Matrix Games Forums

ORIGINAL: Lokasenna

ORIGINAL: Chickenboy

ORIGINAL: Lokasenna
being able to test for antibodies is going to be critical to our understanding. We aren't able to do so yet.

Antibody studies (presumably from blood samples) are a poor way of tracking a fulminant outbreak. Serologic antibody screening is useful for determining historical exposure, but antibody production takes several weeks post-exposure to develop sufficient to detect. It's a 'rear view mirror' sort of diagnostic approach.

Viral detection, whether its RT-PCR, antigen kit detection or virus isolation is the best methodology to detect an outbreak as it is occurring. Serosurveillance will be interesting down the road to look at how many people were affected from a historical perspective.

Right, that's what I meant. It's important to help us learn about what to expect, and if we had a way to test for antibodies for this one in NYC on a widespread basis (for example), it would be helpful knowledge as infections spread to other areas. Correct? If nothing else, it would really help nail down the R0, right?

Sorry, I may have been unclear. When I said 'exposure', I meant historical exposure to the individual unit (a single person), not exposure rates between people (R0). Historical serosurveillance will only really be able to answer the question (weeks/ months later) of whether this person has ever been exposed to and mounted an immune response against this particular viral challenge. With some temporal exceptions (e.g., acute and convalescent serology), historical serosurveillance only really provides a binary "yes/no" answer.

Still, I will be interested to know-down the line-what the overall exposure rates will be for this. One hears politicos grasping hypothetical numbers of "50-70% of the population" will be exposed to this virus. I dunno. It's possible, I guess. But my gut says that's probably too high. Time will tell.

ORIGINAL: Chickenboy

ORIGINAL: Lokasenna

ORIGINAL: Chickenboy




Antibody studies (presumably from blood samples) are a poor way of tracking a fulminant outbreak. Serologic antibody screening is useful for determining historical exposure, but antibody production takes several weeks post-exposure to develop sufficient to detect. It's a 'rear view mirror' sort of diagnostic approach.

Viral detection, whether its RT-PCR, antigen kit detection or virus isolation is the best methodology to detect an outbreak as it is occurring. Serosurveillance will be interesting down the road to look at how many people were affected from a historical perspective.

Right, that's what I meant. It's important to help us learn about what to expect, and if we had a way to test for antibodies for this one in NYC on a widespread basis (for example), it would be helpful knowledge as infections spread to other areas. Correct? If nothing else, it would really help nail down the R0, right?

Sorry, I may have been unclear. When I said 'exposure', I meant historical exposure to the individual unit (a single person), not exposure rates between people (R0). Historical serosurveillance will only really be able to answer the question (weeks/ months later) of whether this person has ever been exposed to and mounted an immune response against this particular viral challenge. With some temporal exceptions (e.g., acute and convalescent serology), historical serosurveillance only really provides a binary "yes/no" answer.

Still, I will be interested to know-down the line-what the overall exposure rates will be for this. One hears politicos grasping hypothetical numbers of "50-70% of the population" will be exposed to this virus. I dunno. It's possible, I guess. But my gut says that's probably too high. Time will tell.

I'm also in the "we'll see" camp.

Somewhere between 5-20% of the US population gets the flu every year (I couldn't find an exposure number), and its R0 is somewhere in the mid 1.something range, as far as a quick google search can tell me. If COVID19 really has an R0 in the low 2.something range, it would stand to reason that the exposure rate would be higher, but I don't know how to model how much higher (just that it wouldn't be linear, so if flu is 1.4 and COVID is 2.1, you can't just multiply the flu rate by 1.5).

FWIW, the guidance we received in my work (we provide social services) from health officials was that ~35% of our workforce could expect to be infected. But it was pretty obvious that this was just a best guess from ~10 days ago.

Today's data:

+3.780 infections
+602 deaths
+408 cured

ORIGINAL: ITAKLinus

Today's data:

+3.780 infections
+602 deaths
+408 cured

Looks like the numbers are going in the right direction. It is now about 14 days after the regional lockdowns started, right?

That's another drop in daily mortality for Italy. Good news, if trending (but last week the numbers flattered only to deceive).

Italy - reason for a dash of hope today. Here's hoping it doesn't reverse, as happened last week.

ORIGINAL: BBfanboy

ORIGINAL: ITAKLinus

Today's data:

+3.780 infections
+602 deaths
+408 cured

Looks like the numbers are going in the right direction. It is now about 14 days after the regional lockdowns started, right?

Yes. However, as I mentioned in another post down in the thread, there have been several steps in the lockdown. So, some cities/provinces have done the lockdown earlier or in a stricter way. Other ones, later.

We can say that it has been 14 days in general, though.

Fingers crossed.

Today's numbers are encouraging and it's the second day in a row, after the peak of 2 days ago.

ORIGINAL: Canoerebel

Italy - reason for a dash of hope today. Here's hoping it doesn't reverse, as happened last week.

