Daily numbers, graphs, analysis thread 13

[PatriciaHolm]

Taking the liberty of starting a new thread as we've just bust the old one, with much thanks to @BigChocfrenzy and I will copy her header..

Welcome to thread 13 of the daily updates

Resource links:

Slides & data UK govt pressers
UK dashboard sub-national data, local authorities
Beta Uk dashboard deaths, cases, hospitals, tests, partially sub-national
UK stats updated daily by PHE & DHSC
ONS UK statistics for CV related deaths, released weekly each Tuesday
PHE surveillance report infections & deaths released every Thursday with sep. infographic
NHS England stats including breakdown by Hospital Trust
FT Daily updates
HSJ Healthcare updates
Worldometer UK page
Plot FT graphs compare countries deaths, cases / million pop. / log / linear
Covidly.com filter graphs compare countries
Plot COVID Graphs Our World in Data

We welcome factual, data driven, and civil discussions from all contributors 📈📶👍

@JulyBreeze

So the two daily figures published by the government - positive tests and deaths - are actually both less useful than the unannounced third stat, hospitalisations....

Hospitalisations are available on the website. They aren't announced as a headline, but they were always on the daily briefing slides back when the daily briefings were a thing; I don't think there is anything underhand or any idea of trying to hide them.

I wouldnt say the published figures were less useful, but they dont show the whole picture.

The increasing test figures are showing that outbreaks are not being stamped on fast enough and that there is still a risk of what has happened in Spain happening here from 1 August when shielding protection ends. If a lot of people refuse to wear masks - or wear them but lean over people to put their shopping on a checkout 5 seconds sooner (happened to me twice in my last 2 supermarket visits, first one no mask) - then we could be looking at rising hospital admissions in 7 days and deaths in 2 weeks.

At the moment it's still finely balanced, so I wont personally be going back to the shops and restaurants. Still looking to bring forward purchases and support takeaways, as we've done throughout this. However the more hissy fits I see over wearing masks the less money I'm going to be spending.

I think the cases and deaths figures are not as helpful as they were because (unless you can dig right down into where testing is happening, what age groups, whether asymptomatic/mild etc) it’s hard to tell what the changes in positive cases are telling us and as previously discussed, death stats just look strange at the moment, possibly because of the ‘anyone ever tested as positive, even if long recovered’ being recorded as covid deaths or something else that would explain what appears to be an overly high mortality rate. In this situation, hospitalisations are most useful and they have always been readily available and clear enough.

The NHS data on deaths in hospitals is also good data to look at for trends. But there's more of a lag on it than with hospital admissions, obviously.

The UK PHE death figures aren't worth looking at IMO as they are clearly wrong. They don't fit with the ONS data that's released weekly at all. When are they sure to complete this review is them does anyone know?

@PatriciaHolm I wasn't suggesting it was underhand, more like not properly thought through maybe!

It's the headlines that the media seize on and (mainly consequently) stay in people's thoughts and influence their decisions.

[quote JulyBreeze]@PatriciaHolm I wasn't suggesting it was underhand, more like not properly thought through maybe!

It's the headlines that the media seize on and (mainly consequently) stay in people's thoughts and influence their decisions.[/quote]
Yes, I agree - we have ourselves in a media driven spiral of having to announce this stuff every day now, which I don't think is helping at all!

The NHS triage reports for 111 and 999 is interesting too - it shows an ongoing low level, consistent with the observed continuing low levels of hospitalisations.

@cathyandclare

I downloaded the figures and Bradford caught my eye, every time I spotted a highish figure ( in the 20s or 30s) it was Bradford. Looking at the beta graph, that's not really reflected yet. I haven't charted it out, so it could be my local bias as I live in Yorkshire!

Bradford certainly has had high levels per 100,000 for some time now.

The above post, plus another poster's comment about huge regional variations, got me thinking that it seems to be the same set of LAs with higher case levels; or at least if not all the exact same LAs a similar set of LAs in terms of local characteristics such as higher population densities, higher levels of deprivation; higher levels of poor / over-crowded housing, and higher levels of inter-generational households.

Comparing the cases per 100,000 in w/e 28th June with the most recent complete week, w/e 26th July a significant number show limited movement up or down, Whilst there will probably be 100 to 200 further cases to add to w/e 26th July it will show a rise on the last couple of weeks and come up almost the same in terms of weekly case numbers as the position four weeks ago.

Bradford was high then and is still high now. A simple scatter chart shows this - dots below the diagonal line have a case rate lower than 4 weeks ago, and those above higher. It highlights the bigger movers - so Barnsley, Rotherham, Doncaster have all come down a fair amount as compared with 4 wks ago; conversely the likes of Oldham, Sandwell and Trafford all up significantly.

