Are The Virus Stats Accurate?

[BilSuth]

If Joe Bloggs self-diagnoses (correctly), he does his 14 day isolation without bothering or notifying any medical authority, then returns healthily to normal activity, how can he possibly show up on the official figures?

[ToraToraTora]

they are only accurate for what they claim to be, Known cases, known deaths. Of course there are many unknowns.

[Zacs-Master]

He doesn’t. That’s why the stats usually (and should always) say ‘confirmed’ cases.

[-SharonA-]

Also I think a lots of these deaths are not correct as I heard of a family whose mother was in hospital with heart failure and was due to die. She died last week and the family are angry that the death certificate says covid 19 when right up to her death, she had none of the symptoms.

[ummmm]

I would guess loads more people have than they know.

[Mamyalynne]

Is that all it states Sharon?

Usually the first listed cause is the one that directly led to the death and then working backwards other causes are listed.

[ichkeria]

The death rate for closed cases is currently 13% and rising, although there will be a lot more unconfirmed cases and probably fewer unknown deaths.

[-SharonA-]

Mamya, I believe heart failure was listed too.

[NickorWan]

I struggle to make sense of the stats. Yesterday over 450 people died of the virus here in Spain. I read somewhere that most deaths usually occur around day 17. That means that the 450 figure is only really relevant against the number that got the virus 17 days ago, but they don’t seem to be the stats we see.

[fiction-factory]

I love statistics but I am very wary of the reported cases figures and resultant percentages, and certainly think international comparisons may not tell us much. The deaths figure is more reliable but even then comparisons between countries may not tell us everything due to recording differences- particularly in the case of multiple causes.
But I am happy for the medical professionals and statisticians, people like Chris Whitty, to use whatever statistics they think are most reliable and helpful

[BilSuth]

Points about precise cause of death are relevant but so long as they always use consistent criteria (e.g. the medical equivalent of the legal concept of proximate cause) the stats won't be skewed. My point is about undisclosed cases. These will probably increase with greater public awareness.
This not scare-mongering. Joe Bloggs as described, if indeed not longer infectious, is apparently no longer a threat. As for these omissions from the official stats, there is no problem as it is the trend that really matters. Bear in mind that the numbers stated today are 10 to 14 days out of date and hopefully the graph will flatten out a bit as the public increasingly observe sensible safety measures (again, it will take to the end of the month to kick in and affect the stats).
Non-disclosure cases have the small disadvantage that it rules out contact listing and tracing. I'm not sure how much they still do that here in UK, maybe the numbers are too great now. In any case it has created problems elsewhere with identities being revealed and the infected folk (even when recovered) being victimised and trolled.
Meanwhile be careful, folks, and stay healthy.

[NickorWan]

Following up my earlier post. We know that in Spain over 450 people died yesterday and we are told that most would have contacted the disease around 17 days ago. Say for example we assume that most people know they have the virus after 7 days - we look at the known cases 10 days ago which was just over 3,000. That's 450 out of 3,000.

They are scary figures, someone tell me what I am doing wrong!

[ichkeria]

You are making quite a lot of assumptions there.
It’s not possible to name any accurate assessment of the percentage of deaths because there are a lot more people probably who have the virus than have tested.
However as I posted earlier, the worldwide death rage for composted cases is 13% which is quite high

[jno]

NickorWan, there are various graphics on this page that may help

https://www.ft.com/coronavirus-latest

[ichkeria]

“death rate for closed cases”
(Sigh)

[ichkeria]

...actually 14% now, which is comparable to your Spanish figures

[steg]

//death rage for composted cases//

I’ve a name for my new death metal band...... braw

[jno]

it does sound like the sequel to 28 Days Later.

[NickorWan]

jno, what I am looking for is a figure that shows the death rate against the number who got the disease at the same time as they did. Death rate against the number who have the disease at the time they died isn't the same thing when the numbers are increasing at the rate they are. If you considered death rate yesterday against numbers with it yesterday the figure would look much better than it actually is.

[NickorWan]

Sorry, should say 'if you considered number of deaths yesterday...'

[ichkeria]

As previously suggested, the "rate" is never going to be absolutely accurate. There are different ways of gathering stats, but as long as we don't actually know how many people have the virus, then percentage death rate is impossible to say.

As good a way as any is looking at the closed (or indeed "composted" if you prefer) cases.