Thread by @GidMK

Lots of people are still comparing the COVID-19 death rate with influenza ("it's only as deadly as the flu!")

This is very misleading

If you compare apples to apples, COVID-19 is MUCH deadlier than the flu

The figure that most people use when talking about the death rate of influenza is 0.1-0.2% (i.e. 1 to 2 deaths per 1,000 cases of flu)

They then say that the lower estimates of COVID-19 fatality rate are similar, so it's comparable

But they really aren't

The 0.1-0.2% figure for influenza is one estimate of the CASE fatality rate - i.e. the number of deaths divided by confirmed cases

The CFR for influenza varies significantly, but is around 1-40 deaths per 1,000 people infected (0.1-4%)

Comparing to COVID-19, we can see that influenza, with its CFR of 0.1-4% is actually quite a bit lower than the CFR of COVID-19 varying from 0.1-20%

This is largely to do with testing - if we only divide deaths by CONFIRMED cases, we ignore all of the people who are never tested for either disease

Which is where the INFECTION-fatality rate (IFR) comes in

A 2014 systematic review looked at CFR and IFR estimates of influenza during the H1N1 pandemic

Now, this was a *bad* year, but it's still interesting to see what the numbers are

ncbi.nlm.nih.gov/pmc/articles/P…

According to this review, if you use the denominator of everyone who is infected (determined from serology tests) the IFR of influenza is 0.001-0.01% (1-10 deaths per 100,000 infections)

In comparison, the LOWEST estimate of COVID-19 IFR that I've seen anywhere comes from an Iranian paper that has been criticized for undercounting deaths

They estimated an IFR of 0.09% (90 deaths per 100,000 infections)

So even using the LOWEST POSSIBLE FIGURE for COVID-19 and the HIGHEST POSSIBLE FIGURE for influenza, the infection-fatality rate is still 9x higher for COVID than the flu

A more realistic estimate, using the plausible ranges for both of these diseases, puts COVID-19 at around 100x more deadly than influenza

That is really quite a lot

It's also worth noting in this thread that even using the influenza IFR is somewhat misleading, because these IFRs are calculated differently

The influenza IFR is generally calculated using modelled EXCESS mortality attributable to influenza

COVID-19 IFR is instead being calculated by simply dividing confirmed deaths by inferred cases

So even this slightly better comparison, that says COVID-19 is about 100x deadlier than the flu, is using different numerators (although the same denominator) and thus may underestimate the risk

To take a leaf out of @d_spiegel's book, this means that you have a similar risk of death from a single COVID-19 infection to 50-100 influenza infections over the course of one year