The Naked Scientists
  • Login
  • Register
  • Podcasts
      • The Naked Scientists
      • eLife
      • Naked Genetics
      • Naked Astronomy
      • In short
      • Naked Neuroscience
      • Ask! The Naked Scientists
      • Question of the Week
      • Archive
      • Video
      • SUBSCRIBE to our Podcasts
  • Articles
      • Science News
      • Features
      • Interviews
      • Answers to Science Questions
  • Get Naked
      • Donate
      • Do an Experiment
      • Science Forum
      • Ask a Question
  • About
      • Meet the team
      • Our Sponsors
      • Site Map
      • Contact us

User menu

  • Login
  • Register
  • Home
  • Help
  • Search
  • Tags
  • Recent Topics
  • Login
  • Register
  1. Naked Science Forum
  2. Life Sciences
  3. Physiology & Medicine
  4. COVID-19
  5. Is there a way to tell how widespread coronavirus is from the numbers?
« previous next »
  • Print
Pages: [1] 2 3 4   Go Down

Is there a way to tell how widespread coronavirus is from the numbers?

  • 75 Replies
  • 33393 Views
  • 0 Tags

0 Members and 2 Guests are viewing this topic.

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Is there a way to tell how widespread coronavirus is from the numbers?
« on: 17/03/2020 15:47:35 »
Germany, 8000 cases 20 deaths, its significantly better than italy, so is south koreas average, how are they doing it ? Better testing or better healthcare, or better climate or something else

Also lombardy in italy has a population of 10 million, to date it has 1,500 deaths, its pretty bad with peak figures at present around 252 death per day, this is pretty bad. Do you think that this region is a good estimator of severity ? It seems pretty likely that this virus has been all the way around the populace by now in lombardy to me. It seems to be maxing out at about 25 deaths per million per day, but for how long is anyones guess.


Edit 8.32pm on 17th march

Its been 10 days since the italian lockdown on the 8th march, these cases are estimated to have been caught before lockdown.

France has less cases yet has a far higher ratio

Edit. 10.30pm.march 17

Lombardy has max 252 not 350 so ammended original post.

Edit 11pm march 17

In hubei the lock down began on the 23rd of january, new cases peaked on the 4th feb but deaths came to a level at the 10th feb of around 100 and continued until the end of February, so thats about 13 days for incubation peak and 19 for peak mortality, continuing for about 3 weeks.

Edit 18th march 2.41pm

The flu pandemic of 2017 2018 peaked for about 5 weeks , it totalled about 80000 deaths for the winter in the us. I would think the vast majority in those 5 weeks. Estimating about 5 deaths per million per day. Current corona virus looks to be about 5 times as lethal for about the same time scale.
« Last Edit: 31/08/2020 17:43:10 by chris »
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 



Offline alancalverd

  • Global Moderator
  • Naked Science Forum GOD!
  • ********
  • 21134
  • Activity:
    69%
  • Thanked: 60 times
  • Life is too short for instant coffee
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #1 on: 17/03/2020 18:43:33 »
You would expect the true deaths/cases ratio to be fairly constant throughout Europe as intensive care is pretty well established everywhere. But whilst death is unequivocal, the reported infection rate depends on self-reporting and accurate diagnosis, and the true infection rate will vary between urban and rural societies, and those with a social emphasis on "pubs and clubs" versus home entertaining. And those countries that suffered extreme civilian casualties in WWII will have fewer vulnerable older folk.....

In a nutshell, you can't derive much sense from the statistics.
Logged
Helping stem the tide of ignorance
 

Offline Bored chemist

  • Naked Science Forum GOD!
  • *******
  • 31101
  • Activity:
    14.5%
  • Thanked: 1291 times
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #2 on: 17/03/2020 19:22:42 »
I gather part of the reason for Italy's  apparent poor figures is that some of the outbreaks were in small towns with small hospitals- once the local hospital has all the intensive care beds full, the death rate goes up like a rocket.

Essentially they had bad luck.
Logged
Please disregard all previous signatures.
 

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #3 on: 17/03/2020 20:17:37 »
Quote from: alancalverd on 17/03/2020 18:43:33
You would expect the true deaths/cases ratio to be fairly constant throughout Europe as intensive care is pretty well established everywhere. But whilst death is unequivocal, the reported infection rate depends on self-reporting and accurate diagnosis, and the true infection rate will vary between urban and rural societies, and those with a social emphasis on "pubs and clubs" versus home entertaining. And those countries that suffered extreme civilian casualties in WWII will have fewer vulnerable older folk.....

