Daily numbers, graphs, analysis thread 8

[Barracker]

Welcome to thread 8 of the daily updates.

Resource links:
Worldometer UK page
Financial Times Daily updates and graphs
HSJ Coronavirus updates
Johns Hopkins Coronavirus Resource Centre
NHS England stats, including breakdown by Hospital Trust
Covidly.com to filter graphs using selected data filters
ONS statistics for CV related deaths outside hospitals, released weekly each Tuesday

Thank you to all contributors for their factual, data driven, and civil discussions.

On BBC Horizon last night they were clear that black and Pakistani/Bangladeshi had a very much increased risk,

Risk of what? And these are not the same.

Pakistani/Bangladeshi are ethnically similar to Indian who have a slightly lower death rate, but Pakistani & Bangladeshi men are extremely likely to work as taxi drivers, which is the profession with by far the highest death rate.

The figures are here

www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/coronavirusrelateddeathsbyethnicgroupenglandandwales/2march2020to10april2020#ethnic-breakdown-of-deaths-by-age-and-sex

And if you look at the data by sex then South Asian men have a higher relative risk than women relative to white men, but black women and men have almost the same risk as each other.

This is likely to reflect differing jobs and other social factors rather than anything inherent to the ethnicity.

It might be that black people are inherently more likely to die, but that's not proven, only that they are more likely to have died, and this might be because of geography, socioeconomic, cultural and/or inherent racial differences

In general there are NO jobs that for women have a higher death rate than the average men, but several jobs for men that carry multiples of the average male death risk, so it's not clear why black women are doing relatively almost as badly as black men when that's not true of any other ethnicity. But probably a large part of this is employment. What employment fields are black men overrepresented in?

Daily Mail reporting no new corona cases in London over the last 24 hours

Yes and 363 deaths, encouraging figures.

BigChocFrenzy, do we have any data that breaks down the "non-Covid" excess deaths by age, please?

You would have to define 'excess deaths'

Excess deaths can be covid deaths that have not been tested. Or they can be deaths of people who have died because they havent got timely medical treatment.

We don't know that there are any 'non-covid' excess deaths.

We can count total excess deaths, but in general deaths by all causes are down except for diseases found only in the very old, which are likely undiagnosed covid-19.

So there is no evidence for covid-19 causing excess deaths beyond those who actually died with covid-19, even if that doesn't appear on their death certificate.

The problem with age groups is that young men in particular MOSTLY die from violent causes (car acccidents, murder, suicide), and we won't have any stats on these for months and months.

So it's meaningless really to look at excess deaths in younger age groups because we'll see fewer deaths being registered because inquests from say murders last August are being postponed, so the deaths aren't being registered at the moment.

We can easily count total excess deaths by week of registration (but not last week because of the bank holiday), but you'd need to try to work out the number of deaths that would normally be of unnatural causes

Bigchoc absolutely agree about deaths included being people who are long term icu patients. This is why I'm watching confirmed cases against no's of people tested (not no of tests ) and no of admissions to hospital and % of critical care beds occupied with Covid patients.

Adding in today deaths with the past 4 days and taking a 4 day average is 310 for UK.
For me this is still too high to be relaxing any more measures having seen the numbers flocking to beaches etc in non family groups and our police not having power to enforce it

@GreyGardens88

Misleading, and frankly irresponsibly dangerous, click bait. Many people would have read the headline and thought “yay, it’s all over”, when in the text it says:

“Officials have played down the numbers, suggesting that they may be the result of a technical hitch known to have happened over the weekend, and explaining that they will rise in the coming days as more results come back. The number should not be interpreted to mean the epidemic is tailing off, they said.

The numbers also showed that only 79 cases were diagnosed across the whole of England - this, too, will inevitably rise in the coming days as more people who were swabbed on Monday test positive. A total of 2,412 people yesterday received positive results from samples taken between May 13 and 18.”

@shoots Horizon broke it down as Black, Indian, Pakistani/Bangladesh and so on, and yes Black and Pakistani were higher than Indian. Risk of death. And they also said when all other factors were accounted for. So I’m afraid different conclusions to you, which is why I wondered where they got info from.

Correlating but non causative factors ?

• skin colour / vitamin D
• occupation / sedentary / taxi drivers
• faith / mosques / religious practices / god's will
• family structure / multi generational / overcrowding

I love stats but unravelling that lot will the the work of an expert

and that is before we add into the mix

• immigration status / hostile environment

affecting willingness to contact the authorities

Re London - the numbers I find the most encouraging are hospital deaths. The total announced today is 10 even though it is Wednesday (so there should be no obvious weekend glitches). 2 go back to March. 4 occurred on May 17, 3 - on May 18, 0 - up until 5pm on May 19.

These are very low numbers. I am wondering what is happening to hospital admissions. The number was 49 a few days ago.

Sorry, the total announced for London is 13, not 10. Have made a typo.

Still a very low number.

How funny, March skiing trips suspected in EU:

https://www.theguardian.com/world/2020/may/20/top-eu-doctor-europe-should-brace-itself-for-second-wave-of-coronavirus?CMP=ShareiOSAppp_Other

@ListeningQuietly I wonder whether language barriers should be in that list as well?

Pakistani/Bangladeshi are ethnically similar to Indian

I think I would describe Pakistani/Bangladeshi as ethnically very different from Indian: different religions - Muslim/Hindu - higher educational attainment among British Indians - consanguineous marriages (and more health problems/birth defects) more likely among Pakistani/Bangladeshi families.

Religious worship must have played a part in the early spread of the disease - both Muslim and Christian festivals/worship have been linked to large clusters in other countries;

BBC website yesterday - "France had a cluster of Covid-19 cases in February that originated at an evangelical church.Thousands of people gathered in Mulhouse for a week of activities. More than 2,500 cases are said to have been linked to it worldwide, Reuters news agency reports."

"UK March - Despite the government advising Londoners to avoid unnecessary social contact, an East London Mosque and London Muslim Centre spokesperson confirmed today that it is open to the public."

"On a normal day, some 10,000 worshipers from all over East London attend the Whitechapel site across five daily prayers."

www.mylondon.news/news/health/london-coronavirus-citys-busiest-mosque-17930813

it's not clear why black women are doing relatively almost as badly as black men.

Link from your link - 'Understanding and reducing ethnic
inequalities in health 2018' -

"Among women, those identifying as Black African had the highest obesity prevalence when waist circumference was the measure (at 53%), and those identifying as Bangladeshi when the waist-to-hip ratio was used (at 50%). Regardless of the measure used, women and men identifying as Chinese had the lowest obesity prevalence."

Unfortunately many of the younger black nurses/healthcare workers who died were noticeably overweight as were the younger white nurses and healthcare workers. I wonder if black women are more likely to be regular church goers as well/sing in a church choir - coronavirus clusters have been linked to choirs.

The link also says; "An investigation in 2011 found that over 90% of GP practices and 100% of hospitals had systems in place for recording ethnicity but data completeness and quality remained poor with useable ethnicity data available for 86% of new inpatients, 50% of new outpatients, 53% of new A&E patients and just over 3 quarters of new primary care registrations in 2011.Ethnic group is not recorded on death certificates in England, so most analysis of mortality rates uses country of birth, an increasingly inadequate proxy for ethnic group."

assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/730917/local_action_on_health_inequalities.pdf

taxi drivers, which is the profession with by far the highest death rate.

Taxi drivers are often older - "the average age of a driver was 47 years old, with 29% of drivers being aged under 40.Those aged 60 or over made up 19% of drivers.There has been a slight shift in the age profile of drivers over the past ten years, with a slightly smaller proportion of younger driver and a slightly larger proportion of older drivers.The two main ethnic groups of drivers were White and Asian or Asian British in 2018/19, making up 42% and 40% of drivers respectively."

London cabbies tend to meet in those tiny little huts during breaks - very squashed together I would imagine - perhaps Asian cab drivers similarly meet in close contact, or perhaps they meet in the local mosque during the day.

Strawberry
The Language barrier is linked to the religion and the hostile environment.
But that mostly affects women .....

ListeningQuietly I wonder whether language barriers should be in that list as well?

A head teacher from London on world at 1 earlier this week said that she felt that language was a significant barrier in understanding around SD among the mostly BAME cohort of pupil parents as demonstrated by how they acted when they met her and various other things she'd dealt with over the last few months..