Blackburn and Leicester still have / had, literally, cases per 100,000 off the chart - and are not shown because the scaling gets a bit silly. At least in Leicester though the rate has fallen a lot from 144.3 to 56.5; to maintain perspective though 56.5 is still around eight times higher than the all England rate per 100,000 for last week.

data set behind the graph

Hospitalisations and deaths remain low in most countries that have come out of lockdown,
because there are far more younger people being infected than before
e.g average age at infection in Germany has dropped to 37 from the peak of age 52 in April

So they aren't really a reliable guide to how many new cases, just to how many become seriously ill

As the age dropped, hospitalisations dropped too, from a peak of 22% to the current approx 10%

and deaths as a % of confirmed cases dropped from a peak of nearly 7% to the current 0.7%,
a factor of 10

PHE must still be updating their deaths - it's taking a long time, but they can't simply use the 28 day cutoff for all cases
because they must include the deaths from COVID that genuinely happen long after the 28 day cutoff
e.g. recent report of the consultant clapped out of ICU after > 60 days - who died suddenly a few days later after a stroke

Sad, but manageable numbers of deaths predicted for the next 3 weeks

https://www.theguardian.com/uk-news/2020/jul/29/uk-health-leaders-call-for-government-to-seek-total-elimination-of-covid-19

Scientists advising the government have predicted that between 43 and 84 people will still be dying from Covid-19 every day by mid-August
....
On Wednesday, the government’s official dashboard showed 83 Covid-19 associated deaths and 763 newly lab-confirmed infections.

Modellers from the MRC Biostatistics Unit at Cambridge University, however, estimate that there are far more new cases – approximately 3,000 a day
– and that deaths will stay high for weeks to come.

For context, around 1,500 people die every day in the U.K.

www.google.co.uk/amp/s/www.bbc.co.uk/news/amp/world-asia-india-53576653

Interesting article about Mumbai slums from which a number of potentially important conclusions can be tentatively drawn:

Assuming a representative sample (which they appear to sought to do), 57% antibody levels indicates the majority of those infected do produce antibodies, and that undermines the theory herd immunity can be obtained with very low levels of antibodies in the population. The vast majority of places are a very long way from achieving it.

The slums seem to be pretty close to achieving herd immunity, and yet this was achieved without a catastrophe. Yes, there were many deaths which is tragic but the rate was 1 in 1,000-2,000. This is consistent with Covid being exponentially linked to age... I believe life expectancy is in the 50s in those slums, a good 20-30 years lower than most developed countries. Given that Covid mortality risk doubles every 8 years, this is consistent with a 0.5-1% death rate in the “West”. Moreover, this indicates Covid is a “wealthy” nation disease, and that developing countries with a low average age would probably be better just to carry on regardless given the overall risk level and the impracticality of social distancing, and that any attempts to impose lockdowns in such places do more harm than good.

Bigchoc say 5-80/3000 ..that still seems v high compared to other countries’ mortality rate?

50-80 not 5-80

Brazil have a young average age and went down the "protect the economy " route - they've had a lot of deaths and overwhelmed health services.

Dont know why anyone would say Bradford doesnt show up on the beta - it does. coronavirus-staging.data.gov.uk/cases?areaType=ltla&areaName=Bradford

Last couple of days need to be ignored but it's quite clear Bradford hasnt got to grips with the virus.

jamanetwork.com/journals/jama/fullarticle/2764787

onlinelibrary.wiley.com/doi/abs/10.1002/jmv.26364

Really interesting that studies in Iran and California found that in around 25% of deaths, patients had a co-infection. Mainly flu as well as Covid.
Does this add weight to the governments flu jab programme? It seems to show co-infection is a big risk factor

@BigChocFrenzy @sleepwhenidie does that 50-80 deaths prediction take into account the current PHE reporting anomaly? Maybe that would explain why it still seems relatively high?

Dont mean to disagree that some poor countries have little choice on what they do because part of their population will die of starvation if they trash the economy. India, South Africa cant stay in lockdown long. But we've seen what happens when countries try to carry on regardless and no-one yet knows the long term health impacts of this disease.

[quote MarcelineMissouri]**@BigChocFrenzy* @sleepwhenidie* does that 50-80 deaths prediction take into account the current PHE reporting anomaly? Maybe that would explain why it still seems relatively high?[/quote]
....
No, because afaik PHE have not yet corrected it

• I presume < cross fingers > they'd announce when they switch to a new system and not just slip it in.

We've done / seen a few estimates of overcounting in the range 7-37 (!) which would still leave England with higher deaths than expected
However, until PHE release their update figures - hopefully they will indeed stop using a different counting system to the rest of the world ! - the official figures are all we have

@alreadytaken

Dont mean to disagree that some poor countries have little choice on what they do because part of their population will die of starvation if they trash the economy. India, South Africa cant stay in lockdown long. But we've seen what happens when countries try to carry on regardless and no-one yet knows the long term health impacts of this disease.

..... There is a balance for each country between risks of COVID

• population age, density, capacity of health service, track & trace, prevlence of T1 & T2 etc -

vs risks of lockdown

• sufficient wealth & infrastructure to support - especially feed ! - workers locked down, strength of economy to withstand some businesses going bust, at least shortterm mass unemployment

imo, in a developing country lockdown is the worse of 2 terrible options,
but some of their government may choose to protect the minority of elderly and middle-aged in their own class

Also, age is so dominant that a country's IFR should - initially - be very low if they have few old people

However, the health service of a developing country is likely to have v few ICU beds and may be overwhlemed with comparatively few serious cases,
after which younger people who cannot get treatment for COVID or any other ailment or accident would die in larger numbers

Someone kindly pointing me in the direction of this data on this thread...

coronavirus-staging.data.gov.uk/testing

Does anyone know if the daily number of tests processed (154496 for yesterday) corresponds to the positive coronavirus tests for that day.

Or if there is a time lag between the two numbers?

Large number of tests yesterday but smaller number of positive tests - hoping a positive sign for August but also wondering whether numbers match or not?

Thanks very much!