Coronavirus - General Chat - No statistics

[servodude]

I think the data is interesting both for the recent indication that harsher lock-down measures in the North-East look to have been delivering some good results, and also that some areas without harsher restrictions at the moment are trying their hardest to play catch-up with some of the more infectious regions.

I've noticed the same looking at the changes in rate of hospital admissions growth (based on exp trend over 30 days ending at 7 day intervals back from latest reported figures at https://api.coronavirus.data.gov.uk/v1/data)
- if the fairies are willing there will be a table below




- sd

[Nimrod103]

#1 comparing age of Covid-19 death with average life expectancy.
Fairly obvious, if somebody has reached the age of, say, 80, their life expectancy is calculated on how long the average 80-year-old lives, not on the whole population (which would include all the people who don't make it to 80).

It does not seem obvious to me - am I missing something?
An 80 year old on average may have 11 years of life expectancy.
But the 80 year old who is vulnerable to dying of Covid (because of a variety of different health factors) may only have a life expectancy of 2 years. Whereas the 80 year old who will survive Covid may have a life expectancy of 15 years.
The fatality rate of 80 year olds is, what, 10%. 90% survive. Those who will die and those who get better are two different populations.

[Sorcery]

I think the data is interesting both for the recent indication that harsher lock-down measures in the North-East look to have been delivering some good results, and also that some areas without harsher restrictions at the moment are trying their hardest to play catch-up with some of the more infectious regions.

I've noticed the same looking at the changes in rate of hospital admissions growth (based on exp trend over 30 days ending at 7 day intervals back from latest reported figures at https://api.coronavirus.data.gov.uk/v1/data)
- if the fairies are willing there will be a table below




- sd

Where is Wales in your list? Or even New South Wales? Or Northern Ireland? I saw what you presented as evidence of a decline of serious cases generally, though for good old Nicola, results might tell.
Edited : I would delete the above if I could, fundamental misreading of doubling time. Begs the question for Nicola is why is she is so keen on action when the results are in the bag?

[zico]

"No more lockdown" is not a new idea. It is what we have lived with for decades. Lockdown is the new stupid idea and I want it to end as soon as possible. I greatly resent not being able to see my close family members for the best part of a year for no good reason.

GS

So why don't you simply go and see your close family members if it's important to you? OK, it's illegal, but then so is driving at 31mph in a 30mph zone, and who hasn't ever done that?
I wouldn't condemn anyone for driving at 35mph in a 30mph zone, if traffic was light and they were alert.
At speeds of 36-45mph, I'd think they couldn't complain if they were caught and fined, because they knew they were breaking the law.
At speeds of 70mph+, I'd think they were insanely reckless people who ought to be banned from driving for a long time.

Similarly with Covid. Breaking the "rule of 6" by having 7 people together, but being careful not to keep social distancing - illegal, but not harmful. (Actually, possible less likely to transmit the virus than 6 people together thinking "we're legal, so whatever we do is OK".)
Joining in with hundreds of others to demonstrate up and down Oxford Street - illegal, and pretty harmful.
Under the new rules, it's legal to attend a business lunch in a restaurant with 29 other people. but it's far more likely to transmit the virus than some prohibited gatherings. (A bit like driving at 29mph on a very busy high street being safer than 35mph in a deserted 30mph zone)

Everyone will have different tolerances to risk/benefit, and sensible people can make sensible choices. However, a minority of people don't make sensible choices, or don't value the potential negative impact on others, which is why many pubs (for example) need to be closed, to prevent people from being reckless.

What bugs me is that when a lot of media commentators want "an end to lockdown", by and large they don't mean the freedom to see close relatives a bit more often, they mean the freedom to have lots of low-wage workers risking their health commuting into central London by public transport in order to give the city "a buzz" and to serve food & drinks to the media folks. (I mention London specifically because it's quite striking to see the difference in attitudes between London and regional media folks.)

[servodude]

I think the data is interesting both for the recent indication that harsher lock-down measures in the North-East look to have been delivering some good results, and also that some areas without harsher restrictions at the moment are trying their hardest to play catch-up with some of the more infectious regions.

I've noticed the same looking at the changes in rate of hospital admissions growth (based on exp trend over 30 days ending at 7 day intervals back from latest reported figures at https://api.coronavirus.data.gov.uk/v1/data)
- if the fairies are willing there will be a table below




- sd

Where is Wales in your list? Or even New South Wales? Or Northern Ireland? I saw what you presented as evidence of a decline of serious cases generally, though for good old Nicola, results might tell.
Edited : I would delete the above if I could, fundamental misreading of doubling time. Begs the question for Nicola is why is she is so keen on action when the results are in the bag?