The drop you mentioned was just a statistical aberrance.

What happened is that the first villages/cities infected got into the total lockdown and therefore there has been a drop in that date. Then, since the virus had already travelled, we had the 2 days ago's peak.
Now we are in the total lockdown, so there shouldn't be places where the virus has arrived without major restrictions in place.

This week is definitely the decisive one. If we fail now, we're screwed.


To make an example, some people got infected in the small village of Codogno. Then, the area went in lockdown but it was too late since the virus had already arrived in nearby cities. That's when we got the lowering of deaths you underlined in red.

After that, everything has been closed and therefore now there shouldn't be places where the virus has arrived without being into the lockdown phase.

That's why I am plenty of hopes but also terribly worried: we cannot fail now, since we are now in the week in which we should see the effects of the full lockdown.

ORIGINAL: Lokasenna
Somewhere between 5-20% of the US population gets the flu every year (I couldn't find an exposure number), and its R0 is somewhere in the mid 1.something range, as far as a quick google search can tell me. If COVID19 really has an R0 in the low 2.something range, it would stand to reason that the exposure rate would be higher, but I don't know how to model how much higher (just that it wouldn't be linear, so if flu is 1.4 and COVID is 2.1, you can't just multiply the flu rate by 1.5).

Yeah, but that R0 for influenza really only is relevant for the negligible (except for vaccination) precautions that we take for it. No masks on commercial flights. Awful handwashing techniques and protocols. Poor 'don't come to work if you're sick' workplace protocols/enforcement. Etcetera. That paradigm has globally changed for COVID-19, so I don't expect a laboratory-determined number to be relevant in the face of such unusually stringent social distancing and other non-pharmacologic prevention efforts.

If there is a bright spot, it's in the Peltzman effect of risk homeostasis: Preventative and protective measures for COVID-19 will likely decrease the overall mortality of the most susceptible groups to other diseases that are similar risks. I expect this flu season's mortality (and next year's) to be below 'average' as a result.

https://en.wikipedia.org/wiki/Risk_compensation.

Your assessment is good to hear, Francesco. If Italy has truly peaked and started it's downward curve, hurrah!

ORIGINAL: mind_messing

Unsure if posted already - well worth a read.

Also, I was impressed with the quality of the data viz - an example of charts done right

https://medium.com/@tomaspueyo/coronavi ... 9337092b56

The most thorough I've seen and the best at showing this through visuals. Thanks

ORIGINAL: Canoerebel

That's another drop in daily mortality for Italy. Good news, if trending (but last week the numbers flattered only to deceive).

Mortality is a lagging indicator of the leading edge of the infection 'wave'. Those that die from this virus and its complications were infected some weeks ago. I too want to see the mortality drop off, but I'm more interested in the waning of 'new cases' to track the progression of the virus through a given population.

Also, I've been following Japan more closely. They've got one of the oldest populations on the planet and they've had comparatively few cases to date, in spite of widespread testing. As far as I'm concerned, they're the champs at social distancing and 'flattening the curve'-a model for others. I'd like to hear more about how they've successfully stemmed the tide so far.

Of course, they're an island. Relatively xenophobic culture. Social distancing is their normative behavior anyways. No PDAs or touchy-feely in public. Perhaps that's the key to all this.

Question, Andre, from me of peanut gallery credentials: aren't the new case numbers highly erratic/suspect so that trying to detect trends is much less certain than using mortality? And, since mortality lags, doesn't that make it a more reliable indicator that the pandemic is declining, once a true downward trend develops?

ORIGINAL: Lokasenna

ORIGINAL: alanschu

A friend of mine shared that she may have had COVID-19 in Florida in January. She had some flu-like illness that wasn't the flu with several days of fever that ibuprofen didn't do much for, and intense shortness of breath and was out for about 10 days.

Of course it's just speculation as she was never tested for the disease (it wasn't really on most of our radar in January), but I could see it being present around that time.

Indeed just speculation, although theoretically possible. Blow her case up to the rest of the population and it becomes obvious why being able to test for antibodies is going to be critical to our understanding. We aren't able to do so yet.


Worth noting that there was a particularly nasty strain of seasonal cold or flu going around in December/January. Both people in my household had it, as did just about my entire office. It started making the rounds in the first week or so of December, which was before this coronavirus really appeared.

I got it in November. And kept feeling like it hung around until late February. Interestingly this is just after my wife returned from a trip to China. No idea if there is a connection, but considering she was not in Wuhan, I'm pretty positive it wasn't Covid-19. No fever or cough.

ORIGINAL: Canoerebel

Your assessment is good to hear, Francesco. If Italy has truly peaked and started it's downward curve, hurrah!

Now there are two major considerations:

I) The decline in the numbers is a statistical aberrance. In line of principle we should see a decline during the week, therefore we can say that the current positive trend might have arrive a little bit too early than supposed.

