General > General Technical Chat

Covid 19 virus

<< < (12/381) > >>

splin:

--- Quote from: Nominal Animal on March 03, 2020, 06:36:11 am ---I'm used to working with noisy data, so I know that if your dataset size is ~ 10,000 and you look at differences in the .1% range, you're basically just waving your arms without any statistical reliability.
--- End quote ---

Maybe but we aren't looking at .1% differences are we? Other than regular flu having a well understood death rate of around .1%

Ncov-19 has an infection rate of around 2 to 3, twice that of regular flu. That's 100% higher. The death rate is uncertain at this stage but appears to be between 1 and 3% or 1 to 2 thousand percent higher. There are no 0.1% differences here.


--- Quote ---
--- Quote from: splin on March 03, 2020, 05:41:28 am ---Covid 19 has a mortality rate of somewhere between 1.3% and 3.5%, so at least as bad, if not worse than the 1918 influenza pandemic.
--- End quote ---
It is also nowhere as easily transmitted as most influenza variants, as evidenced by the efficacy of the Chinese quarantine methods.
--- End quote ---

"As evidenced"? Sorry, but that is just too daft to laugh at. The best current estimates from health professionals and scientists is that it is around twice as infectious. No doubt that will get revised as better statistics become available with increasing numbers.

I'm pretty certain that the extreme quarantine measures taken by the Chinese would have had at least as dramatic impact on the transmission of regular flu if not more.


--- Quote ---  Plus, the mortality rate is actually under 0.8% for those under 50 years of age.  It is not at all clear what the actual numbers are, since we don't even have reliable detection methods yet.
--- End quote ---

So what? That's still 800% higher than regular flu. And why be selective with the figures? What about the elderly with much higher mortality rates? Don't you care about them? Accurate numbers will become available in due course but it seems clear that they are much higher than 'regular' flu.


--- Quote ---
--- Quote from: splin on March 03, 2020, 05:41:28 am ---I'd love to hear your explanation of why those dramatic images of conditions inside Chinese hospitals are no worse than conditions seen every 3 decades or so.
--- End quote ---
Because in bad flu years, there are old people dying on beds in hospital corridors in just about every European country?
--- End quote ---

Yes, but they just about cope, ususally. It's not uncommon for non emergency procedures to get postponed to accomodate more serious flu outbreaks, but it's likely to be as nothing compared to a serious epidemic of ncov-19 "as evidenced" by China having to build a whole new temporary hospital to handle the large numbers of acute patients. I doubt they've ever done that for 'regular flu' outbreaks.


--- Quote ---Most of those who are likely to contract the virus, have already contacted it in most European cities at this point, I'm guessing.
--- End quote ---

Hardly. 'Experts' here in the UK are expecting it to peak by around June. Its barely started yet. The UK government seem to be preparing for up to 80% of the population to catch the virus - but are hoping for good weather to slow it down to reduce pressure on the health services.

Maybe the public can yet be persuaded to adopt harsh isolation measures to limit its spread but I think too many people are too selfish these days for this to work. The days of the wartime spirit with everybody doing their bit for the common good seem to be long gone. I hope I've misjudged the public.

SiliconWizard:

--- Quote from: Cerebus on March 03, 2020, 02:56:56 pm ---
--- Quote from: splin on March 03, 2020, 05:41:28 am ---Without the dictatorial powers available to the Chinese state, I don't see how the rest of the world can hope to address the problem until public opinion finally kicks in. That will be very late in the day imo.

One thing is for sure, it won't take long to find out who is right and I sincerely hope it's not me.

--- End quote ---

You don't know your own countries 'dictatorial powers' well enough. Check out the Civil Contingencies Act 2004 and the Emergency Powers Act 1964. In case of an 'emergency' the government can grant itself pretty much any power and suspend pretty much any existing law that it wants to. Most 'democracies' have some such set of dictatorial powers lurking on the books.

--- End quote ---

Yup. We have actually used this extensively over here a few years back when there had been a wave of terrorist attacks, and some of the dispositions from the "state of emergency", when it was finally ended, eventually made it to the general laws. Did all this cause some heavy discussions over here? Yup it did. Was the government not able to proceed due to people's concerns though? Nope.

Similar things happened in the USA back in early 2000's (also due to terrorism), and it had lasting effects.

Those powers are not just lurking. They are effectively used, and more often than you think.

Nominal Animal:

--- Quote from: splin on March 03, 2020, 03:40:36 pm ---Ncov-19 has an infection rate of around 2 to 3, twice that of regular flu. That's 100% higher. The death rate is uncertain at this stage but appears to be between 1 and 3% or 1 to 2 thousand percent higher.
--- End quote ---
Where did you pull those number from?  As long as there is no reliable test, and the majority of infected only show mild cold/flu-like symptoms they are likely to weather at home without medical intervention, the numbers will be overestimates.

You obviously have no experience with noisy data, because at these scales comparing such small numbers with such high noise has huge error bars, "1 to 2 thousand percent higher" being as accurate as "enormous" or "tiny".  Essentially, the error bars at these scales are larger than the values themselves.  You can make statistical predictions based on them, but their reliability is very poor.

