Daily coronavirus-related chat, questions, and minor updates - April 23

NDR Coronavirus Update Podcast with Christian Drosten; here's my cliff notes from yesterday's episode.

You can go look for the details in here: https://www.ndr.de/nachrichten/info/coronaskript182.pdf if you're curious, just use deepl to translate what you're interested in. The notes are in order to help you look. Also have a look at the references at the end of the script.

Most infectious 1 day before symptoms start, 4 days after mostly not infectious, 44% of infections before begin of symptoms; why hospital outbreaks then?

Explanation: The studies indicating infectiousness relate to normal everyday life; domestic situations e.g.; Intubation might be risky for medical professional even if it happens ~1 week after symptoms start => more direct access to deeper lungs makes this potentially more hazardous. Weaning of intubation might be done without high PPE standards because the patient might (!) not be infectious.

Risks to other organs include: Lung and lung embolism, kidneys, neurological issues (sense of smell, e.g.). Studied observationally currently and we're also looking for more effects. Sense of smell does not imply risk for brain apparently.

Relaxation Orgies: (Opening discussion orgy (Öffnungsdiskussionsorgie) is a term that has entered parlance recently after being used by Merkel)

People think "the worst is mostly over, just a little longer then we'll open up. But: UK reports excess deaths in early april. This demonstrates thoroughly that this is not "just a flu" but is in fact way worse. I don't have to tell you that excess deaths comingles direct and indirect deaths and that's ok. Apparent cases of indirect excess deaths at home might actually be caused by influenza. An old person in fragile health might not show that they're dying right now to a flu. With all that: A influenza season (christmas to carnival) might get you a few 100 excess deaths in germany.

https://www.nytimes.com/interactive/2020/04/21/world/coronavirus-missing-deaths.html displays a lot of that all across the board; approximately doubling of deaths in spite of measures we're taking. Let's hope this trend doesn't continue. Germany is in a good spot because we started early.

Politicians apparently need two conditions to enact measures: 1) An image of where we don't want to go. Refrigerated trucks, hospital beds in the hallways, etc. and 2) "The problem is here!" - 2) happens usually because of deaths in the country. But that's a lag of 1 month approx.

Germany started early because of testing. We didn't need deaths because we tested a lot, so our politicians acted sooner. For example, we noticed community spread initially because a lab decided "hey, we're doing a batch of influenza tests, let's run a covid test on them too." e voila, our first case of community infection that we were not able to trace back to a known case.

Christian Drosten regrets that we seem to be wasting our good position: We're in a unique position because we a) are populous b) have decreasing case counts and c) know well what's going on here. Politicians seem to be intent on wasting that in favor of relaxation measures when we could hope for a quicker reduction in case numbers instead. A rare show of an explicit political opinion from the man. Relaxation measures btw that effectively open up shopping malls again.

Some French study:

R decreased from 3.3 to 0.5. Models suggest that when the lockdown (11th of may) is over, maybe 6% of people in france have had covid-19. Even after the entire moving-around-ICU-patients-by-train in ile de france and grand est, only ~12% there have had the disease.

They also modelled severe and lethal cases, but exclude elder care facilities from their modeling here:

2.6% need hospitalization

IFR projected to be 0.53% (in line with current estimates)

for (older than 80, not living in elder care) persons, IFR: 8.3%

What this means for germany: Most of the data is applicable, we're also seeing an increased involvement of the elderly and elder care facilities. Case fatality in germany will go up.

More on the situation in germany:

Mild lockdown with huge success. Contrast FR, IT, ES where you can't even go on a run. It should really have been possible for us to keep this up some more. (Note from points he previously raised: Keeping R just under 1 will result in a geographical and social diffusion of cases without clear clusters. A worsening of R will see a hard to contain outbreak all over. A relaxation triggering this is what worries him most I think.)

R < 1 and every slice of the german economy gets weird phantasies of how they can reopen with weird protective measures; or they just think that it's their turn to get some relaxation next. (You want relaxation, go use your god damn free time and trade backrubs with your wife. Argh!) Christian reports this because apparently people email him that kind of stuff.

On mutations of the virus:

We could previously observe mutations that did not affect how the virus acts. Study from china, 11 patients, early outbreak (22 Jan - 4 Feb), already heavily discussed online to wrong conclusions: So: Phenotype charactecization is what this study is about. We can judge a virus' dangerousness by observing its reproduction in a cell culture. Additonally we can directly observe the cells' immune response in a cell culture. In these kinds of tests we could expect to see evidence of a strain being more virulent, e.g. by evolving an answer to cell immune responses. This is all a bit more involved, all the way up to animal tests.

The study in question shows particular representativeness of the current genealogy of the virus, i.e. these are more or less the roots of the current genetic tree. On to the spicy part: Apparently, the difference in growth of the virus amongst some of the samples is a factor of 270. In a cell culture. Easy to make scare headlines out of that, we all know how it goes.

No Peer Review though. At the tail end of exponential growth, things naturally deviate very widely; this was hardly controlled for. A closer look: it's not a matter of one sample being 270x worse than the rest or anything. It's a matter of the total difference between best and worst is 270x; seems to be a worse outlier, and in fact that's usually the case (your average mutation is bad). So this is a spread in replication, not an increase. 270x also isn't that bad if the replication overall is a 100000x or more. Additionally, a technical error: the initial dose of virus injected into the cell cultures deviates by a factor of 100. That's a big deal. He'd tell his PhD students: Cool, super interesting, but you're going to have to do all that over again.

TL;DR: Meaningless study.

And that's that. I wrote quite a lot, but that's from 40 minutes of an expert talking. Suuuper interesting podcast, wanted to share.

ETA: Can I just say how much of a goddamn good source this guy is? And he has this podcast. If anyone wants to explore the entire series by auto-translating it: https://www.ndr.de/nachrichten/info/Corona-Podcast-Alle-Folgen-in-der-Uebersicht,podcastcoronavirus134.html - I'm always ready to check the translations if you've got suspicions.