In a nutshell, you can't derive much sense from the statistics.
Of thse points,

The ww2 generation has gone, so are most hitler youth (think pope benedict) although germany did have the soviets (famed health care though) and any losses of younger would not factor to that degree.
Lombardy is a very rich highly populated area like london of the uk.
I doubt very much that people in the country side are running around getting tested.

Perhaps the famed italian efficiency showing up the lackadaisical german attitude again.



Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 

Offline alancalverd

  • Global Moderator
  • Naked Science Forum GOD!
  • ********
  • 21134
  • Activity:
    69%
  • Thanked: 60 times
  • Life is too short for instant coffee
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #4 on: 17/03/2020 23:01:04 »
Quote from: Petrochemicals on 17/03/2020 20:17:37
The ww2 generation has gone,
As have their first generation children. On business visits to Minsk over the last 20 years it was clear that there were almost no people older than me, the city and surrounding areas having suffered something like 95% extermination between 1941 and 1944. I am 75, so in at least one potential pocket there are very few candidates at high risk.
Logged
Helping stem the tide of ignorance
 



Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #5 on: 18/03/2020 18:00:24 »
Quote from: alancalverd on 17/03/2020 23:01:04
Quote from: Petrochemicals on 17/03/2020 20:17:37
The ww2 generation has gone,
As have their first generation children. On business visits to Minsk over the last 20 years it was clear that there were almost no people older than me, the city and surrounding areas having suffered something like 95% extermination between 1941 and 1944. I am 75, so in at least one potential pocket there are very few candidates at high risk.
Nope my moms 1944 and my aunt 1937, she still remembers the war, uncle 1939.  They would have been young though. Since when has minsk been in Germany? I thought it was in Romania ?

Reallyi suppose this is a bad flu year without the aid of vaccines, I do not know when they came in?
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 

Offline alancalverd

  • Global Moderator
  • Naked Science Forum GOD!
  • ********
  • 21134
  • Activity:
    69%
  • Thanked: 60 times
  • Life is too short for instant coffee
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #6 on: 18/03/2020 21:29:07 »
No, it's the capital of Belarus, and has been for almost 1000 years. The population was mostly butchered by the occupying Nazis, then bombed by the advancing Soviets in 1944.
Logged
Helping stem the tide of ignorance
 

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #7 on: 18/03/2020 22:37:43 »
Quote from: alancalverd on 18/03/2020 21:29:07
No, it's the capital of Belarus, and has been for almost 1000 years. The population was mostly butchered by the occupying Nazis, then bombed by the advancing Soviets in 1944.
Must be why bulgaria has no cases then !
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #8 on: 18/03/2020 23:40:26 »
Ten plus days after lockdown, new cases look likely to have peaked in lombardy  like china,  peak being on the 14th, that being 7 days. Do you think a 7 day period for incubation?  If that is true 6 days after the infection peak deaths should also be peaking, which means that italy will now begin to drop gradually as of the ~20th~. It may continue for 2 weeks at a similar level ?

Maybe the china lockdown may have forced close quarters with infected people, thus extending the infection period ? Where as lombardy may have already done its damage.
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 



Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #9 on: 22/03/2020 17:19:53 »
Prevailing climate seems to be playing a part. Spain and the Uk are comparible in healthcare provision and populace but spain has the mediteranian extra life expectancy. Italy is also comparible to the uk but they did not have hind sight.
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #10 on: 23/03/2020 23:30:36 »
Thats lombardy peaking right on time, 13 days in and new cases peaked identically to wuhan. Strange thing is though that deaths in lombardy are not continuing to to rise unlike wuhan which continued to rise for 6 days after peak new cases. This number of deaths is not just anomolous, and is not linked to new cases, rather than people who where infected 13 days ago or became apparent ~7days ago~ which there was a continued dip in new cases after the 14th (7 days after lockdown) which continued until the 19th and new cases then increaced greatly.

