Mar 11, 2020, 10:13 am
Last edit by: Ocn Vw 1K
In order to reduce noise in the Coronavirus / Covid-19 : general fact-based reporting thread, and to create a central place to invite any member to ask a basic question about the impact of COVID-19 on travel, your moderators have decided to open this separate "lounge" thread for related discussion that isn't strictly fact-based reporting.
Any member who can provide a constructive, helpful answer to a question; or post constructively in reply to a member's point-of-view, is welcome to post.
All FT rules apply, including avoiding personalized, snarky, political, other off-topic, commercial, and repeatedly disruptive content.
Discussion of general economic impacts of Covid-19 belongs in the OMNI forum, not here.
Discussion and critique of political/government actions to aid the economy or which is far more political than related to COVID-19 is for the OMNI/PR forum, not here.
This is a protocol for posting adopted by the forum Moderator team:Please follow this protocol, based on FlyerTalk Rules and long-standing FlyerTalk best practices. Doing so will help keep the thread open, and allow our moderator team to aid members, rather than having to resort to discipline.
•Constructive, respectful posts, views, opinions, questions, and replies, related to the topic are welcome. Avoid commenting on members personally, or posting off-topic or political messages.
•While respectful disagreement of a posted view is allowed, don’t call-out posters to prove their points. FlyerTalk has never required discussion standards at the level of a Ph.D. dissertation defense, or a trial court witness cross-examination.
•After a reasonable exchange of views on a point, please yield the floor so that others may bring up different topics, questions or points.
•Especially important in this time of pandemic, when normal life and travel have been upended: please take regular breaks from the thread.
Please stay healthy,
your FT Coronavirus and Travel Moderator Team.
Any member who can provide a constructive, helpful answer to a question; or post constructively in reply to a member's point-of-view, is welcome to post.
All FT rules apply, including avoiding personalized, snarky, political, other off-topic, commercial, and repeatedly disruptive content.
Discussion of general economic impacts of Covid-19 belongs in the OMNI forum, not here.
Discussion and critique of political/government actions to aid the economy or which is far more political than related to COVID-19 is for the OMNI/PR forum, not here.
This is a protocol for posting adopted by the forum Moderator team:Please follow this protocol, based on FlyerTalk Rules and long-standing FlyerTalk best practices. Doing so will help keep the thread open, and allow our moderator team to aid members, rather than having to resort to discipline.
•Constructive, respectful posts, views, opinions, questions, and replies, related to the topic are welcome. Avoid commenting on members personally, or posting off-topic or political messages.
•While respectful disagreement of a posted view is allowed, don’t call-out posters to prove their points. FlyerTalk has never required discussion standards at the level of a Ph.D. dissertation defense, or a trial court witness cross-examination.
•After a reasonable exchange of views on a point, please yield the floor so that others may bring up different topics, questions or points.
•Especially important in this time of pandemic, when normal life and travel have been upended: please take regular breaks from the thread.
Please stay healthy,
your FT Coronavirus and Travel Moderator Team.
COVID-19: Lounge thread for thoughts, concerns and questions
Mar 22, 2020, 9:55 pm
#1381
FlyerTalk Evangelist
Join Date: Aug 2014
Programs: Top Tier with all 3 alliances
Posts: 11,669
Sorry but this is completely invalid.
Firstly, you can’t assume the same baseline death rate for a sample that wasn’t randomly selected from the population. The expected number of all-causes deaths for that 60k sample is just not known.
Secondly, 60k is the number of cases, not the number of infected people (which epidemiologists in Italy have suggested is certainly well into the hundreds of thousands already). So the expected number of deaths, whatever it is (see above), in the sample is much larger than you’re saying.
Firstly, you can’t assume the same baseline death rate for a sample that wasn’t randomly selected from the population. The expected number of all-causes deaths for that 60k sample is just not known.
Secondly, 60k is the number of cases, not the number of infected people (which epidemiologists in Italy have suggested is certainly well into the hundreds of thousands already). So the expected number of deaths, whatever it is (see above), in the sample is much larger than you’re saying.
As for the difference between diagnosed and infected, we have discussed this before. The Vo Euganeo and the WP princess data suggest only another as many as the infected would be asymptomatic (and thus likely undiagnosed). From this, you can basically only double the infected numbers. I will give you 10x instead, for argument's sake. So, let's say the truly infected in Italy are 600k, 10x the officially diagnosed. Then the usual mortality rate for a group of 600k Italians will be 17 / day. We can also multiple that by 10, for the supposedly 10x unrepresentativeness in mortality sample I mentioned above. This makes 170 deaths per day, and I gave you extreme benefit of the doubt to account for your arguments. This is still 651-170 = 481 excess deaths per day due to covid-19. (using today's deaths)
Last edited by nk15; Mar 22, 2020 at 10:04 pm
Mar 22, 2020, 10:04 pm
#1382
Join Date: May 2017
Posts: 2,281
Just look at the distribution of cases by age compared to the general breakdown of the Italian population. If this was a relatively random sample I would expect it to match. Currently 70+ are accounting for 36.2% of their cases. A quick google of the age pyramid for Italy demonstrates their population is definitely not 1/3 aged 70+. This alone shows that the cases don't represent a random sample. I'm sure we could stratify further by comorbidities on top of age and see how certain groups are more likely to be identified as cases than not.
Mar 22, 2020, 10:10 pm
#1383
Join Date: Nov 2018
Posts: 765
Taking a flawed methodology and then multiplying the numbers by 10 doesn’t magically correct the flaws
The cases are nothing like a random sample. The difference between infection and case doesn’t just represent asymptomatic / symptomatic. There are limited testing resources and health workers are busy. If testing wouldn’t change the medical decision making it’s not being done. In other words, if you likely have Covid-19 but your symptoms are minor, you won’t be tested. This skews the cases towards more serious cases. More serious cases are also more likely to die, and more serious cases are also more likely to be older or have poor overall health - bringing us back to why the Covid-19 cases are nothing like a random sample of the population.
The cases are nothing like a random sample. The difference between infection and case doesn’t just represent asymptomatic / symptomatic. There are limited testing resources and health workers are busy. If testing wouldn’t change the medical decision making it’s not being done. In other words, if you likely have Covid-19 but your symptoms are minor, you won’t be tested. This skews the cases towards more serious cases. More serious cases are also more likely to die, and more serious cases are also more likely to be older or have poor overall health - bringing us back to why the Covid-19 cases are nothing like a random sample of the population.
Mar 22, 2020, 10:11 pm
#1384
FlyerTalk Evangelist
Join Date: Aug 2014
Programs: Top Tier with all 3 alliances
Posts: 11,669
Just look at the distribution of cases by age compared to the general breakdown of the Italian population. If this was a relatively random sample I would expect it to match. Currently 70+ are accounting for 36.2% of their cases. A quick google of the age pyramid for Italy demonstrates their population is definitely not 1/3 aged 70+. This alone shows that the cases don't represent a random sample. I'm sure we could stratify further by comorbidities on top of age and see how certain groups are more likely to be identified as cases than not.
Mar 22, 2020, 10:18 pm
#1385
FlyerTalk Evangelist
Join Date: Aug 2014
Programs: Top Tier with all 3 alliances
Posts: 11,669
Taking a flawed methodology and then multiplying the numbers by 10 doesn’t magically correct the flaws
The cases are nothing like a random sample. The difference between infection and case doesn’t just represent asymptomatic / symptomatic. There are limited testing resources and health workers are busy. If testing wouldn’t change the medical decision making it’s not being done. In other words, if you likely have Covid-19 but your symptoms are minor, you won’t be tested. This skews the cases towards more serious cases. More serious cases are also more likely to die, and more serious cases are also more likely to be older or have poor overall health - bringing us back to why the Covid-19 cases are nothing like a random sample of the population.
The cases are nothing like a random sample. The difference between infection and case doesn’t just represent asymptomatic / symptomatic. There are limited testing resources and health workers are busy. If testing wouldn’t change the medical decision making it’s not being done. In other words, if you likely have Covid-19 but your symptoms are minor, you won’t be tested. This skews the cases towards more serious cases. More serious cases are also more likely to die, and more serious cases are also more likely to be older or have poor overall health - bringing us back to why the Covid-19 cases are nothing like a random sample of the population.
Last edited by nk15; Mar 22, 2020 at 11:10 pm
Mar 22, 2020, 10:21 pm
#1386
Join Date: May 2017
Posts: 2,281
Sorry, that's not how statistics work. All we know is Italy on an average day will experience ~1750 deaths, and currently has ~700/day attributed to coronavirus. Given that generally people who are dying on a daily basis have comorbid conditions, and the majority of people dying from coronavirus (as someone provided above) have comorbidities, there is likely significant overlap between those two groups. Beyond that we can't calculate additional deaths attributable due to corona virus until we get a large sample of days and are looking retrospectively.
Mar 22, 2020, 11:10 pm
#1387
FlyerTalk Evangelist
Join Date: Dec 2006
Location: Pacific Northwest
Programs: UA Gold 1MM, AS 75k, AA Plat, Bonvoyed Gold, Honors Dia, Hyatt Explorer, IHG Plat, ...
Posts: 16,857
Take this article from today: https://www.rte.ie/news/world/2020/0...ew-death-toll/
“the total number of victims who have died of Covid-19 in Italy is 5,476”
”The hardest-hit northern region of Lombardy remained in a critical situation, with 3,456 deaths”
Lombardy has a population of 10 million per Wikipedia. What’s the expected death rate for 10 million compared to the covid death rate? I think it looks much worse than if you include all of Italy.
Mar 22, 2020, 11:25 pm
#1388
Join Date: May 2017
Posts: 2,281
But the covid-19 deaths are predominately in the north. For example, I read somewhere today that Rome had ten deaths.
Take this article from today: https://www.rte.ie/news/world/2020/0...ew-death-toll/
“the total number of victims who have died of Covid-19 in Italy is 5,476”
”The hardest-hit northern region of Lombardy remained in a critical situation, with 3,456 deaths”
Lombardy has a population of 10 million per Wikipedia. What’s the expected death rate for 10 million compared to the covid death rate? I think it looks much worse than if you include all of Italy.
Take this article from today: https://www.rte.ie/news/world/2020/0...ew-death-toll/
“the total number of victims who have died of Covid-19 in Italy is 5,476”
”The hardest-hit northern region of Lombardy remained in a critical situation, with 3,456 deaths”
Lombardy has a population of 10 million per Wikipedia. What’s the expected death rate for 10 million compared to the covid death rate? I think it looks much worse than if you include all of Italy.
Mar 22, 2020, 11:33 pm
#1389
FlyerTalk Evangelist
Join Date: Dec 2006
Location: Pacific Northwest
Programs: UA Gold 1MM, AS 75k, AA Plat, Bonvoyed Gold, Honors Dia, Hyatt Explorer, IHG Plat, ...
Posts: 16,857
Sure but basically the statement from you that I quoted quoted, while perhaps having accurate numbers, is still not helpful as it takes a basis of the whole country (that would be like applying Hubei deaths to all of China...). I understand that you may not have better or more granular data, but ultimately I think it makes the whole debate moot.
Mar 23, 2020, 12:51 am
#1390
Join Date: May 2017
Posts: 2,281
On the Diamond princess cruise ship which was a concentrated cesspool of viral load, everyone was tested, and where people with interior rooms were only allowed out for 1 hour a day, there were:
3,711 passengers
3,077 never tested positive (still surprising for something this contagious)
3,711 passengers
3,077 never tested positive (still surprising for something this contagious)
I didn't read into their methodology. But if they're only testing people who required hospitalization, then it would make sense their Hospitalizations+ICUs+Death = Total positive cases
Mar 23, 2020, 12:53 am
#1391
Join Date: Oct 2003
Location: Seattle
Programs: AS MVP100K, Hilton Diamond, IHG Gold
Posts: 3,215
I have found articles on countries closing their borders, but I don’t see if they’ve set a deadline for reopening them.
Obviously anything can happen within a month, but I am curious to know when Germany’s “official” re-opening of their border will be? Also Cambodia’s? This is to apply to US citizens (the only passport I hold)
Obviously anything can happen within a month, but I am curious to know when Germany’s “official” re-opening of their border will be? Also Cambodia’s? This is to apply to US citizens (the only passport I hold)
Mar 23, 2020, 1:22 am
#1392
Suspended
Join Date: Mar 2002
Location: Canada, USA, Europe
Programs: UA 1K
Posts: 31,452
It is very fluid, and nobody Is giving any firm guidance on what will happen. Just a lot of ‘at least’.
Mar 23, 2020, 1:23 am
#1393
Suspended
Join Date: Jul 2001
Location: Watchlisted by the prejudiced, en route to purgatory
Programs: Just Say No to Fleecing and Blacklisting
Posts: 102,095
Whatever the numbers are now, it should be clear by now that downplaying the risks from this virus spreading and the lack of national preparation for it are key ingredients in the recipe for increased problems from the virus spread.
Mar 23, 2020, 1:45 am
#1395
Join Date: Sep 2015
Location: Between Seas
Posts: 4,754
I have a bit of good news 
Let us hope the vastly ramped-up testing succeeds in ways that can still inform strategic decision-making.
Sunnyvale company wins FDA approval for first rapid coronavirus test with 45-minute detection time
- Under the current testing regime, samples must be sent to a centralized lab, where results can take days.
“With new tools like point-of-care diagnostics, we are moving into a new phase of testing, where tests will be much more easily accessible to Americans who need them,” U.S. Secretary of Health and Human Services Alex Azar said on Saturday.
The diagnostic test for the virus that causes COVID-19 has been designed to operate on any of Cepheid’s more than 23,000 automated GeneXpert Systems globally, the company said. The systems do not require users to have specialty training to perform testing, and are capable of running around the clock, Cepheid President Warren Kocmond said in the statement. The company said the test would begin shipping next week, but did not give further details or say how much the test will cost. -
- Under the current testing regime, samples must be sent to a centralized lab, where results can take days.
“With new tools like point-of-care diagnostics, we are moving into a new phase of testing, where tests will be much more easily accessible to Americans who need them,” U.S. Secretary of Health and Human Services Alex Azar said on Saturday.
The diagnostic test for the virus that causes COVID-19 has been designed to operate on any of Cepheid’s more than 23,000 automated GeneXpert Systems globally, the company said. The systems do not require users to have specialty training to perform testing, and are capable of running around the clock, Cepheid President Warren Kocmond said in the statement. The company said the test would begin shipping next week, but did not give further details or say how much the test will cost. -