Read on Twitter
This article was just published on how far we are from herd immunity for COVID-19. Although the title is definitely correct, the article has a surprising number of important issues and errors in it. https://www.nytimes.com/interactive/2020/05/28/upshot/coronavirus-herd-immunity.html
First, they show graphics of seroprevalence in different locations as if they are accurate estimates of the population exposed. There are 7 locations in the main figure. I might have missed something but I have not seen a paper of any kind for any of the studies. None.
Given possible biases in serosurveys the results could be off by sizeable amounts (e.g. 2x). Again, biases aren't going to change conclusion that we're far from herd immunity, but showing seroprev # as if they are from an unbiased published study when they aren't isn't helpful.
Error #1: Article says infection fatality rate of flu is 0.1-0.2% with no source. That is much higher than estimates I've seen. This thread w/ refs suggests flu IFR: 0.02-0.05%, or 2-10x lower than NY Times article. Perhaps they mean the CFR? https://twitter.com/AdamJKucharski/status/1243466404415909889
This is not helpful given the huge controversy over IFRs, and confusion b/w IFR and CFR: https://twitter.com/DiseaseEcology/status/1252844190070829056
Finally, there is a grossly oversimplified quote at end of article by @AndrewNoymer. I think goal was to say that no immunity exists to SARS-COV-2, but all 328M Americans are NOT equally susceptible to infection, & this matters for herd imm: @mgmgomes1 https://www.medrxiv.org/content/10.1101/2020.04.27.20081893v3
You can do better @nytimes.
