Read on Twitter
The study (or rather a short preliminary summary) about the German town that was initially hardest-hit by COVID-19 was greeted with enthusiasm yesterday, today there is a backlash. Here is a good article with the main points: https://www.zeit.de/wissen/gesundheit/2020-04/heinsberg-studie-coronavirus-hendrik-streeck-storymachine-kai-diekmann/komplettansicht
[If you don't know German, http://DeepL.com does excellent translations.] The critique is along the lines of my own objections yesterday: Unclear whether specificity is really that high and there are hardly any false positives for immunity. ...
A rush to conclusions that seems to be driven by other considerations than scientific ones. Lack of any information for underlying data or methodology, not even a semblance of a peer review, at least misleading communication as if results were representative for Germany.
One point that I missed and that seems important is that persons were counted and not households (as one person would probably infect everybody, which is different from spread across the population). Even if results hold up, there are some questions that need to be asked here.
What turned me off in particular was the herd immunity angle: 15% immunity were presented as a big step in the direction while even that would only slow growth somewhat. There seems to be a bias in that direction with the lead researcher.
I have read an interview where he makes a weird argument that a lockdown has also the downside that it slows the progess to herd immunity down. By his own estimates, herd immunity would mean at least 150k deaths in Germany. I find such nonchalance pretty appalling.
This here is even more damning. To be fair: The summary came across as so shoddy that it is unlikely the PR agency had a hand in it. They would have done science in Powerpoint, not in Word! https://www.sueddeutsche.de/wissen/heinsberg-studie-herdenimmunitaet-kritik-1.4873480
Here is a study they have done in Austria. Testing there is roughly on a par with testing in Germany. The results would point to two cases missed per one case confirmed: ie. 0.3% of the population are active cases in reality: https://www.tagesschau.de/ausland/oesterreich-corona-101.html
