Daily numbers, graphs, analysis thread 6

[Barracker]

Welcome to thread 6 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.

That would give a crude death rate (just deaths / number infected) of 1.25% for New York. However deaths will continue to rise and the number infected has probably gone up too.

But 20% of 8m 1.3m
So 21k/1.3m
1.3% death rate. ??

X post

Yes keep we saw a massive drop over the week before lockdown in numbers. Especially after Ireland announced closures. I'm sure NI and Scotland did too at that point?

21k is the number of deaths in the whole state.

NYC is 11.2k.

@BigChocFrenzy
Thanks for posting a link to the paper about years of life lost due to Covid19. It seems that the main message of that paper is that years of life lost is higher than some people are saying because people of that age with the comorbidities appearing on the death certificate often have a life expectancy of several years more. But, unless I'm missing something (quite possible!) it doesn 't seem to reflect the fact that people can suffer from comorbidities to different extents and it seems intuitively likely that someone who is suffering very badly with those is also more likely to die from Covid19 (so eg an 80 year old with lung cancer might live another 2 years on average but an 80 year old with lung cancer who died from Covid19 might have been expected only to live for another 6 months had they not died of that). So I think it's a really important thing for us to understand but I'm not actually sure whether that paper helps with that or not. Do you have any thoughts?

PilbrookesI think that is covered in the paper:

Methods:
We first estimated YLL from COVID-19 using standard WHO life tables, based on published age/sex data from COVID-19 deaths in Italy.

We then used aggregate data on number/type of LTCs to model likely combinations of LTCs among people dying with COVID-19.

From these, we used routine UK healthcare data to estimate life expectancy based on age/sex/different combinations of LTCs.

We then calculated YLL based on age, sex and type of LTCs and multimorbidity count.

20% of 8m is very obviously not 1.3m

implied IFR is around 0.65%, but probably will go up as some of the 21% infected will die, but haven't died yet.

However true infection rate could also be higher than 21%?

This thread is now my major source of coronavirus info and I was just wondering if anyone has any data to help solve a debate at our house.

We live fairly near the main road to a reasonably sized teaching hospital. A few days before lockdown we could hear what felt like constant sirens along the road (at least 5-10 a day). We thought it was just going to get worse but a few days later it got a lot quieter (1-2 a day) until a few days ago when the number of sirens went back up again and we've had the air ambulance over too. We've been debating a few scenarios to explain it.

Scenario 1
Patients were transferred out of London when the outbreak began to pick up to free up space and are being transferred again as our country hasn't had too many cases so has capacity

Scenario 2
At the beginning they were admitting people with milder symptoms so were admitting more people and they are doing the same again whereas in the middle part they were trying to keep people at home therefore less ambulance activity

Scenario 3
There was more traffic on the road pre lockdown and now so there is the same number of ambulances moving around but they now need to use sirens to warn other road users.

Scenario 4
Just coincidence

Any data to shine a light on this would be great.

To answer the question what happened on 18th.

I think we have to consider the week of 16-23 March saw many changes. the government had started adding measures and other countries in U.K. took further measures encourage English to follow suit.
Alongside this around the 12/13-20 there was also an huge increase in numbers of people in supermarkets panic buying.
The gyms bars and restaurants closed on 20th so even though lockdown officially started 23rd that weekend had seen a HUGE drop in social mixing.
I know by that point ds disability swimming comp (National) has already been cancelled and that wasn't until April. He was also doing a regional comp that weekend and it was cancelled. Swim England had been posting advisories and we'd changed swim training that week to avoid squads swapping sessions in changing room and limit contact. Showering before swimming etc.

Was pleased but surprised in the large drop of deaths today when it usually can still be weekend catch up.

Also that FT graph is both scary with regards total deaths being over 50% higher than recorded but I'd like to know how they came to those estimated figures re home and care home deaths. Surely they won't grow at same rate due to populous?

@BigChocFrenzy
Thanks but that supports my point doesn't? The analysis assumes that someone aged 80 with eg lung cancer and hypertension who dies of Covid19 is simply'chosen at random' from all people in that group, while, in reality, they are more likely to be iller than average in that group.

Pilbrookes I read that differently:

"We then used aggregate data on number/type of LTCs to model likely combinations of LTCs among people dying with COVID-19."

We will have to agree to disagree

@BigChocFrenzy
Thanks again. I'm struggling to interpret it the way you are tbh. Maybe the paper could have been a bit clearer on this.

Unfortunately, you would expect IFR in the US and especially in the big cities (like NYC) to be higher than the average. A meaningful percentage of the US population does not have access to healthcare and this will impact the numbers.

P1brookes The authors took a variety of combinations of comorbidities and ages,
using the data on the dead that they had
and the statistics about life expectancy with those kind of combinations

That looks to be as good an estimate as one can reasonably get,
unless you want to assume only the terminally ill died

@BigChocFrenzy
I don't want to assume anything! That's why I was interested in the research. It might be the best that n be done with the existing data but if it doesn't take into account the severity of tbe comorbidities I don't think the results can be particularly meaningful.

PBrookes - my reading is the same as yours. They are assuming that those who die with e.g. 2 co-morbidities aged 80 had the same average life expectancy as all 80 year olds with 2 co-morbidities. That is likely to be an over-estimate, but as they can never know when someone would die, it means that the study is an upper bound.

All the COVID survival rates given by doctors for age, comorbidity etc are averages in each case

There is also no consideration of will to live, family etc

An 80-year old might survive, but a 50-year old might not

That is as good a study as can currently be formed

The study may be an average, not an upper bound

But if it's a systematic overestimate and we have no idea of how much it is likely to overestimate by ...

As Eyewhisker says it's an upper bound. We should be careful of drawing conclusions from this.

There is a continuing narrative that the dead would have died anyway within a few months

That is necessary in order to "other" them and make more people regard the elderly and those with comorbities as expendable

There are 1.5 million in the Shielded category and about another 15 million in the Vulnerable
That is a lot of people to write off

@BigChocFrenzy
OK, you've got an agenda here and I'm j ust interested in the truth so there's no point in discussing this further.

The conclusions I tentatively draw - until I have better evidence - is that most people who died had a probable life expectancy of several years

but those who want to end lockdown and go for herd immunity prefer to continue with the theme that the dead would have died within a few months

I want to end lockdown in stages, on the basis of expert scientific advice,
not to abandon caution because too many people want to get back to full wages and normal life.

Bigchoc - if the death rate is less than 1% and a third of the population are vulnerable, they clearly are not being written off.

Where is your evidence or paper that the life expectancy of the dead was very short ?

Going by feelings, not evidence, is what I consider having an agenda

If you mean the claims of the Social Darwinists like Toby Young, then they all look at average life expctancy for the elderly,
instead of the much longer life expectancy for someone who has already attained a particular old age