324 Covid Deaths Yesterday, And 2,095 Cases Are They Easing Lockdown Too Early ?

[Gromit]

https://coronavirus.data.gov.uk/

[fiction-factory]

//ff, Today's figures for the UK (latest update) are 38,489 deaths from a total of 274,762 - by my reckoning (ratio of the two) that is just a hair above 14% turning into deaths??

KARL. I think your calculation is pretty meaningless. The deaths figure is meaningful and can be justified by the body count (accepting it may understate excess mortality/overstate OF Covid) . Your denominator of 264762 represents what? It's only known infections isn't it.

[jim360]

I'd have to agree with ff. The 14% is the Case Fatality Rate and is also dependent on how many people were tested, what the criteria for testing are, and so on. Particularly early on in the progression of the disease, the UK was only testing those who were likely to be positive -- which isn't at all unreasonable -- and ignoring those who were either asymptomatic or only slightly suffering.

A more important measure would be the Infection Fatality Rate, which takes into account also the untested (and is therefore only possible to estimate). I think for the UK the total number of people to have had the disease is variously estimated to be between 7% and 30%, although I think the consensus is probably somewhere in the region of 15% nationwide. This is almost exactly 10 million people so even if it is wrong (which is almost certainly true) then at least it provides a useful order of magnitude estimate, and suggests an IFR of around 0.4% (1 in 250), which is in line with worldwide averages and other studies.

This is a back-of-the-envelope estimate and comes with a hefty uncertainty, but it is still more meaningful than the 14% figure, which relies too heavily on confirmed cases. However, the most truly meaningful measure remains excess mortality, which is significant and horrific.

[Deskdiary]

Jim, you’re wrong; the correct POV is how many have died, and it ain’t that many.

You may be correct with all the numbers and stats you are throwing around, but ‘on the ground’ and in the the real world you couldn’t be more wrong.

The fact is this virus is not affecting very many people. It just isn’t, and no amount of spin or clever stats will affect that fact.

[jim360]

The growth rate of the virus in March and April strongly suggests that if we had carried on as normal then, yes, we would be looking by now at well in excess of 100,000 extra deaths, with more to come. The virus may not have affected "many people" (although how you come by that, since even falling sick is unpleasant and is likely to have lasting health effects eg lung damage, is anybody's guess). But that is only because we have taken the threat seriously.

It is a sad irony that the relative success of the lockdown is now used as evidence against its ever having been needed. That's what you're doing. If you open up the country, if you ignore the threat, then you'll see just how dangerous it can be. Let's hope that we never find out.

[jim360]

Also "the correct POV is how many have died, and it ain’t that many."

How does anyone look at excess mortality running at 67% or so, corresponding to tens of thousands of extra people dead in March/April/May than would be normal, and suggest that it "ain't that many"? That sort of death rate doesn't normally happen, except in (serious) wars or in serious pandemics. We are in the second of these. You can't ignore it.

[10ClarionSt]

How does anyone know that there would have been in excess of 100,000 deaths? How do you know that Jim? On your own admission, it is a suggestion. Nobody knows but some are pretending that they do.

[jim360]

I didn't say I knew, I said that the evidence "strongly suggests" it. Nobody can know. But it is difficult to see how the death toll could have been less if we had carried on as normal and allowed a far greater number of people to catch the disease by now. Add to that the risk of exhausting hospital capacity, leading to knock-on effects, and, yes, the death toll would have been higher.

One feature of the weekly death rate data is that it's possible to see an apparent gap between the number of excess deaths and the number of Covid-related deaths. As I understand it, this gap isn't understood fully yet, but I think some are attributing it in part to people not going to hospital for conditions that would be treatable but are otherwise fatal. Again, then, imagine what would have happened -- indeed, what very likely will happen, in other countries -- had our healthcare system been overloaded with the sick and dying. It wouldn't bear thinking about.

[10ClarionSt]

Well, we were told to protect the NHS, when my understanding was that the NHS was there to protect us. More brainwashing. Say it often enough and people will fall for it. As some on here clearly have done.

[jim360]

NHS workers have also died in the crisis, and the message of "protect the NHS" was, presumably, a sentence fragment that could be readily understood as "make sure that the NHS isn't overwhelmed so that it can continue protecting us". So I'm not sure what point you're trying to make.

[KARL]

ff and Jim, If we ditch the published figures, as produced by the UK authorities, in favour of ones we pick, naturally the result will be what we prefer. No discussion, nothing worth discussing.

[fiction-factory]

I am not suggesting that we ditch the figures, Karl, and I doubt jim is either, but we should interpret them properly and not use them to do the calculations you were doing to compare countries. The numerator, deaths figure is broadly comparable; the denominator reflects the testing processes

[jim360]

It's more a case of recognising their context. Take Brazil as another example. It has reported, now, half a million cases and about 29,000 deaths, but has tested less than a million people. It is entirely sensible to suggest that the true number of cases is rather a lot higher, since around 99.5% of all Brazilians won't have been tested. In that case, the only fair way to compare like with like is to find a measure that isn't dependent on how many tests are being carried out or on who is tested in the first place. Hence the attempts to estimate the Infection Fatality Rate, or, one step further, by stripping out any direct reference to confirmed/suspected cases and just looking at the excess mortality. Ironically, then, this is actually the way to generate the fairest comparison, and hasn't anything to do with a preference other than that.

As another case in point, consider the picture in Russia, which, by the same measure, is reporting a Case Fatality Rate of 1.1%, or something like 13 times "better" on the face of it than in the UK. Assuming that the figures are entirely accurate, which is already a bit of a stretch, there are at least three points that make the comparison misleading:

1. Russia has, it claims, done more testing at a greater rate per head than the UK;
2. The disease is at an earlier stage in its progression in Russia as compared to the UK, so it is likely that their death rate will rise for a while when the UK's is falling (hopefully).
3. Most importantly, the criteria for counting deaths in Russia are different, relying on full autopsy reports supporting death from Covid-19, as opposed to a death included if Covid-19 is mentioned on the death certificate.

Finally, of course, there's the question of whether the data from Russia (and Iran, China, North Korea, etc) are even trustworthy. It's been suggested that earlier in the year Russia reported that there were no Coronavirus cases, but a curious spike in pneumonia deaths at the same time. Again, the only way around these anomalies is to hope that people can at least count actual dead bodies correctly.

None of this means that the Case Fatality Rate isn't worth measuring, but it can't be used in the way you used it at first, ie when comparing the UK's and Iceland's response. Which is a shame, because it's probably right to consider Iceland's response as more effective. In that sense I kind of feel like it's a "right for the wrong reasons" post: Iceland's response has been more effective, but it isn't the Case Fatality Rate that tells us so.

[KARL]

Pixie, Apologies, mine of 20.31 reversed the facts - from some weeks earlier and until 22nd April Icelandic nationals were quarantined on arrival from abroad while foreigners came and went unrestricted. From that date (i.e. once the epidemic was firmly under control) every individual arriving in the country has been required to enter quarantine.