COVID-19 Lethality: How Not to Do It

Covid deaths in Sweden: Prediction by Imperial College London model (no measures and moderate measures) versus reality without lockdown (HTY/FOHM).

Published: June 23, 2020; Languages: DE, EN
Share on: Twitter / Facebook

A German medical professor claims the Covid-19 infection fatality rate (IFR) is “five to ten times higher” than the IFR of seasonal influenza. But he makes two classic mistakes.

He first compares the seasonal influenza IFR (up to 0.1%) to early model predictions of the Covid19 IFR. But these early model predictions turned out to be completely unrealistic (see above). Besides, even the US CDC in May reduced its (still conservative) best-estimate Covid IFR to just 0.26%.

He then claims antibody studies in Brazil and Spain showed an IFR of 1% or more. But they didn’t: the professor confuses the so-called “crude IFR” (deaths divided by infections in a study group) with the population-based IFR, which is adjusted for the age and risk profile of a population.

This distinction is crucial, as Covid-19 mostly affects elderly high risk groups: in fact, 40% to 80% of “Covid-related” deaths occurred in nursing homes. But if ten of one hundred nursing patients die, this doesn’t mean 10% of the whole population will die.

As an example, the famous Diamond Princess cruise ship with its mostly elderly passengers had a crude IFR of about 1.5%. Based on this value, Stanford professor John Ioannidis, already in March, calculated a population-based Covid IFR of about 0.13% for the US society as a whole.

The German professor does mention the German antibody study of the Gangelt hotspot which showed a population-based IFR of 0.36%. The professor doesn’t mention, however, that this was an upper bound: the adjusted IFR was 0.27% and the median age of death was 81 years.

Most antibody studies showed a population-based IFR between 0.1% and 0.3%, which is comparable to a severe influenza. In fact, for people under 50, the Covid19 IFR is rather lower than for influenza. A few hotspots did show locally higher IFRs of up to 0.7%, but these places usually were affected by a collapse in elderly care due to infections or panic.

More recently, immunological research has revealed that serological (i.e. blood) antibody studies detect at most 20% of infections, as most people neutralize the coronavirus with their mucosal or cellular immune system without even needing to develop (permanent) antibodies in the blood.

This means the real Covid19 IFR may drop well below 0.1% and thus into the range of seasonal influenza. It also means Covid19 “immunity passports” and mandatory vaccines won’t work. And it explains why even hotspots like NYC and Stockholm found antibody values no higher than 20%.

Despite its rather low lethality (IFR), mortality can still reach locally and temporarily very high levels if the coronavirus is spreading fast and manages to infect high risk groups in care homes and hospitals, as indeed happened in many hotspots like northern Italy or eastern France.

But even so, cumulative all-cause mortality since the beginning of the year even in hard-hit countries like the US and the UK or in Sweden (without a lockdown) remained in the range of a strong influenza season. Countries like Germany and Switzerland saw a mild “influenza season”.

The following chart shows that global Covid-19 mortality is comparable to the notorious Swine Flu “non-pandemic” of 2009/10 and remains an order of magnitude below the flu pandemics of 1957 (Asian flu) and 1968 (Hong Kong flu). These were serious, but life still went on as normal.

Importantly, while these death rates are population-adjusted, they aren’t even age-adjusted: the age group most affected by Covid-19 today (80+) hardly existed back in the 1950s. Hence adjusting for age would make Covid-19 look even less dramatic in comparison.

In contrast, the UN warns that due to the unprecedented political response to Covid-19, half of the world’s workers or 1.6 billion people are now at immediate risk of losing their livelihood. In the US alone there are already 46 million unemployment claims.

See also: Facts about Covid-19

Covid-19 death toll compared to earlier epidemics and pandemics (DB Research)

Share on: Twitter / Facebook

Up ↑