Over The Peak?

[Dagman]

Hi folks, the government inform us the UK is over the peak of the virus, I don't see it looking at the figures. How are they coming to this conclusion? I can see some sad losses down some days but shoot back up again the next, undoing them! TIA.

[woofgang]

because less people are testing positive, ending up in hospital and dying over a 7 day rolling average.

[tiggerblue10]

Today's figure looks bad but that's because of the Bank Holiday weekend lag. Not all deaths were recorded at the weekend.

[Dagman]

Ah, rolling averages and delayed figures. Didn't know numbers were behind and added on later. Makes more sense in that case. Thanks both!

[Buenchico]

From Aunty Beeb's wesite:
https://ichef.bbci.co.uk/news/660/cpsprodpb/5018/production/_112240502_optimised-european_countries_sm_deaths_standard_12may-nc.png

[tiggerblue10]

No worries.

I really hope that we see a much lower figure tomorrow like 200-300. Sadly still too high but a downward trajectory is more positive.

[Dagman]

Thanks Buenchico, I can see the UK graph showing regular peaks and troughs. I guess I was hoping for a steeper quicker fall in that average line but it is at least falling. Hope the next couple of weeks it doesn't trend upwards again with the new lockdown measures. Cheers.

[fiction-factory]

The deaths figures reflect mainly recent deaths but can include some from a week or so ago. It typically takes 10-20 weeks for someone to die once diagnosed so to get an indication of the projected future falls in death rates in the newt few weeks you can look at the figures for hospital admissions and numbers in ICUs now- and these figures have been falling noticeably.
The recent relaxation may lead to an increase in infections and admissions in the next few weeks but this won't impact on the deaths until June in most cases

[fiction-factory]

10-20 weeks? I meant 10-20 days!

[Milo2]

Our local paper reports daily on the number of deaths in local hospitals. Some reported today were from the middle of April but have only been counted in today’s numbers.

[lankeela]

Why are they no longer giving the number of daily deaths in hospitals - surely this must have gone down a lot if the majority are now in care homes?

[Peter Pedant]

they do it with death certificates
and that is a science by itself
and think of cheques clearing - the same effect the DCs from one day will be spread over and processed over a few days

I take them very generally as a neighbour died in april and we were allowed the death cert in November

but they are on the way down

[Peter Pedant]

// Why are they no longer giving the number of daily deaths in hospitals//
it is a partial total - and the general public were confused
(think of the usual whining innumerate Abers but 60 m of them)

the system was never designed to give daily totals
but the public have been surprised
however I have done it a few times
no my name is not Jesus Christ, I just know what to do with Death Certs

causes of deaths from DCs are so inaccurate as to be a waste of time

[jim360]

Another, potentially more solid sign that we are over the peak is in the ONS weekly death figures. Painfully high across April but in the most recent release, this morning, "only" 8,000 excess deaths (+80%) were recorded in the week ended May 1st as opposed to 12,000 and 11,500 (+110%) in the two weeks previously (England and Wales figures only).

[Peter Pedant]

hi JIm
America looks as tho it is going to do the crazy test - let it rip and see what happens ....

anyway you must have an idea of the halving time from your own graph. any idea what it is
AND - - - the area under the curve - any idea what that will be? - begtween 50 and 100 ?

[jim360]

My own most recent focus is tracking the NHS England hospital data, and I've sort of put analysis of that on hold for a while as I was considering a more ambitious way of tracking and modelling it. Given the recent changes in policy it will be interesting to see if the current track is stable or starts to change in a couple of weeks.

[jourdain2]

Soo many different figures going around. I'm beginning to wonder if there has been a slight over-reaction. I read that a team in Oxford say that the virus is not an epidemic in Britain, with a tiny %age infected. Royal Collage of G.Ps. has figures suggesting fewer than 0.24% of adults have the virus. An epidemic needs 40+ per 10,000 and current figures suggest it is between 3 and 24 per 10,000. (I'm just giving you what I think seem to be reliable published figures.)

Radio 4 had a discussion (may have been on 'More or Less', but I'm not sure, which I find interesting, although not a mathematician) which basically seemed to agree that everything was based on guesswork and that some modelling didn't stand up to examination.

Are we witnessing a media-enhanced epidemic possibly? Just a question and I know that people are dying and it is horrible - but one doctor was estimating that 20,000 extra cancer patients will die this year (who would have survived) because their treatment has been postponed or not begun. Same with other life-threatening diseases.

It's just food for thought.

[jim360]

We've seen already something like 50,000 excess deaths in the UK in the last few weeks. That's a lot. Maybe it's not world-ending, but I would say that world-ending is too high a threshold for a crisis.

Still, concerns about excess deaths being driven in other areas are real and reasonable. But the Government could only have reacted to what was the best information at the time, which was that there could well have been excess deaths in the coming months measured in the hundreds of thousands without some level of intervention. Anybody would panic if presented with solid evidence for that prediction.

[jourdain2]

I tend to agree with you about the reason for a government to panic - but people around here seem to be frightened for no particular reason. We are in a very low infection area. We are taking reasonable precautions - have hand-gel and space out when meeting other dog-walkers, but that's about it and t.b.h. we may as well carry on as usual.

That is, of course, the problem with a broad sweep of figures. Perhaps a regional approach might be better and a more precise definition of figures given. I hear that the vast majority of infections and deaths reported are in hospitals and care homes (in fact a lady I knew was sent from her care home for investigations into her heart and was returned with Covid-19 - she died 10 days ago and others in the home are following her).

Lack of clarity and accuracy is rather a problem.

[jim360]

As to the regional approach -- indeed, and I believe this is part of the Government's plan. Although it will be tricky to be perfectly clear about how this will work in practice.

[woofgang]

jourdain you can see deaths by county but i can't find the link. Its been on here before, perhaps someone else can help?