[smoyer]

I saw a statistical analysis of the flatten the curve idea but assuming we couldn't ramp up the number of intensive care beds, it took 3600 days to fit everyone in under the "available health care" line.

It's still a good strategy for right now because it gives us some extra time to ramp testing and healthcare up.

[samatman]

The more optimistic way of looking at this is: the 'flatten the curve' scenario has to get so close to R0 < 1, that we may as well put in a bit of extra effort to actually get R0 < 1, like China did. At which point the virus will start to isolate in pools and eventually be extinguished.

This is difficult, not impossible. We'll see what actually happens.

[neuronic]

Looking at the people happily strolling around and gathering in public cafes as if nothing was going on, I severely doubt it. (large city in Germany btw)

[asdfasgasdgasdg]

They will get with the program once one or two people they know die.

[Touche]

Those Darwin award winners will be the first to get it.

[throwaway894345]

They’ll be the first, but almost all of them will be fine. They’ll just pass it onto the elderly, cancer patients, etc who won’t be fine. No Darwinian justice here, I’m afraid.

[mathdev]

Of course there is.

[lainga]

How does that work? Killing the elderly reduces the support network available to rear your children, and so reduces your ability to sustain your genes?

[miskin]

Not in the system where state is paid from everyone's taxes to take care both about kids and elders :(.

[mirimir]

Maybe in some countries. But it's a big world.

[phkahler]

If it's drawn out slow enough, the spread should nearly stop without lockdown once 20-25 percent of the population has recovered. Assuming prior infection offers immunity of course.

[fuckknows]

I think it's a pretty safe prior to have that our immune system will work effectively against this virus post exposure. This isn't the first time a virus has crossed the species barrier in all of our evolutionary history and we've survived so far.

[shock]

I don't know we can count on that at this stage: https://thehill.com/changing-america/well-being/prevention-c...

[milanove]

The article said the patient showed signs of recovery and then later on started showing signs of the virus again. I don't know if that means they conducted an exit test to confirm that she was negative for coronavirus after her first bout with it. Maybe it died down in her system, but wasn't completely gone when she was deemed healthy again and later on it flared up again for some reason.

[shock]

There's another man who tested negative and then again got symptoms and tested positive again: https://www3.nhk.or.jp/nhkworld/en/news/20200315_13/

[Engineering-MD]

Why do you get 25%? I thought it largely depended on the R0 value

[Misdicorl]

Assuming immunity in the recovered population, R0(25% immune) = 0.75R0(~0% immune)

[tonystubblebine]

Was this the article you were reading with the 3600 days analysis? I just read it. Not sure what to make of it yet.

https://medium.com/@joschabach/flattening-the-curve-is-a-dea...

[makomk]

The "Containment works" part seems questionable. In particular, South Korea almost certainly wasn't "tracking its first 30 cases very well, until patient 31 infected over 1000 others on a church congregation". Patient 31 didn't have any known source of exposure - she hadn't travelled abroad or had any contact with known cases, didn't even meet South Korea's criteria for testing, the most likely place she got Covid-19 is from the congregation. There was an entire thousand-person cluster that had slipped through the net and only got spotted because one person happened to get in a car accident and the doctors tested her on a hunch despite not meeting their testing criteria.

If things had gone slightly differently, it's entirely possible that we'd be pointing at Italy as an example of successful containment that South Korea should have followed. They tested and contact-traced and thought they were doing quite well for the first 30 or so cases too, but reality hit them once it was far too late to do anything.

[roenxi]

A very important part of the public discourse is how many tests can conducted and at what cost.

For something with a long incubation period and potentially mild symptoms (ie, infected might not go to a hospital) only widespread statistical sampling will give a reliable picture of what is happening.

When I heard countries talking about 'testing criteria' that involved travel - I think this was in mid/late February - it seemed inevitable that things were going to drift out of control. There was no way that would pick up the mild inbound cases that could spread to become clusters. It probably developed that way because actual monitoring programs were infeasible but it'd be nice to know.

[ArnoVW]

It is my understanding that the church is sort of a cult, and it's followers hide their membership.

This illustrates the challenge with tracing approaches: if people are not open, or not sufficiently aware of the consequences of their actions, it's very easy to miss infections.

[neuronic]

Today and yesterday I have seen countless people who just dont care. Viral pandemic currently causing society to shut down?

Let's go meet in a cafe!

