COVID-19 Megathread 4: Grandma Got Run Over by the Dow Jones

[parochial boy]

Wow, look at the tracker today. Nearly every state seeing decreases in new cases compared to yesterday. Looks like the tide has turned more quickly than anyone expected and Trump has saved the country from disaster, expect a big approval bump.

That tracker you're citing may not be entirely accurate:

Coronavirus death toll: Americans are almost certainly dying of covid-19 but being left out of the official count

As if we ever come close to to accurately counting deaths from the flu or other respiratory viruses that primarily kill older folks with serious preexisting conditions.  There’s an almost unanswerable argument here about how to code these deaths in people who are really sick - does blaming the coronavirus or the flu (or anything else) for the death of someone who would have died in a few days from a severe underlying illness make sense? 

The article does actually explain it... you can calculate a fairly accurate death toll by simply comparing the number of people dying over a time period versus the number of people you would normally expect to die over the same time period. So someone who was already terminally ill wouldn't get included in that sort of calculation if they already would have died within the time period during which the epidemic occured.

[parochial boy]

Oh I know it was bureaucratic nonsense but the buck stops with Trump. It was his own people doing the bungling. He did finally put his foot down and got the testing flowing, like it is now. It was just a few weeks too late. We can only speculate if Hillary or someone else would have done better, but Trump's government screwed it up. That is what we know for sure.

As I said my understanding that it's what experts-epidemiologist (not bureaucratic) of CDC decision to proceed like this. If there is any evidence that experts of CDS were demanding more testing, but Trump/his bureaucrats screw it up, link to it.

I already linked to CNN article that explained about CDC routines, here is 538 about testing and diff vs South Korea
https://fivethirtyeight.com/features/how-coronavirus-tests-actually-work/

Compare that to South Korea, where people can get their results in about a day. There, the government had been stockpiling the necessary chemicals for years after COVID-19’s cousin MERS briefly hit that country in 2015. That helped the country move quickly to approve and decentralize testing as soon as COVID-19 arrived.

See, a structural problem that US governments didn't stockpiled, not specifically Trump's or Obama's. Most European Countries are doing worse/same.

https://twitter.com/natesilver538/status/1245449988186025985

Bug and feature of US system.

Even now states like mine just aren't testing enough and they aren't recommending tests unless you fall under a strict criteria.

Take a look at Europe. Even stricter criteria, even less testing for most countries. Explanation is not that their current governments were "bungling", but that their systems are structurally different from US and that most governments have been "bungling" in decades.

Germany is an obvious exception. Fast and massive testing from the beginning.

Except I'm not really sure that the point holds that well.

If you look at the systems that have been doing really badly in terms of ramping up testing you get:

France/Italy/Spain - all have overwhelmed healthcare systems where testing has been dropped as a priority just to be able to keep up. In Italy and Spain's case you are coming off initially missing the boat on the emergency exacerbated by a decade of of political turmoil and austerity politics that have clearly damaged the healthcare systems. They're also, don't forget, much poorer countries than the USA. In France, well anybody who follows French politics will be perfectly aware that the French healthcare was already heading into a crisis before the pandemic, driven by funding cuts and a "privatisation" of the system.

The UK - funding cuts that have deeply damaged the system over the last decade, plus a political response that makes that was even more incompetent than Trump's - pretendig there was no problem for weeks after it already clearly was. (both the French and the Brits took far too long to cancel their soccer leagues, for example, which have turned out to have been a massive incubator for the disease).

The Netherlands - which recently liberalised/privatised it's system, and has one of the least "socialised" systems in Europe.

Doing the best, as in better than the USA, you have Germany/Austria/Denmark/Norway/Iceland which all have "socialised" healthcare with differing levels of centralisation. Followed by Switzerland, which has a less socialised system, but where testing and care has been overwhelmingly taken on by the cantonal authorities and the public hospitals - the private insurers were actually one of the principal reason Switzerland was kind of slow of the marks.

I mean, the fact that the US has driven straight past all of these countries in terms of per capita cases and (except for Switzerland - which was basically hit at point blank range by what happened in Italy) and per capita death,s even despite the being a long haul flight away from any of the outbreak zones while all the aforementioned countries being far more exposed to Italy and Spain kind of shows you immediately that the US response was not great.