This i summise means that when people are in lockdown transmission is taking place, 7 days between catching the virus and exibiting symptoms, and 6 to 7 days between new cases and death. Estimating that mortality in lombardy will decline until the 25th and then increace until the peak on the 27th. If so this means that the virus was not entirely spread throughout lombardy at lockdown, but it does mean that lombardy is  progressing differently than wuhan due to the 2 (one estimated) peaks in mortality. Does this mean that there where two peaks in wuhan, but due to chance the peaks and troughts are virtually undistinguishable ?

Also, we are 2 weeks behind italy, we could lend them ventilators as our first week peak (7-14  days) will coincide with there  21 to 28 day maximum , but im sure the usa or other countries could.
« Last Edit: 24/03/2020 00:55:49 by Petrochemicals »
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 

Offline Bored chemist

  • Naked Science Forum GOD!
  • *******
  • 31101
  • Activity:
    14.5%
  • Thanked: 1291 times
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #11 on: 24/03/2020 08:33:56 »
Quote from: Petrochemicals on 23/03/2020 23:30:36
we could lend them ventilators
I forget which way round it was but Germany offered to help out France (or vice versa).
Quote from: Petrochemicals on 23/03/2020 23:30:36
This i summise means that when people are in lockdown transmission is taking place,

It's pretty much guaranteed to. If you shut the family up in one house 24/7 then, if anyone has it, they will all get it.
Logged
Please disregard all previous signatures.
 

Offline set fair

  • Sr. Member
  • ****
  • 467
  • Activity:
    1.5%
  • Thanked: 19 times
  • Naked Science Forum Newbie
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #12 on: 24/03/2020 23:15:39 »
I'm persuaded by the arguement that it has to do with the way Germany compiles it's statistics. Their age demographic is second to Italy in Europe, so not that. If they had found a better way of treating patients then wouldn't we have heard about it?
Logged
 



Offline RD

  • Naked Science Forum GOD!
  • *******
  • 9094
  • Activity:
    0%
  • Thanked: 163 times
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #13 on: 31/03/2020 19:46:31 »
Now ~linear growth: ~2.5K new cases per day in UK ...


* UK COVID-19 daily increase.png (33.49 kB . 802x587 - viewed 9398 times)

https://coronavirus.jhu.edu/map.html
« Last Edit: 31/03/2020 19:48:52 by RD »
Logged
 

Offline alancalverd

  • Global Moderator
  • Naked Science Forum GOD!
  • ********
  • 21134
  • Activity:
    69%
  • Thanked: 60 times
  • Life is too short for instant coffee
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #14 on: 31/03/2020 23:18:06 »
Be wary of mortality statistics. Cardiologists were puzzled for years as to why the French, who smoke more and eat more dairy products and red meat than the British, had fewer deaths from heart disease. Red wine? Extramarital affaires? Turned out that they just didn't like writing it on death certificates in the absence of objective postmortem  evidence - a cultural matter rather than a scientific one.

Now here we have a disease that is particularly virulent among the elderly with cardio-respiratory comorbidities. So did he die from COVID, pneumonia, or cardiac insufficiency? Pretty much a matter of opinion. Testing doesn't really help - it may confirm that the patient died with COVID, but not from it. 
 
And we still have a huge imbalance of numbers and types of tests between countries. Whilst death may be fairly unequivocal (even allowing for a variance in reporting preferences) the number of confirmed cases depends on the number of credible tests and the pool of infected persons. Plus there is a variation in test outcomes: an antibody test will quickly indicate ongoing infection or a successful recovery from an infection that the patient did not consider sufficiently serious to refer, whereas a PCR test for viral RNA (a) takes a lot longer to report and (b) only detects active virus. 

Once you have resolved the morass of statistical uncertainties, you might be able to discern a difference in treatment outcomes, but that would be better investigated by direct comparison of procedures, which may vary as much within a country as between countries.
Logged
Helping stem the tide of ignorance
 

Offline RD

  • Naked Science Forum GOD!
  • *******
  • 9094
  • Activity:
    0%
  • Thanked: 163 times
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #15 on: 02/04/2020 20:53:12 »
I spoke too soon ...


* UK COVID-19, new cases, daily increase new ,1st April 2020.png (44.92 kB . 977x631 - viewed 8048 times)

https://coronavirus.jhu.edu/map.html
Logged
 

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #16 on: 02/04/2020 22:21:42 »
Quote from: RD on 02/04/2020 20:53:12
I spoke too soon ...