II) The south. Southern Italy has so far basically no cases but we don't know yet whether it's because we somehow blocked the arrival of the virus to those places or we have yet to see the big wave of infections. We'll know that during the course of this week. To be sure, over the course of the next couple of weeks.


On a side note.
We, Italians, we are the worst possible ambassadors of our country, because we always tend to give extremely negative impressions of it around the world when it comes to things like this emergency.
In reality, I think that the country is managing the emergency in a surprisingly efficient and organised way. Also, it looks like the population is respecting the quarantine quite well, all considered. Of course, we are far from the "chinese way", but I think that in western societies that would have been impossible. Finally, I think that even the Chinese have experienced many, many troubles in the logistics when they had to lockdown Hubei.
The biggest issue is probably the economy, since it's quite likely that the "nordic club" (Netherlands and Germany in primis) within the EU won't help us much (if at all). I mean, Czech Republic literally stole face masks and respirators sent by China to Italy.

ORIGINAL: Chickenboy

ORIGINAL: Canoerebel

That's another drop in daily mortality for Italy. Good news, if trending (but last week the numbers flattered only to deceive).

Mortality is a lagging indicator of the leading edge of the infection 'wave'. Those that die from this virus and its complications were infected some weeks ago. I too want to see the mortality drop off, but I'm more interested in the waning of 'new cases' to track the progression of the virus through a given population.

Also, I've been following Japan more closely. They've got one of the oldest populations on the planet and they've had comparatively few cases to date, in spite of widespread testing. As far as I'm concerned, they're the champs at social distancing and 'flattening the curve'-a model for others. I'd like to hear more about how they've successfully stemmed the tide so far.

Of course, they're an island. Relatively xenophobic culture. Social distancing is their normative behavior anyways. No PDAs or touchy-feely in public. Perhaps that's the key to all this.

There is a major flaw in this reasoning. It relies on a number, the new infections, that is not really reliable.
Deaths are much easier to count and are "certain" by definition (unless the population starts to massively hide dead bodies, something I find quite... unlikely).

Now, it's probable that the number of infected people is grossly underestimated and that it reflects more the capability of a system (region/city/country) to test its citizens. Basically, if I have 10.000 positive people per-day and I can do 3.000 tests, I will never ever be able to track/test all of the 10.000.

On the other side, deaths are very easy to track even if the information we receive is delayed.


I suppose another important indicator is the number of people who got accepted in intensive care units. Still, also there, there can be under-tracking due to the fact that some people don't reach the hospital etcetc. Also, since the number of ICUs is finite, we can have that the number represents more the capacity of the hospitals, rather than information about the spreading of the virus.



I'm a strong proponent of deaths as the most reliable information. Secondarily, people admitted to ICUs and hospitalized.

On a sidenote, there were some critical observations regarding the claimed quantity of tests.

Perhaps we should differentate between high valid and work intensive laboratory tests and more 'quick-n-dirty' on location sticks

https://www.independent.co.uk/news/worl ... reddit.com

(Please let us overlook the political bias g'men, this is not the intention; unfortunately I couldn't find a more serious source in English, the relevant content was copied from a reliable german newspaper).
Meanwhile the Koreans are producing an advanced second generation. The problem of course is the unfomfortable high amount on 'false positives', which is the reason it was not choosen by the German health Departement as far as I know, only locally in some hospitals.

Regarding the real trustworthy tests, one might realize that they are not only quite work intensive, but could also only conducted by highly trained personal.

https://out.reddit.com/t3_fnhn9t?url=ht ... ame=mweb2x

“ASCLS believes we are unlikely to see the needed SARS-CoV2/COVID-19 testing capacity for at least a month and perhaps more, even under the best of circumstances.”

Finally, an overlook about the WHO's position and the reality in field...

https://euobserver.com/coronavirus/147830

It's obvious that the few capacities atm are best used for medic key personal and other decision makers.
(Sry for spelling, it's the crappy phone again as internet went down - in Hamburg?!)

ORIGINAL: obvert

ORIGINAL: mind_messing

Unsure if posted already - well worth a read.

Also, I was impressed with the quality of the data viz - an example of charts done right

https://medium.com/@tomaspueyo/coronavi ... 9337092b56

The most thorough I've seen and the best at showing this through visuals. Thanks

+1

warspite1

Its expected that Boris will announce much stricter rules tonight. I hope he doesn't go so far as the Belgians - although at least they are trying to be sensible about it. It's only non-essential threesomes that are a problem. Phew.... that's a relief.

ORIGINAL: warspite1

Its expected that Boris will announce much stricter rules tonight. I hope he doesn't go so far as the Belgians - although at least they are trying to be sensible about it. It's only non-essential threesomes that are a problem. Phew.... that's a relief.

Don't forget this important caveat: "in indoor areas"

So threesomes on the balcony or in the park are presumably OK, maybe even encouraged?