This is a perfect example of a dataset where your numbers are from a biased subset.  The apparent numbers do not reflect the true characteristics.


--- Quote from: splin on March 03, 2020, 03:40:36 pm ---"As evidenced"? Sorry, but that is just too daft to laugh at. [...]
I'm pretty certain that the extreme quarantine measures taken by the Chinese would have had at least as dramatic impact on the transmission of regular flu if not more.
--- End quote ---
Ah, you are the type that goes "that's too daft, as I'm pretty certain".  I apologise for mistaking you for someone who considers things in a rational analytical fashion, instead of just going by the feel of your pants.

This is the first time the spread of an infectitious disease seems to have been completely stopped by quarantining the affected regions.  You might be pretty certain of whatever you like, but history shows us neither European countries nor Northern America has managed that yet.  Thus far, serious disease has only been managed through vaccination programs only.  We know quarantine works, and epidemiologists have been talking about self-imposed quarantines as an effective method for these viral outbreaks for decades; average westerners have just ignored their advice completely, and keep sneering at Asians wearing breathing masks (whose intent is not to protect themselves from infection, but from spreading possible infection to others; this seems a concept too hard to understand for many).


--- Quote from: splin on March 03, 2020, 05:41:28 am ---Yes, but they just about cope, ususally.
--- End quote ---
Be careful, you're waving your hands so hard you might start flying.

Are you sure you're not trying to explain things according to your own fixed internal model, instead of looking at the various possibilities that could explain the phenomena we are seeing?  I am trying to do the latter (because I am painfully aware of my internal tendencies to do the former, and absolutely hate it when it happens: it feels to me like getting lost because of being too lazy to look at the map often enough), and have pretty consistently described these as my current opinion.  I could be wrong -- and do feel free to disagree, I and others here value the viewpoints -- but when you start saying things like "that's too daft to even laugh at" followed by "I'm pretty certain that" you sound like having to question your own beliefs or understanding is your worst fear.  It shouldn't be; it is best to be your own advocatus diaboli, to inspect the roots of your beliefs and understanding, to find out where you truly stand.

As an example of weaknesses in my own understanding, I really do not know how dangerous this situation is, because we really do not know the true number of infected.  We know the lower bound, from people admitted to hospitals, but we have no reliable detection method, so the number of infected but with minor symptomps could be huge.  In fact, the scariest scenario is that there are actually more than one virus involved here, possibly describing the reinfection cases.  (The worst case scenario still is that the mutation rate is so high that human immune systems cannot keep up with it; this also makes any sort of effective vaccine highly unlikely.)

If anyone asks me how dangerous this is, I can only say truthfully "I don't know", or lie.  I just don't have anything to correlate this with.  So, instead, I just live with it, knowing that this epidemic or something like it will repeat at least once more in my lifetime, and I will then be even more susceptible (due to age) than I am now.

hamster_nz:

--- Quote from: Nominal Animal on March 03, 2020, 09:01:09 pm ---
--- Quote from: splin on March 03, 2020, 03:40:36 pm ---Ncov-19 has an infection rate of around 2 to 3, twice that of regular flu. That's 100% higher. The death rate is uncertain at this stage but appears to be between 1 and 3% or 1 to 2 thousand percent higher.
--- End quote ---
Where did you pull those number from?  As long as there is no reliable test, and the majority of infected only show mild cold/flu-like symptoms they are likely to weather at home without medical intervention, the numbers will be overestimates.

--- End quote ---

Where do you get your numbers from? That R0 number is very freely available, from various studies - eg:

https://www.ijidonline.com/article/S1201-9712(20)30091-6/fulltext

Of cause R0 varies depending on where you measure it, a jail or hospital is different from a subway or a isolated tropical island.

The 'reliable test available' is more of a people/political thing than a technology thing. in some countries they are being actively used. South Korea have been doing 2,138 test per million people. Finland have done 23 test per million people, and USA less than 2 - but now they have withdrawn the numbers.. https://www.worldometers.info/coronavirus/covid-19-testing/

Apart from the CDC fiasco, false negatives in are mainly due to the samples not having enough RNA material to amplify, not due to failures of the test kits.

Rumor has it that the first positive test in the rest home was by a researcher who made their own test, as they couldn't get one through the CDC.

If the testing is unreliable (with lots of false negatives) then the actual R0 must be higher that indicated through testing... so you've shot yourself in the foot. Or are you saying it has false positives?

chickadee:
Ok, just a little bit of hilarity from the AP news:

https://apnews.com/dcd130364f2bbec8142d0ae47f805cba

So basically the USA media needs to be negative and sensational about everything, so this is what they came up with...

Meanwhile, in the same few days, well over 10,000 (and likely closer to 20,000-30,000!) Americans died from other things - traffic fatalities, abortions, diseases induced by all the plastics in humans bodies and other pollutants from the environment...  But this is old news to them and you won't even find it on the last page   :-//

I don't know when journalism died, but it must have been a while ago  :-\

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod