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.

I wonder if this learned and balanced groups might help - it's with regards to the hot topic of children returning to nurseries where social distancing is impossible (specifically nurseries and not schools). Is the any data from other countries about what happens when children return? I'm concerned England is going too soon but can really only see biased arguments from the government and unions.

summary of the research on children www.rcpch.ac.uk/resources/covid-19-research-evidence-summaries

As for blood clots - medics have known about this for some time and if in hospital you will possibly be given anticoagulants. It wasnt well known at the start. It is one of the reasons the outcome for those admitted more recently is likely to be better than those first admissions. Treatment is already improving a bit even if nothing revolutionary has been found.

Is there any explanation as to why Brazil is so badly affected, but other South American countries have got off pretty much scot-free? Argentina for example have less than 400 deaths, Paraguay and Uraguay have none.

I worry that the chart BigChoc posted showing worldwide numbers decreasing is skewed by lack of reporting from some countries.

I thought the explanation for lower death rates in poorer countries might be linked to lower life expectancy (fewer old people!) but the median age in Brazil is low, and almost the same as the rest of South America.

Because the Brazilian pm is a bad as trump

Isn’t Bolsonaro not only worse than Trump but also lacking the system of individual State Governors who have the authority (and hopefully sense) to impose lockdowns?

"Since Brazil confirmed its first coronavirus case on 26 February, Bolsonaro has continually minimised the pandemic, rejecting media “hysteria” over its dangers and suggesting Brazilians could swim in excrement and emerge unscathed.

The Trump-admiring populist has also purposefully undermined social distancing guidelines, mingling with supporters and sacking his health minister on 16 April after he publicly challenged the president’s behaviour."

God help Brazil, because their president wont.

Actually Brazil's best hope is that the president contracts the virus, which is quite likely.

Oh, I get he's an arse, but surely it would be impossible to keep Argentinian figures as low as they are, even with their strict lockdown.
No deaths at all in neighbouring countries to Brazil also seems unlikely.

Russia's death to case ratio also seems indefeasibly low. I suppose we will never really know what is going on in many countries.

Is there any data regarding outcomes for Covid admissions who are already taking anticoagulants?
Also interesting study ( not peer rev yet but with so much new Covid related study going on it may be some time before review is available) but I have been wondering whether a decreasing vulnerable pool is effecting the infection rate. www.medrxiv.org/content/10.1101/2020.04.27.20081893v2

*Russia's death figures are v unlikely, especially considering their exceptionally low life expectancy for a Eiropean country.

Significantly, doctors who post about COVID patients there keep falling out of windows in statistically unlikely numbers

Some S American countries may not be counting COVID deaths properly, or not have the systems in place to be able to do so
However, Argentina's median age of only 32 would lead us to expect a much lower death rate than most of Europe

Bolsanaro's wilful irresponsibility - indeed like a Brazilian Trump - seems to have massively raised their death rate compared to previous years,
e.g. looking at the FT charts of total death from all causes, the city of Manaus seems particularly badly hit - quite likely to be others as well

The low numbers of deaths in Australia and NY, with higher median ages than S Amerian countries (but still higher than Europe)
show what is possible in a country with low population density that locks down early and hard

Oops, Australia and NZ
(Warning to others: I should not have downloaded the last IOS update last night - my keyboard is giving the wrong characters and the usual fixes aren't working)

This thread has been really interesting reading. Has anyone come across any data for survival rates by age and/or ethnicity for those who have been infected?

oralengineer interesting paper. It fits in with the discussion we were having about the low R in London, some degree of herd immunity and the infection getting the 'low hanging fruit' .

thattone ONS chart for deaths by age & sex.

Also, the study of all UK patients - which also includes stats on comorbidities:

Features of 16,749 hospitalised UK patients with COVID-19 using the ISARIC WHO Clinical Characterisation Protocol

https://www.medrxiv.org/content/10.1101/2020.04.23.20076042v1

The median age was 72 years
....
47% had no documented reported comorbidity.

Overall, 49% of patients were discharged alive,
33% have died and
17% continued to receive care at date of reporting.

There was a murder discussed locally two days ago on fb that's not been reported on the news; so definitely still happing.

"decreasing vulnerable pool" would probably play some part

However, NYC with a worse death rate had 21% with antibodies, so London is likely less
Spain found about 5% on average nationally with antibodies
France estimated around 5% too

Looks mainly that lockdown got exponential growth under control, for countries / regions that were experiencing it
and continuing lockdown or - in other European countries - social distancing after relaxing lockdown, has got R0 under 1.0

Murders and suicides will obviously be continuing,
but both are orders of magnitude lower than the spike over the last couple of months in total deaths from all causes

Thank you BigChocFrenzy, that's useful to know.

@cloudsinspring see www.medrxiv.org/content/10.1101/2020.04.18.20070912v2.full.pdf

at page 14, table 2.

They estimate

0.004%-0.15% with a mean of 0.04% for age 0-20
0.0005%-0.08% with a mean of 0.015% for age 21-40
0.002%-0.17% with a mean of 0.04% for age 41-50
0.01%-0.29% with a mean of 0.10% for age 51-60
0.55% - 1.48% with a mean of 0.93% for age 61-70
3.35%-6.34% with a mean of 4.54% for age 70-80
6.39-11.57% with a mean of 8.50% for age 80+

If we look at the UK stats the death total doubles every 6 or 7 years which is likely a reasonable estimate of added risk. The estimates above have a very high degree of uncertainty for the younger age groups - the numbers overlap. It's almost certainly untrue that 21-40 is less at risk than 0-20, but because the number of deaths in their sample was so small, that result occurred by chance. The much larger number of deaths in the older age group makes the results much more certain.

I.e. it's vanishingly unlikely that the death risk is as high as 1 in 600 (the upper bound of the confidence interval) for the 0-20 age group, since the upper bound for 21-40 is 1 in 1250, and we see smaller numbers of deaths in each successive age bracket in the UK, which does suggest a strict exponential age/risk progression.

I am not sure the exact underlying risk of illness causing death by age, but we know that mortality is lowest among pre-teen children, but this is more because they don't commit suicide, etc. as much as older groups - the illness impact is not that significant. There is some evidence children do have an advantage with infection with covid-19 even if in general you aren't necessarily more likely to die of something at 25 than 5.

Also men are probably 50% higher than the numbers above, women 50% lower. And with in the age group, the oldest will be around 3x at greater risk than the youngest.

The government has a list of 'shielded' people

www.gov.uk/government/publications/guidance-on-shielding-and-protecting-extremely-vulnerable-persons-from-covid-19/guidance-on-shielding-and-protecting-extremely-vulnerable-persons-from-covid-19#who-is-clinically-extremely-vulnerable

There is quite a short list.

For example, of the 385 people dying aged

shoots thanks for these posts & the papers referenced

111 today even though it’s the weekend.

Presumably the ‘vulnerable pool’ would also be enlarged if there is indeed a section of the population with natural immunity - which has been discussed here before. Does anyone know of the existence of natural immunity is being investigated? I guess by its nature it would require initial self reporting by those who have definitely been exposed but not become ill...

Sorry, the population outside of the vulnerable pool would be enlarged, not the pool itself!

www.bbc.co.uk/news/health-52662065?at_custom1=%5Bpost+type%5D&at_custom2=facebook_page&at_custom3=BBC+News&at_medium=custom7&at_campaign=64&at_custom4=9F1CF172-9761-11EA-92F0-4FC996E8478F

This link was on the bbc about blood clots.