Unfortunately the welsh and NI data are not gathered the same way
- the welsh gather it all confused like
- and the NI data seems to get reviewed continuously so the last couple of values are always well wrong (and typically 7 strangely)
- have a look at the disclaimers in the coronavirus.data.gov.uk site

can you explain what you think doubling time is?
I've posted pictures in the past that might help you see what I mean (perhaps more recently on the science board - including wales and NI for you )
- this is just a least squares fit of the fn a^e(bx) where x = days
- so the time that takes to double is ln(2)/b

New South wales haven't had many cases recently
- you'd have to go back to the start of august till even Victoria's cases matched England's (but I suppose that's not adjusting for it having ~10% the population)

- sd

EDIT: brackets ! always bloody brackets

[servodude]

#1 comparing age of Covid-19 death with average life expectancy.
Fairly obvious, if somebody has reached the age of, say, 80, their life expectancy is calculated on how long the average 80-year-old lives, not on the whole population (which would include all the people who don't make it to 80).

It does not seem obvious to me - am I missing something?
An 80 year old on average may have 11 years of life expectancy.
But the 80 year old who is vulnerable to dying of Covid (because of a variety of different health factors) may only have a life expectancy of 2 years. Whereas the 80 year old who will survive Covid may have a life expectancy of 15 years.
The fatality rate of 80 year olds is, what, 10%. 90% survive. Those who will die and those who get better are two different populations.

The change in fatality rate with age is nicely covered in this meta analysis: https://www.medrxiv.org/content/10.1101 ... 895v6.full

- sd

[servodude]

Its more of a question as to whether this second increase in hospital admissions has a seasonal element or not.

Hence if we look at countries that have contrary seasons then we can look to find how many of them have a second increase in cases (aka wave) and whether is is arguable therefore that there is not a seasonal element to the increase in cases/admissions in the northern hemisphere.

Obviously the disease cannot start infecting people in a country until it gets there

Once we conclude that it does have a seasonal element there is then the question as to what proportion of the increase in hospital admissions is driven by that and/or the lack of immunity from prior infection (either by this virus or a similar one)..

I think I see what you're saying
- and I totally agree that this will have a seasonal component if it becomes endemic
- not sure though how much of a component this will have (or how visible it will be) as it starts out on its inaugural world tour
And I say that notwithstanding that there are clear effects already shown on the virus and its transmission mechanisms and how they vary with environment (stay outdoors people!)

Thought I'd compare Chile with Australia from that data set you had (from worldometers) to see how closely the waves coincided because they have the same seasons

- to be fair though most of that second wave in Aus is just Melbourne (https://covidlive.com.au/) but I'm confident that the shape of it is down to more than the changing seasons


- sd

[johnhemming]

Thought I'd compare Chile with Australia from that data set you had (from worldometers) to see how closely the waves coincided because they have the same seasons

There is, however, a good argument that New Zealand and Australia actually managed to control the first wave (unlike large numbers of other countries) and hence they have a true second infection wave as pockets of infection kick off. They will also then be likely to get a seasonal wave in about 6 months time.

[Nimrod103]

#1 comparing age of Covid-19 death with average life expectancy.
Fairly obvious, if somebody has reached the age of, say, 80, their life expectancy is calculated on how long the average 80-year-old lives, not on the whole population (which would include all the people who don't make it to 80).

It does not seem obvious to me - am I missing something?
An 80 year old on average may have 11 years of life expectancy.
But the 80 year old who is vulnerable to dying of Covid (because of a variety of different health factors) may only have a life expectancy of 2 years. Whereas the 80 year old who will survive Covid may have a life expectancy of 15 years.
The fatality rate of 80 year olds is, what, 10%. 90% survive. Those who will die and those who get better are two different populations.

The change in fatality rate with age is nicely covered in this meta analysis: https://www.medrxiv.org/content/10.1101 ... 895v6.full

- sd

Doesn’t answer my point that those 80 year olds predestined to die of Covid if they catch it (vulnerable to rather than predestined, if you prefer), are very unlikely to have an 11 year life expectancy. So to assume 11 years of life has been lost, cannot be correct.

[XFool]

Proportion of people in England with Covid antibodies has fallen, study says

The Guardian

Figure has dropped by over a quarter in three months, fuelling concerns over reinfection

[swill453]

It does not seem obvious to me - am I missing something?
An 80 year old on average may have 11 years of life expectancy.
But the 80 year old who is vulnerable to dying of Covid (because of a variety of different health factors) may only have a life expectancy of 2 years. Whereas the 80 year old who will survive Covid may have a life expectancy of 15 years.
The fatality rate of 80 year olds is, what, 10%. 90% survive. Those who will die and those who get better are two different populations.

The change in fatality rate with age is nicely covered in this meta analysis: https://www.medrxiv.org/content/10.1101 ... 895v6.full

Doesn’t answer my point that those 80 year olds predestined to die of Covid if they catch it (vulnerable to rather than predestined, if you prefer), are very unlikely to have an 11 year life expectancy. So to assume 11 years of life has been lost, cannot be correct.

This Telegraph article is behind a paywall, but the headline is:

Men who die of Covid-19 losing average of 13 years, while women have 11 years cut off life expectancy

https://www.telegraph.co.uk/news/2020/0 ... tists-say/

Scott.

[dealtn]

https://wellcomeopenresearch.org/articl ... NXlbTrBwoo

Thank you, that's very interesting.

Unfortunately this is from early April, and is reliant on only one sample from Italy, with some modelling from Scotland, and use of co-morbidity assumptions from influenza.

Now don't get me wrong I think it is a good source, and the conclusions intuitively look a long way from the (lazy) assumption "it's only old and sick people that are going to be dying soon ..." you often see. Far enough to make those assumptions look very weak.

The strengths and limitations section by the authors admit though that the study was "conducted rapidly under the pressure of time", better data will be available in time, etc. but also that they were unable to adjust the model for socio-economic, smoking, occupation, and others. Also the model "didn't converge", "had wide posteriors", and hence was by their own admission "indicating substantial uncertainty". A small sample size, and limited data both on co-morbidity, and how those co-morbidities (when there are more than one) act together, will do that. So its not a criticism of the paper, and they admit as much.

You would hope follow up investigations and statistical analysis would produce additional, and hopefully better results. Much like a lot of thinking around the pandemic in April a lot will have changed I suspect. It may well now be that with fewer dying, that have been affected, the YLL will have fallen too, and it is closer to true that it is more of the deaths are the (really) old and sick. Regardless I can't see how if this statistical study, flawed though it is by the author's admissions, is even close true, that the argument that YLL is as low as "weeks or months" can be correct.

Thank you again.

[dealtn]

https://www.medrxiv.org/content/10.1101 ... 20050559v2

"A substantial limitation of our work is that we have not considered factors apart from age that affect the IFR of COVID-19. For example, a recent U.K. study found that mortality outcomes are strongly linked to specific comorbidities such as diabetes and obesity but did not resolve the question of whether those links reflect differences in prevalence or causal effects on IFR."

Not sure I like this line from the other link you provided in demonstrating the robustness of the research with respect to YLL conclusions.

[servodude]

https://www.medrxiv.org/content/10.1101 ... 20050559v2

"A substantial limitation of our work is that we have not considered factors apart from age that affect the IFR of COVID-19. For example, a recent U.K. study found that mortality outcomes are strongly linked to specific comorbidities such as diabetes and obesity but did not resolve the question of whether those links reflect differences in prevalence or causal effects on IFR."

Not sure I like this line from the other link you provided in demonstrating the robustness of the research with respect to YLL conclusions.

I posted a link to an IFR vs age meta analysis a few posts ago (I think in reply to Nimrod)
- you might find it an interesting read
- as with most of this stuff it's mostly "we don't really know" bit it did a good job of collating quite a few sources

-sd

[vrdiver]

It does not seem obvious to me - am I missing something?
An 80 year old on average may have 11 years of life expectancy.
But the 80 year old who is vulnerable to dying of Covid (because of a variety of different health factors) may only have a life expectancy of 2 years. Whereas the 80 year old who will survive Covid may have a life expectancy of 15 years.
The fatality rate of 80 year olds is, what, 10%. 90% survive. Those who will die and those who get better are two different populations.

The change in fatality rate with age is nicely covered in this meta analysis: https://www.medrxiv.org/content/10.1101 ... 895v6.full

- sd

Doesn’t answer my point that those 80 year olds predestined to die of Covid if they catch it (vulnerable to rather than predestined, if you prefer), are very unlikely to have an 11 year life expectancy. So to assume 11 years of life has been lost, cannot be correct.

I'll repeat the link I provided, and for your convenience highlight from the abstract that answers your point:
https://wellcomeopenresearch.org/articl ... NXlbTrBwoo

Abstract
Background: The COVID-19 pandemic is responsible for increasing deaths globally. Most estimates have focused on numbers of deaths, with little direct quantification of years of life lost (YLL) through COVID-19. As most people dying with COVID-19 are older with underlying long-term conditions (LTCs), some have speculated that YLL are low. We aim to estimate YLL attributable to COVID-19, before and after adjustment for number/type of LTCs.
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.
Results: Using the standard WHO life tables, YLL per COVID-19 death was 14 for men and 12 for women. After adjustment for number and type of LTCs, the mean YLL was slightly lower, but remained high (13 and 11 years for men and women, respectively). The number and type of LTCs led to wide variability in the estimated YLL at a given age (e.g. at ≥80 years, YLL was >10 years for people with 0 LTCs, and <3 years for people with ≥6).
Conclusions: Deaths from COVID-19 represent a substantial burden in terms of per-person YLL, more than a decade, even after adjusting for the typical number and type of LTCs found in people dying of COVID-19. The extent of multimorbidity heavily influences the estimated YLL at a given age. More comprehensive and standardised collection of data on LTCs is needed to better understand and quantify the global burden of COVID-19 and to guide policy-making and interventions.

So yes, in people with six or more co-morbidities, the YLL (years of life lost) is estimated at <3, but as the study points out: After adjustment for number and type of LTCs, the mean YLL was slightly lower, but remained high (13 and 11 years for men and women, respectively)

To the point about the 80 year olds and 11 years of life expectancy. When I addressed dealtn's query re sources, I didn't say that the average life expectancy of an 80 year dying of CV19 is 11 years. I said #2 the average lost life span is around 11 years. The reference to an 80 year old did not claim that they had 11 years of life left, but if somebody has reached the age of, say, 80, their life expectancy is calculated on how long the average 80-year-old lives, not on the whole population.

The two points are separate and shouldn't be conflated.

VRD

[servodude]

Thought I'd compare Chile with Australia from that data set you had (from worldometers) to see how closely the waves coincided because they have the same seasons

There is, however, a good argument that New Zealand and Australia actually managed to control the first wave (unlike large numbers of other countries) and hence they have a true second infection wave as pockets of infection kick off. They will also then be likely to get a seasonal wave in about 6 months time.

That sounds dangerously close to saying that lockdowns plus quarantine are effective as control mechanisms (for that is all that NZ & Aus did)
- or at least sufficiently effective to outweigh seasonality?

I'll agree again that if this becomes endemic it will be seasonal (because it's a respiratory virus wrt transmission so exposure and moving indoors does as it does for all respiratory transmitted pathogens).. but the whole point of having a f******g "pandemic response" should be that the buggering thing doesn't become endemic in the first place.

-sd

[servodude]

The two points are separate and shouldn't be conflated.

VRD

Is there a mechanism to request a BBC bytesize about "maths what can help in a pandemic'

-sd

[johnhemming]

That sounds dangerously close to saying that lockdowns plus quarantine are effective as control mechanisms (for that is all that NZ & Aus did)- or at least sufficiently effective to outweigh seasonality?

I'll agree again that if this becomes endemic it will be seasonal (because it's a respiratory virus wrt transmission so exposure and moving indoors does as it does for all respiratory transmitted pathogens).. but the whole point of having a f******g "pandemic response" should be that the buggering thing doesn't become endemic in the first place.

I think it is already endemic in most of the places it has been encountered in the world. The difficulty for NZ and Aus is that they are unlikely to be able to prevent it becoming endemic.

It is clear that the processes of infection happening now are not increasing at the rate they were earlier in the year. The question you need to look at then is whether that is because of the current restrictions or because of current levels of resistance. I would argue that both will have some effect, but the levels of resistance have a greater effect.

The evidence for that is that as restrictions were released earlier in the year the level of hospital admissions etc went down.

[XFool]

Survey uncovers widespread belief in 'dangerous' Covid conspiracy theories

The Guardian

False claims that pandemic is a hoax or was started deliberately are attracting adherents around world

[vrdiver]

It is clear that the processes of infection happening now are not increasing at the rate they were earlier in the year. The question you need to look at then is whether that is because of the current restrictions or because of current levels of resistance. I would argue that both will have some effect, but the levels of resistance have a greater effect.

The evidence for that is that as restrictions were released earlier in the year the level of hospital admissions etc went down.

Can you reconcile this with your views on seasonality/weather? Restrictions were lifted as the weather improved and more people spent more time outside (I assume).