Again reupping this paper on challenges of CFR estimation in outbreaks/pandemics with @neil_ferguson @_MiguelHernan and others. journals.plos.org/plosntds/artic…
Infectious disease epidemiologists have been working for decades -- since SARS 1 -- to develop methods to address these biases. A less technical summary of issues here washingtonpost.com/opinions/2020/…
A tour de force designed by Joe Wu and @gmleunghku for which we helped out just published today nature.com/articles/s4159… estimates symptomatic case-fatality risk of 1.4% overall rising sharply with age.
People are beginning to ask me about this amateurish paper thelancet.com/journals/lanin… that slipped through peer review. Ignoring all the work in this field over 2 decades it clumsily accounts for delay distributions that CFR while ignoring ascertainment biases that counteract this.
We are all working fast and making some errors no doubt. But a basic rule in science is "If I just found something no one else has, good chance I am wrong and they are right. My job is to make the case water tight, including by reading papers in the field."
Best scientists work hard to find the flaws in their reasoning before anyone else does and before publication. Peer review is a backstop. In this case both failed.
