COVID-19 Megathread 5: The Trumps catch COVID-19

[Smeulders]

Facts.

“For the first time, the CDC has attempted to offer a real estimate of the overall death rate for COVID-19, and under its most likely scenario, the number is 0.26%. Officials estimate a 0.4% fatality rate among those who are symptomatic and project a 35% rate of asymptomatic cases among those infected, which drops the overall infection fatality rate (IFR) to just 0.26% — almost exactly where Stanford researchers pegged it a month ago.”

“We destroyed our entire country and suspended democracy all for a lie, and these people perpetrated the unscientific degree of panic. Will they ever admit the grave consequences of their error?”

https://www.conservativereview.com/news/horowitz-cdc-confirms-remarkably-low-coronavirus-death-rate-media/

I'm not that optimistic about the death rate. I think one of the big priorities now should be vaccines or the protective antibody treatments that may be rolled out much, much sooner.

The CDC has put out an estimate, without any information as to how they got to that number. The CDC's estimate diverges strongly from other studies, which do tell us how they got their estimates. There is no reason to believe the CDC numbers over the others until they show their methods. (Consensus IFR seems to be in the 0.5 - 1% range)

[Smeulders]

First off, I do not necessarily think they are lying. It is not impossible that they have data to support their estimate. If so, they should produce it and people can try to figure out why estimates differ so much.

If they have no data to support the estimate, or if a reasonable analysis of their data does not support the estimate, it is not hard to guess what is happening. The CDC saying "it's not so bad" is helpful as the administration is pushing to reopen as quickly as possible.

[Smeulders]

Facts.

“For the first time, the CDC has attempted to offer a real estimate of the overall death rate for COVID-19, and under its most likely scenario, the number is 0.26%. Officials estimate a 0.4% fatality rate among those who are symptomatic and project a 35% rate of asymptomatic cases among those infected, which drops the overall infection fatality rate (IFR) to just 0.26% — almost exactly where Stanford researchers pegged it a month ago.”

“We destroyed our entire country and suspended democracy all for a lie, and these people perpetrated the unscientific degree of panic. Will they ever admit the grave consequences of their error?”

https://www.conservativereview.com/news/horowitz-cdc-confirms-remarkably-low-coronavirus-death-rate-media/

I thought at least 80% of cases were asymptomatic.  Why would they estimate such a larger % of symptomatic cases, while also estimating a much lower IFR than previous studies have concluded.  FWIW, that IFR estimate seems implausibly low based on the serology studies in NY and Spain.

A lot of the reports of very high asymptomatic rates are early in outbreaks, with more people developing symptoms later on, or in populations not representative of the general population (young navy personel for example). That part at least is plausible. Definitely agree that the CDC estimate is not easy to reconcile with the results of large scale serology studies.

[Smeulders]

2) contact tracing with enforced mandatory self-isolation is not a realistic outcome in the U.S. or any other Western democracy, and the social/political costs of such aren’t worth the risks.

Really? It worked pretty well here.

Greece is doing less testing than Kyrgyzstan (population adjusted, Greece ranks #81 in testing; the US ranks #36, ahead of Germany and Switzerland). 

So no, “your country” isn’t doing this.  Did you even watch the video?  No country is forcibly isolating all virus contacts

Circle A represents Greece having a contact tracing system.
Circle B represents testing per capita.

As you can see from this Venn Diagram, there is no overlap.

You're not going to get a lot of mileage out of contact tracing in a country that is not testing adequately to identify the vast majority of cases (Greece has only identified 3k cases, in a country of almost 11 million).  Ramping up testing is a necessary precedent of implementing robust contact tracing (that is, assuming containment is off-the-table).  That's how the two are related.

"(Not) testing adequately to identify the vast majority of cases" depends a lot of the number of cases you have. Greece appears to have very little because they intervened early and strongly. If you have very low numbers of cases, then you have very low numbers of contacts and you do not need many tests.

You may be mistaken here because there has been a lot of talk about how testing needs to increase for contact tracing (I'm pretty sure I was involved in a few of those discussions on this thread). This point was assuming the current situation in the US and other European countries with large outbreaks. This does not necessarily hold for countries where the outbreak is much smaller.