So maybe to a degree the level of decentralisation matters, if it gives healthcare systems some degree of insulation from political decisions, but overall there isn't honestly a lot of overlap between "socialised" healthcare and an especially efficient or inefficient ability to ramp up levels of testing. I very much get the feeling that the political response, including in the run up to the crisis, combined with the economic context seem to be much bigger predicators of how well or badly a healthcare system has reacted than any actual structural aspects within the healthcare system.

[parochial boy]

Data as of Apr. 8 at 4:03 a.m


https://www.washingtonpost.com/graphics/2020/national/coronavirus-us-cases-deaths/


Cases:

Worldometers says over 33,000 cases yesterday. What explains the discrepancy?

Timing I suspect, I have not been following the US numbers in detail, but I know that Worldometers often gets the Swiss numbers wrong if they miss one one of the larger cantons that release their numbers in the evening. As they are showing 27k cases for the US on Monday and 33k for Tuesday, I expect it's a similar issue where someone released some data that Worldometers didn't pick up until the next day - they're a pretty good aggregator by all means, but I wouldn't take their word as gospel.

[parochial boy]

...That’s not how it works. Only if you do nothing and if the disease doesn’t mutate is herd immunity the eventual outcome. With a significant lockdown, we can reduce the r0 factor to the point the disease dissipates over time, leaving it only in certain areas which we can close off when necessary while reopening the rest of the country.

There was a (German language unfortunately) youtube video that got a fair amount of buzz here, and looks at some of the various epidemiological studies that have been done on how achievable "herd immunity" is. Basically the conclusion is that it isn't achievable.

In order to achieve it ether you have let the thing run it's course, and completely explode the medical system in a way that is just not feasible (and it's not even up for debate any more that this would be the result). Or, even in a country with a lot of intensive care capacity, in the less optimistic scenarios, you need a lockdown-slash-social-isolation that reduces the R0 to below 1.1 and would need to keep those measures in place for as long as two years. Again, completely impossible as an option given the economic and social consequences.

So the conclusion is, the only thing you can do is reduce the number of infections back to a level where contact tracing is feasible, at which point you can ease the restrictions. The hope being that these should be more succesful the second time round as jurisdictions and healthcare systems are prepared, but people themselves are also going to be more careful about avoiding big events, washing hands, social distancing as they know what will happen if they don't. In that way, we can have some degree of normality until the vaccine is ready, but maybe don't expect many big sporting events between now and the end of the year.

[parochial boy]

I wouldn't rule out counting methods as having a pretty dramatic impact either. Just eyeballing the data being released by some of the Swiss cantons, about half of all recorded fatalities have been outside of hospitals, mostly in care homes, and this has been a proportion that has been growing since the start of the epidemic. Over the last few days, a majority of recorded fatalities have been amongst non-hospitalised victims, and that has been the reason that the total number of deaths has only just started to decline a couple of weeks after the peak in new infections and even after a steady decline in the number of hospitalised deaths. I am getting the impression that this picture has basically been replicated amongst other countries that have been reporting both hospitalised and non-hospitalised deaths.

The upshot of that is clearly that jurisdictions that are only recording deaths of people in hospital who have tested positive (thinking New York and the UK in particular here) are going to be pretty dramatically understating the real impact of the disease with their official figures.

[parochial boy]

Important data points here:

Sweeping testing of the entire crew of the coronavirus-stricken U.S. aircraft carrier Theodore Roosevelt may have revealed a clue about the pandemic: The majority of the positive cases so far are among sailors who are asymptomatic, officials say.
...
The Navy’s testing of the entire 4,800-member crew of the aircraft carrier - which is about 94% complete - was an extraordinary move in a headline-grabbing case that has already led to the firing of the carrier’s captain and the resignation of the Navy’s top civilian official.

Roughly 60 percent of the over 600 sailors who tested positive so far have not shown symptoms of COVID-19, the potentially lethal respiratory disease caused by the coronavirus, the Navy says. The service did not speculate about how many might later develop symptoms or remain asymptomatic.

https://www.reuters.com/article/us-health-coronavirus-usa-military-sympt-idUSKCN21Y2GB

this is around the same % that we got in the italian town of Vo'

I'm not sure if the insane asymptomatic rate on this thing is a good or bad thing. On one hand it's good because it means the mortality rate is extremely inflated and it's probably actually more like the flu than SARS in regards to danger, on the other hand it makes the virus harder to control and record.