* UK COVID-19, new cases, daily increase new ,1st April 2020.png (44.92 kB . 977x631 - viewed 8048 times)

https://coronavirus.jhu.edu/map.html
Quarantine uk comes to the 13 day period around sunday 19 the following saturday, so a way to go yet. Most worrying thing is our hospitals where not full at the 200 mark, yet 200 per day where passing even with full medical treatment.

Lombardy  is appearing different to wuhan as its peak mortality was 13 days in also, appears to be fully spread in lombardy, as from the 13day post quarantine to the 19 mortality is equal ie about a 1 to 1 transmission , maximum 550, but really maxing at about 350, or 35 per million per day. That is 2000 per 60 million per day if all at once or staggered over the country at about 1500, so these quarantine measures are levelling the curve throughout the country. After 14 days quarantine appears to have had all of its desired affect and at 28 days the virus appears to have run its course in Wuhan. I wonder really how long lombardy will take to descend from the 350, +28 days is the 11th april, how to calculate the infection percent ?

There has been some question on the wuhan numbers, claims of mis reporting/ under reporting.

« Last Edit: 04/04/2020 23:29:56 by Petrochemicals »
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 



Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #17 on: 10/04/2020 18:06:19 »
Varying degrees of lockdown



Thats 19 days for the uk today since lockdown on the 10th march

19 for new york yesterday on the 9th or today on the 10th

France and spain reached 19 days on the 2nd of april

Lombardy has reached the 28 day since lockdown with the same pattern as Wuhan, 2 weeks-ish post 13 day  peak on the and deaths have continued at a steady level of around the 350 average. Looks alot like the chinese clamped down alot faster, but i cannot see any indication of severity of infection in lombardy other than it had two peaks in mortality, one on the 13 and one on the 19 day after lockdown. It is currently at a mortality rate of 0.1% on the lombardy population.




« Last Edit: 10/04/2020 18:09:16 by Petrochemicals »
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #18 on: 15/04/2020 22:46:47 »
Quote from: RD on 31/03/2020 19:46:31
Now ~linear growth: ~2.5K new cases per day in UK ...


* UK COVID-19 daily increase.png (33.49 kB . 802x587 - viewed 9398 times)

https://coronavirus.jhu.edu/map.html
Strangely its now linear in the uk around the 4000 5000 mark, has been since the 1st april which is around a week into lockdown, even after 21 days in lockdown and looks like it will be linear for about 3 weeks.  Cases did peak on the 13th day after quarantine, but deaths peaked the 17 rather than the 19
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 

Offline Petrochemicals (OP)

  • Naked Science Forum King!
  • ******
  • 3629
  • Activity:
    8%
  • Thanked: 182 times
  • forum overlord
Re: Is there a way to tell how widespread coronavirus is from the numbers?
« Reply #19 on: 06/05/2020 21:33:14 »
Uk becomes the worst afflicted in europe, with it spread to most major conurbations. We have also passed another milestone, passing the yearly death toll for flu colds and pneumonia.

https://www.dailymail.co.uk/news/article-2804670/80-people-die-Britain-DAY-flu-pneumonia-Europe.html

Although University Oxford estimates 600 a year,.

https://www.express.co.uk/life-style/health/1237961/coronavirus-fatality-rate-vs-flu-how-deadly-is-coronavirus-flu-death-toll-2020
Logged
For reasons of repetitive antagonism, this user is currently not responding to messages from;
BoredChemist
To ignore someone too, go to your profile settings>modifyprofie>ignore!
 



  • Print
Pages: [1] 2 3 4   Go Up
« previous next »
Tags:
 
There was an error while thanking
Thanking...
  • SMF 2.0.15 | SMF © 2017, Simple Machines
    Privacy Policy
    SMFAds for Free Forums
  • Naked Science Forum ©

Page created in 0.514 seconds with 70 queries.

  • Podcasts
  • Articles
  • Get Naked
  • About
  • Contact us
  • Advertise
  • Privacy Policy
  • Subscribe to newsletter
  • We love feedback

Follow us

cambridge_logo_footer.png

©The Naked Scientists® 2000–2017 | The Naked Scientists® and Naked Science® are registered trademarks created by Dr Chris Smith. Information presented on this website is the opinion of the individual contributors and does not reflect the general views of the administrators, editors, moderators, sponsors, Cambridge University or the public at large.