And that's not even the, sorry, assholes at the supermarkets. It was insane today.

[asdfasgasdgasdg]

The supermarket is not necessarily an optional trip. People have to eat, and if they're not eating in cafes, they're going to get their food at the market.

[neuronic]

For clarification, I went to the supermarket myself precisely because of the point you raised. My argument was more against the behavior that people display.

[mirimir]

And looking for TP and other paper products.

[JamesBarney]

Many countries seemed to operate with the assumption community spread wasn't happening. And due to this assumption didn't test for community spread.

[smoyer]

That is the article ... I don't think it had the "My back-of-the-envelope calculation is not a proper simulation, or a good model of what’s going on either. Don’t cite it as such" part at the bottom when I read it a couple days ago though.

[netsharc]

Yeah, I noticed the flatten the curve graphs have a horizontal line about "available beds". It probably also represents available healthcare workers, equipment, etc. Surely this line can be made to move. E.g. when China sent a plane with doctors and equipment to Italy. Or simpler, find suitable buildings and make hospitals out of them (easier said than done, obviously).

[rtkwe]

The issue is the people who need to be hospitalized in the worst cases go on to require ventilation to force air into their lungs and those machines have a limited supply. On top of that people on ventilation require pretty close observation so you need about 3:1 patient to nurse ratio. Those are the numbers that are hardest to increase rapidly, nurses and ventilator equipment.

[jodrellblank]

I heard a radio program about the life of a UK (I think) surgeon who worked in war zones and crisis regions like Syria, and one thing he found in a hospital was an intensive care unit staffed mostly by Skype cameras to a hospital in Washington, so the very few nurses on site could be directed to the patients most in need, and the remote nurses could watch the vital signs of many patients without having to spend time walking around.

If that was long enough ago to become part of a biography, it must be even more possible these days to get cameras and connectivity to remote hospitals anywhere on Earth with spare capacity of nurses.

[rtkwe]

That would allow you to move capacity around. Not sure if intubated patients could be monitored remotely though I'm just not sure if they only need vitals monitoring or if they require a present person doing something there.

That is a neat thing though I didn't know about. Wonder if hospital networks could handle the feeds properly, getting the equipment setup would take a bit but it should all be doable with consumer grade stuff.

[treis]

I seem to post this everywhere, but the flatten the curve math doesn't work. There's very little excess capacity in the system and in a scenario where half the country is infect about 1% of people would get care. Even if we perfectly flattened it over the rest of the year.

[netsharc]

Assuming the curve represents the amount of people needing hospitalization; realistically, right now, even flattened, the peak will go above that capacity limit line, and everything inside the curve but above that line are the people who need the hospital but won't get it, so... dead people?

But hey, without the flattening, the number of dead will be a shitton, and they will come fast. Probably causing social panic. With the flattening (but still not enough capacity), there will be a few deaths a day but no "Holy fuck a whole Titanic's worth of people died today"...

[rtkwe]

The actual required time is probably much lower than that. Not everyone who gets the virus needs critical care so what we have to do is keep those numbers below the healthcare capacity line.

[dharmab]

Presumably not everyone needs a hospital bed. Most are fine to take a few days of bed rest, many who would normally be hospitalized out of caution can be cared for by family at home with teledoc guidance, it's a small percentage at high risk who need hospital care.

[smoyer]

Yes ... it used the current percentage of those who get it and require hospital stays and those who need the IC along with the average number of days that they stay. Unfortunately, 70% of our capacity was full before the Coronavirus arrived so there's simply not that much capacity.

[aqme28]

About 10-20% of people need a hospital bed and we have concurrent beds for about a hundredth of that.

[bosie]

a small percentage of a large number is still a large number. 80% infected within 1 year and ~10% of those require hospitalization. usa: 350m => 80% is 280M => ~30 million hospital/ICU beds for 2-4 weeks. this is on top of regular/ongoing hospitalisation. hence it hitting any african country or india with the same force would be devastating.

[markhahn]

much less than 10% require hospitalization, though.

[bosie]

the beauty of the equation is that you can fiddle with the numbers as much as you want, it will always lead to a collapse of the health care system

[DoofusOfDeath]

How would that analysis change if we (optimistically) assume that in 2 years, we have a vaccine that's 90% effective?

I would assume that under that scenario, the backlog would disappear shortly after the 2-year mark.

[jopsen]

We'll have vaccine within the 3600 days -- I hope :)