Read on Twitter
1/
In the German-language part of Twitter, the top-trending hashtag right now is #SterbenmitStreeck
- an extremely heated discussion about the scientific work (incl. the "Gangelt study": https://www.nature.com/articles/s41467-020-19509-y), false predictions & well-publicized opinions of @hendrikstreeck https://twitter.com/benthamsghost/status/1332972489354436609
In the German-language part of Twitter, the top-trending hashtag right now is #SterbenmitStreeck
- an extremely heated discussion about the scientific work (incl. the "Gangelt study": https://www.nature.com/articles/s41467-020-19509-y), false predictions & well-publicized opinions of @hendrikstreeck https://twitter.com/benthamsghost/status/1332972489354436609
2/
@hendrikstreeck contends that #SterbenmitStreeck supporters wish his own death - but in rebukes, they seek to draw attention to the fact that he predicted no second wave and (maybe) herd immunity. Streeck now maintains that he was misunderstood re: forecasting NO second wave.
@hendrikstreeck contends that #SterbenmitStreeck supporters wish his own death - but in rebukes, they seek to draw attention to the fact that he predicted no second wave and (maybe) herd immunity. Streeck now maintains that he was misunderstood re: forecasting NO second wave.
3/
The Gangelt study has been criticized for - despite focusing, including in the title, on the infection fatality rate - only following infected patients for 2 weeks (deaths usually peak around 14 days after an infection was detected).
The Gangelt study has been criticized for - despite focusing, including in the title, on the infection fatality rate - only following infected patients for 2 weeks (deaths usually peak around 14 days after an infection was detected).
4/
The Gangelt study found an eye-popping high rate of previously or currently infected patients (by PCR and antibody testing): 15.5% inhabitants of a village that experienced a super-spreading event, while the official number of patients was much lower, 3.1%.
The Gangelt study found an eye-popping high rate of previously or currently infected patients (by PCR and antibody testing): 15.5% inhabitants of a village that experienced a super-spreading event, while the official number of patients was much lower, 3.1%.
5/
At the same time, @hendrickstreeck, the first author, and colleagues reported an infection fatality rate (IFR) of only 0.36% - lower than expected 0.5% to 1.5% or so. A recent meta-analysis inferred an IFR of ~0.7%.
At the same time, @hendrickstreeck, the first author, and colleagues reported an infection fatality rate (IFR) of only 0.36% - lower than expected 0.5% to 1.5% or so. A recent meta-analysis inferred an IFR of ~0.7%.
6/
The study's conclusion fueled the public debate and informed politicians such as the position of North Rhine Westphalia's prime minister @ArminLaschet.
As @capitalMagazin's articles describe, eg
https://www.capital.de/wirtschaft-politik/pr-fuer-heinsberg-studie-streeck-und-laschet-gehen-in-deckung
there was a PR firm supporting these positions and people
The study's conclusion fueled the public debate and informed politicians such as the position of North Rhine Westphalia's prime minister @ArminLaschet.
As @capitalMagazin's articles describe, eg
https://www.capital.de/wirtschaft-politik/pr-fuer-heinsberg-studie-streeck-und-laschet-gehen-in-deckung
there was a PR firm supporting these positions and people
7/
A big problem with the study is that some deaths might not have been counted. E.g. @ntvde and @medwatch_de report that the IFR might been twice what has been published - because patients were only followed for ~14 days.
A big problem with the study is that some deaths might not have been counted. E.g. @ntvde and @medwatch_de report that the IFR might been twice what has been published - because patients were only followed for ~14 days.
8/
Whoever's side you support - I hope you agree that we need rigorous research on this question. What seems clear is that IFR depends on the age distribution (eg a cluster in a nursing home in this small town would have thrown off the numbers in the other direction).
Whoever's side you support - I hope you agree that we need rigorous research on this question. What seems clear is that IFR depends on the age distribution (eg a cluster in a nursing home in this small town would have thrown off the numbers in the other direction).
9/
Science lives from lively debate and new data has to be incorporated into the existing study of work - e.g. Prof. #Ioannidis of @METRICStanford & @METRIC_Berlin is updating his meta-analysis.
Without facts accepted by all, this ain't a discussion but a messy food fight!
/fin
Science lives from lively debate and new data has to be incorporated into the existing study of work - e.g. Prof. #Ioannidis of @METRICStanford & @METRIC_Berlin is updating his meta-analysis.
Without facts accepted by all, this ain't a discussion but a messy food fight!
/fin
