Ioannidis et al (2020) forecasters.org/blog/2020/06/1… make the claim that the IFR from Spanish Flu was ~100 times worse than that if #COVID19. This claim seems to stem from estimates of 100 million deaths due to Burnet (1979) pubmed.ncbi.nlm.nih.gov/85614/ and
Johnson & Mueller (2002) ../1
Johnson & Mueller (2002) ../1
jstor.org/stable/44446153 .
But these are extreme high end estimates. Patterson & Pyle (1991) jstor.com/stable/44447656 estimate no more than 30 million deaths, and report several lower estimates ../2
But these are extreme high end estimates. Patterson & Pyle (1991) jstor.com/stable/44447656 estimate no more than 30 million deaths, and report several lower estimates ../2
More recently Spreeuwenberg et al (2018) academic.oup.com/aje/article/18… estimate less than 20 million deaths and yield global excess death rates of around 1% of the adult population ../3
Furthermore studies of the situation in India, the worst hit well documented area confirm that excess deaths rates were less than 5% Chandra & Kassens-Noor 2014 bmcinfectdis.biomedcentral.com/articles/10.11… ../4
Whereas a demographic analysis me-jaa.com/October2015/Li… Bhagawati & Choudury (2015) find excess deaths of about 2.3% in India in 1918. Life expectancy at birth in India before the pandemic was an astonishingly low 25 years; so many in India were particularly susceptible ../5
So what can we conclude? The IFR of Spanish Flu depends on how many were infected; but if we assume an attack rate of 50% globally we arrive at a IFR of around 2% or 3%. 100 times less, so ~0.025%. But estimates for #COVID19 are 0.65% or higher. cdc.gov/coronavirus/20… ../6
So surprise the Spanish Flu was not 100 times more deadly than #COVID19. Perhaps around three times worse would be closer to the mark. So the claims of Ioannidis et al (2020) are totally spurious. 7/
@threadreaderapp unroll