In the guardian link below (and I'm linking here again) Dr Andrea Ammon, director of the European Centre for Disease Prevention and Control (ECDC), states that shielding the elderly only is not appropriate given the numbers of young also affected.

https://www.theguardian.com/world/2020/may/20/top-eu-doctor-europe-should-brace-itself-for-second-wave-of-coronavirus?CMP=ShareiOSAppp_Other

Shoots I haven't read the pillars document. I get all the documents from gov.uk (and read any stats ones or school, or disability ones).
Do you have a link?

Can I also add that I know the testing increases are track and trace etc and clusters.

"coronavirus clusters have been linked to choirs."

Yes, it is a big problem, especially with large enthusiastic congregations crammed together indoors, singing & praying lustily for 2 hours

My late mum's CofE church on a typical Sunday had about 10 people and a lost dog in their huge church building,
but some sects or evangelical churches can have a cast of thousands

e.g. an astonishing amount of SKorea's early infections came from one odd sect:

www.bbc.com/news/world-asia-51701039

(2 March) About 60% of the country's more than 4,000 confirmed cases are sect members.

It is good to see some dynamic evaluation of the rapidly evolving Covid picture in this thread rather than dogmatically sticking to one theory and gathering statistics to prove that idea.
Historically we have seen many scientists take the medical profession down the wrong path, for example saturated fats and heart disease. It is now well documented that refined sugars are the problem. Many medics have quietly come to this conclusion well before the ‘scientists’ accepted they’d got it wrong.
Good science is happy to accept it may be wrong, as opposed to proving they are right by any means.

@NeurotrashWarrior yes that was exactly the report that I was thinking of when I commented!

It is good to see that theories are either backed up by data, or rejected by data

There is likely to be an increasing number of insolvencies among care homes due to:

the number of residents who have died
falling numbers of new residents
large costs of Coronavirus - e.g. PPE, care in isolation

Large numbers of feaths have economic consequences, even deaths of those no longer economically active

Some care homes in the NE have already called a Major Incident re their financial crisis

www.bbc.com/news/world-asia-51701039

Adam Kucharski@AdamJKucharski (Mathematician/epidemiologist at LSHTM WellcomeTrust)

This modelling study from Sweden suggests 26% of Stockholm county will have been infected by May 1st
(folkhalsomyndigheten.se/contentassets/…).

However, our estimates of under-reporting suggest only 5-10% have been infected so far
(cmmid.github.io/topics/covid19……^^).^
So what's going on? 1/

The model uses a study that found 2.5% of a random sample tested positive in Stockholm County early April.
As only 150–200 cases were being reported each day during that period, it suggests there was a lot more infection out there.
But how much exactly? 2/

The key issue here is that the number currently infected (i.e. prevalence) isn't the same as the number of new cases (incidence).

To account for this, the model fits to both observed cases and the number infected (i.e. the 2.5% prevalence estimate).
And here's the issue... 3/

To work out under-reporting we need to compare that 2.5% to how many active infections were being reported at same point.

The model assumes people are infectious for 5 days,
i.e. it effectively compares 5d worth of reported cases (i.e. ~800) with the 2.5% prevalence estimate.^ 4/^

This suggests about 70-80 infections for each reported case, which leads to 26% estimate.^
^
But are people really infected for only 5d?
People may be infectious for relatively short period, but they can test +ve for longer: possibly 2 weeks on average
medrxiv.org/content/10.110……^^ 5/^^
^
This suggests we shouldn't be comparing 2.5% with cases over 5 days - we should be looking at the previous 2 weeks.
Looks like ~2400 cases were reported during this period, suggesting around 25 infections per reported case.
In other words, 1/3 of the value they estimate... 6/

This would imply 5–10% will have been infected by 1 May, consistent with our estimate (and Imperial:

mrc-ide.github.io/covid19estimat…).

The above is of course just a rough estimate, but it shows that it's always worth checking results against other data sources. 7/7

Models are varying so widely, depending on key assumptions and methodology

Now we have reliable antibody tests, we desperately need national sampling in several countries and also major cities such as London