Of course, more testing is always better. If Greece ramps up testing they can start testing even people with the slightest symptoms. They'll probably find a few more cases, prevent a few more cases down the line and save a few more lives. That doesn't mean that the current contact tracing is not sufficient to keep the outbreak under control.

[Smeulders]

https://twitter.com/JoeBiden/status/1187829299207954437

A reminder that it didn't need to be this way.

Well of course no country on earth is prepared for a pandemic if you ask government epidemiologists and big pharma lobbyists.  How else would they get their $$$? 

This is like when the American Academy of Civil Engineers every year gives the vast majority of states a bad grade on road infrastructure.  Serious risk-seeking and adverse incentives are at play.   

As someone who’s parents worked with the NIH, you are absolutely insane. Government epidemiologists and NIH workers are so limited they can’t even do things that seem perfectly ethical and normal, much less can they even think about being paid to shill. There are so many regulations preventing that, it just doesn’t happen. Of course, facts don’t matter when you are pushing an anti-scientist narrative.

As someone who has worked in both Federal and state executive service, what makes you think I don’t know what the functions of regulatory big-wigs are?  Self-preservation is the name of the game.  Government agencies and their associated professional/interest groups constantly lobby for increased resources.  Saying your regulatory area is chronically underserved is one of the best ways to do that.  The CDC or NIH are no different.

I swear, for a political forum it never ceases to amaze me the juvenile approach many take to politics here.  This isn’t a conspiracy, just a general note on agency dynamics and intergovernmental lobbying you would have learned in any 300 level poli sci class at a third-rate commuter college.  The NIH and CDC aren’t exempted from iron triangles because doctors are “good guys” or whatever. 

"People always overestimate the importance of what they do, so I will disregard experts when they talk about their fields?" That has got to be one of the laziest ways of dismissing experts I've ever seen. Pretty sure you only troth that one out when experts say something you disagree with though. When Elon Musk was criticizing safety regulations you applauded him instead of noting disregarding regulations would make him money.  

I also note that you didn't even attempt to engage on the substance. That would have been pretty hard given how the lack of preparedness for pandemics is being demonstrated every day this thing goes on. No, just tried to get in a swipe at experts. You can't deny the inconvenient truths, but you sure can shoot the messenger.

[Smeulders]

Latest US. vs. European case and death graphs

Today was a funny report in the US.  It appears the US reported its second lowest number of new cases since March 27th (May 11th was slightly lower), which seems especially surprising on a Tuesday.  However, there was almost no decline in week-over-week deaths today after six consecutive strong days on that metric.  Among the states with the five highest death rates, NY and MA continue strong declines while NJ, IL, an PA may have stalled somewhat.

Declines in Europe really slowed this week.  It may just be that France, Germany, and Italy now believe deaths are low enough to be at an acceptable long term level not worth continuing restrictions.  UK is still in pretty bad shape, and Spain has been reporting almost no deaths for over a week.

You should be taking the Spanish numbers out of your posts. They are bordering on the fraudulent at this point.

The short version is that in their daily updates, they only count people who died the previous day (so if it takes more than 24 hours to register the info, they are excluded.) At the end of each week, they re-add the dead who were registered later. The daily numbers are thus a huge undercount. (Source: https://www.ft.com/content/77eb7a13-cd26-41dd-9642-616708b436)

[Smeulders]

https://twitter.com/schaller67/status/1272878014355984384

A dangerous analysis. By their very nature, you'll see more extreme outcomes with small counties. There are also more of them than large urban Clinton counties. Even if there is no link at all, you'd expect the top of the death per capita lists to be small (Trump) counties.

[Smeulders]

With the recent spike in cases in the South, it occurred to me to check something out with respect to the common argument that deaths are a lagging indicator of infections/cases.  Thus when we see a spike in cases, we shouldn't expect to see a spike in deaths until several weeks later.

