, 1.3 percent for ages 60-69, 4.6 percent for ages 70-79, and 25 percent for ages 80 and above"

@CoffeeandCroissant

Yes, rather than focus on overall IFR, it's more useful to look at the age-mortality curve. Variation in overall IFR is largely due to population age distribution and who gets infected. So for example countries with a population of lower average age than others should have a lower overall IFR.

That said, if you want a guideline figure for Western Europe, US, Canada etc then there is a fair amount of consensus around an overall IFR figure of around 0.6% with a range of 0.5 to 1%. But as said, probably not that meaningful when it varies so much by age.

As another example, the large serological study done recently in the UK gave an IFR figure of 0.9%
www.imperial.ac.uk/news/201893/largest-study-home-coronavirus-antibody-testing/

For an age based breakdown of likely IFR, see the figures here for an example:
mobile.twitter.com/GidMK/status/1300644364117356545

Of course it's not just an outcome of you will either die or you will be fine, a minority will become seriously ill but recover and some may experience longer term illness, even from relatively mild or moderate cases. How many and how long term we don't yet know but it's a reason to remain cautious.

That twitter link is helpful thank you

"This paper (preprint) assesses the age specificity of the infection fatality rate (IFR) for COVID-19 using seroprevalence results from eight national studies , regional studies of fifteen locations in Europe and the United States, and five countries that have engaged in comprehensive tracing of COVID-19 infections. The estimated IFR is close to zero for children and younger adults but rises exponentially with age, reaching 0.4% at age 55, 1.3% at age 65, 4.5% at age 75, and 15% at age 85. We find that differences in the age structure of the population and the age-specific prevalence of COVID-19 explain 90% of the geographical variation in population IFR. Consequently, protecting vulnerable age groups could substantially reduce the incidence of mortality."^
www.medrxiv.org/content/10.1101/2020.07.23.20160895v4

The recently leaked SAGE report was working on the basis of a 0.7% death rate iirc?