IDK whether halfish asymptomatic cases (showing up pretty consistently now between Italy, Iceland, the Diamond Princess and now the Theodore Roosevelt) is a good thing or a bad thing?  It would lower the fatality rate of course,  but only by about 1/2, which if true takes from perhaps 10X more dangerous than seasonal flu to perhaps 5X.  Still quite scary. The flip side is that it makes this thing extremely hard to contain.  If it was 10X more dangerous than flu but only people who already had a high fever were contagious, it could be a lot easier to contain with health screenings.

IMO you either want to hope there are 10 or 100 contagious asymptomatic cases per person who gets sick (to massively drive down the fatality rate and quickly achieve herd immunity) massively or none (so that you can control it with fever screenings).  Halfish doesn't seem to help.

Think it's worth pointing out that the big asymptomatic case numbers that were being reported from the Diamon Princess were at the time of testing. If you follow up from the perspective of several weeks later, then, the number of asymptomatic cases was... signficantly lower.

As much as anything, there have been 12 deaths from 712 reported cases so far. Seeing as it's pretty unlikely that there were a huge number of non-identified cases on the ship, I feel that should probably call into question some of the more optimistic assessments of what the fataility rate for Covid is.

[parochial boy]

Important data points here:

Sweeping testing of the entire crew of the coronavirus-stricken U.S. aircraft carrier Theodore Roosevelt may have revealed a clue about the pandemic: The majority of the positive cases so far are among sailors who are asymptomatic, officials say.
...
The Navy’s testing of the entire 4,800-member crew of the aircraft carrier - which is about 94% complete - was an extraordinary move in a headline-grabbing case that has already led to the firing of the carrier’s captain and the resignation of the Navy’s top civilian official.

Roughly 60 percent of the over 600 sailors who tested positive so far have not shown symptoms of COVID-19, the potentially lethal respiratory disease caused by the coronavirus, the Navy says. The service did not speculate about how many might later develop symptoms or remain asymptomatic.

https://www.reuters.com/article/us-health-coronavirus-usa-military-sympt-idUSKCN21Y2GB

this is around the same % that we got in the italian town of Vo'

I'm not sure if the insane asymptomatic rate on this thing is a good or bad thing. On one hand it's good because it means the mortality rate is extremely inflated and it's probably actually more like the flu than SARS in regards to danger, on the other hand it makes the virus harder to control and record.

IDK whether halfish asymptomatic cases (showing up pretty consistently now between Italy, Iceland, the Diamond Princess and now the Theodore Roosevelt) is a good thing or a bad thing?  It would lower the fatality rate of course,  but only by about 1/2, which if true takes from perhaps 10X more dangerous than seasonal flu to perhaps 5X.  Still quite scary. The flip side is that it makes this thing extremely hard to contain.  If it was 10X more dangerous than flu but only people who already had a high fever were contagious, it could be a lot easier to contain with health screenings.

IMO you either want to hope there are 10 or 100 contagious asymptomatic cases per person who gets sick (to massively drive down the fatality rate and quickly achieve herd immunity) massively or none (so that you can control it with fever screenings).  Halfish doesn't seem to help.

Think it's worth pointing out that the big asymptomatic case numbers that were being reported from the Diamon Princess were at the time of testing. If you follow up from the perspective of several weeks later, then, the number of asymptomatic cases was... signficantly lower.

As much as anything, there have been 12 deaths from 712 reported cases so far. Seeing as it's pretty unlikely that there were a huge number of non-identified cases on the ship, I feel that should probably call into question some of the more optimistic assessments of what the fataility rate for Covid is.

That population skewed older than average which would almost certainly cause a higher mortality rate. Also it’s important to note that the patients on the cruise ship may have been exposed to higher initial viral load as compared to the general population (outside of hospitals)
Taking this into account, I still would believe the true mortality rate is around 1%.

Probably yes. I wouldn't like to guess myself, but from what I've seen, the mid point of the majority of estimates of the moratility rate is around 1% or slightly lower. It's just more a note of caution to the people who have been repeating the rate from that one study from Germany just because it seems to show what they want it to show.