This may intuitively feel true, but empirically it has not been true, at least in places that have had major outbreaks. I looked through my data for the peak in the seven-day averages for cases and deaths of the worst hit countries and states.  Most hit their peak in both within a few days of each other:

Countries:
Italy: Cases peaked March 26, deaths peaked April 2
Spain: Cases peaked April 1, deaths peaked April 3
France: Cases peaked April 9, deaths peaked April 8
UK: Cases peaked April 14, deaths peaked April 13

Early states:
Washington: Cases peaked April 9, deaths peaked April 10
New York: Cases peaked April 10, deaths peaked April 13
Michigan: Cases peaked April 7, deaths peaked April 16
New Jersey: Cases peaked April 7, deaths peaked April 21 (though there was a long plateau for both)
Connecticut: Cases peaked April 22, deaths peaked April 25
Pennsylvania: Cases peaked April 10, deaths peaked May 5 (this is the biggest exception)

More recently:
Illinois: Cases peaked May 12, deaths peaked May 13
Maryland: Cases peaked May 24, deaths peaked May 1 (?!?, also a long plateau for both)
Virginia: Cases peaked May 31, deaths peaked May 28

Interesting, I wouldn't have expected that. One thing to keep in mind though is change in testing. If testing is increasing, you'd expect the peak of people testing positive (=\= actual cases) to be closer together than the peak of actual cases and deaths.

Comparing peak to peak is also dangerous. I don't know how the distribution of time until death after diagnosis is, but this might give issues. For example, if 5% of people die the die after diagnosis, and then 1% the next 10 days thereafter, then the peak of deaths is going to be quite close after the peak in cases. New cases at time t is still going to be important for deaths at time t+10. (This is mitigated somewhat by the trajectory of cases, which descend slowly after the peak)

[Smeulders]

It is like if you find out that you have cancer. One response is to immediately seek treatment and follow the course of treatment recommended by the doctor. Another response is to maybe google a bit about the cancer, go and talk to a doctor about it, and the doctor says that you need chemotherapy, and then maybe you show up for the first chemotherapy session, but after that to decide "Hmm, isn't this fake news? Cancer is not real!" and not show up for the rest of the treatment. After all, maybe everything will get better on it's own! The human body is a wonderful thing, it can heal itself!

People who get infected now are much more likely to tested and get treatment.  We wouldn’t even hear about most of these cases if people weren’t seeking diagnosis and treatment relatively early on.

This was not true of the majority of cases in Italy and Spain (in contrast, as I mentioned above, to Portugal).  You really think a lot of people who test positive for the virus are ignoring the treatment suggestions from doctors?  I haven’t really heard much indication of that.  

We have so many more effective treatments now than we did a few months ago.  Besides all the news we’ve heard about remdesivir, plasma transfusions, and steroids, some are as simple as laying patients on their stomachs ( https://www.nytimes.com/2020/05/13/health/coronavirus-proning-lungs.html).  Whether the death rate continues to fall as it has for the past two months may depend on whether these improvements can continue to keep pace with new cases.  Maybe I’m just more optimistic about this than others, but the rapid progress we’ve seen so far is pretty miraculous.

I think you probably are. We've heard a lot about new treatments because those are interesting news stories. From the numbers I have heard, none of these improved treatments have impacts that would explain the differences in case fatality rates. I can be convinced otherwise, but if I had to guess the largest effects would be increased testing catching many more mild cases, and simply different reporting standards. We've seen that there are large differences in excess deaths vs. reported covid deaths between countries and regions. As the epidemic moves into Republican controlled states, I expect increased underreporting.

[Smeulders]

I'm just shocked at the people who think everyone should wear a mask 100% of the time.

What the hell does that nonsense even mean? Is there anyone out there saying that people should wear masks in their houses? If not, then you are just babbling.

About a month ago, I actually saw an article saying people should wear masks when home alone.

No you didn't. You are just making sh**t up.

If I recall correctly, Bandit is one of those who are getting their information through a "lockdownskeptics" discussion board on Reddit. Some idiot somewhere probably did say people should always wear masks at home. The lockdownskeptics echo chamber then spent days amplifying this and pretending this was a common opinion, so they could pretend to be smarter than the people advocating for more caution.

[Smeulders]

[..]

What is going on here? I would think, or at least certainly hope, that the most likely explanation is that this is in fact COVID-19 and there is something wrong somehow with their tests. However, if so that doesn't explain how Kazakhstan and Kyrgyzstan are able to confirm other patients as being COVID-19 positive... apparently they do have some tests that work to be able to do that... So if they have tests that work, presumably they tried using those same tests on these "pneumonia" cases...

[...]

Countries may be able to test for Covid, but they may not be able to do so at scale, or consistently. Sure, a bunch of unidentified pneumonia cases is consistent with a new epidemic starting, but it is also consistent with bad testing for Covid. I'd put my money on Covid.