A common canard in NZ, spread by talk show hosts, misguided academics and conspiracy theorist political candidates, is that the @WHO revised down its estimate of the death rate of #Covid19 since March. This notion is based on a misunderstanding of the science and maths. (1/10)

There are two different types of "death rate" figure for a disease: the case-fatality rate (CFR) and infection-fatality rate (IFR). The CFR is simple math - the number of confirmed deaths divided by the number of confirmed cases. (2/10)

Because, as with other diseases, many #Covid19 cases have mild or no symptoms, the true case count is likely to be much higher than the # of cases confirmed via testing. Excess death statistics show deaths are likely also being undercounted, though not to the same degree. (3/10)

So while the CFR is a helpful metric in the early stages of an outbreak of a new disease to get a ballpark idea of the severity of the disease (does it kill 1 in 10 like SARS, 1 in 2 like Ebola, or 1 in 1000 like pandemic influenza?) it is not an a perfect measure. (4/10)

That's where the IFR comes in. This is exactly what it sounds like - the true number of deaths divided by the true number of cases. This is a figure that can never be determined with 100 percent accuracy because both numerator and denominator will always be estimates. (5/10)

So, back to the @WHO. On March 3, the WHO warned the CFR of #Covid19 was 3.4 percent. This wasn't a scaremongering tactic, as some might suggest, but a simple act of division: 3110 global deaths over 90,893 global cases makes 3.4 percent. (6/10) https://www.who.int/dg/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19---3-march-2020

The WHO never stated or even implied that 3.4 percent of all #Covid19 cases would die. The specific quote was maths, not prediction: "Globally, about 3.4% of reported COVID-19 cases have died". Since then, the WHO and other agencies have released estimates of the IFR. (7/10)

These estimates are, of course, much lower than the early March CFR (which, by the way, is about the CFR of #Covid19 today as well - 848k deaths/25.3m cases = CFR of 3.35%). The WHO figures the IFR is around 0.5-1% while the CDC says 0.65%. (8/10)

Those are both still many times the IFR for pandemic influenza - these figures are not small potatoes. But it is important to remember that the IFR and CFR are different. Suggesting the WHO has revised down the death rate is false and nonsensical - it's apples and oranges. (9/10)

It really is like saying the WHO has revised down its estimate of how much fruit weighs, because in March it said apples weigh 170 grams and in July it said oranges weigh 130 grams. CFR and IFR are two different things and eliding them as "death rates" is misleading. (10/10)

Addendum: There's a similar false claim going around that the US' CDC has revised down its death *count* to just 6 percent of the original number. That's equally untrue, as @BadCOVID19Takes demonstrates: https://twitter.com/BadCOVID19Takes/status/1300478066448269317

A correction: The likely IFR for #Covid19 is many times the *CFR* for other influenza pandemics. IFR for these other pandemics is even lower. https://twitter.com/marcdaalder/status/1300639413701996544?s=19