On Tuesday the 6th of October Donald Trump posted the following on Twitter and something similar on Facebook after having recovered from COVID-19:
"Flu season is coming up! Many people every year, sometimes over 100,000, and despite the Vaccine, die from the Flu. Are we going to close down our Country? No, we have learned to live with it, just like we are learning to live with Covid, in most populations far less lethal!!!"
The posts were quickly censored, but still - could it be that he might be partly correct? After all, he said “most populations”, and we already know that the flu and COVID-19 affect age groups differently. Could it be that for very young populations, the flu actually is more lethal? The answer is not obvious.
There is much talk about how COVID-19 compares compares with influenza. Firstly, this is not an easily done comparison since “the flu” can be caused by several different viruses with different fatality rates from year to year. Secondly, since the flu is not considered as deadly as COVID-19, it is actually harder to find good research and detailed figures on infection fatality rates for the various age groups. Thirdly, because a huge portion of the world population has already had many or most types of flu viruses, in general they do not get as sick as with a new virus like SARS-CoV-2, or they even don’t get sick at all.
2017-2018 was a bad flu season in the US compared to the others in 2010-2020, as seen from the table on https://www.cdc.gov/flu/about/burden/past-seasons.html .
https://www.cdc.gov/flu/about/burden-averted/2017-2018.htm shows various estimates on the 2017-2018 flu season for various age groups. One can assume that the figures in table 1 is fairly accurate, especially when it comes to medical visits, hospitalization and deaths. The numbers on asymptomatic cases are missing.
https://www.cdc.gov/flu/about/keyfacts.htm estimates the asymptomatic percentage of infections to be around 45%-ish.
If assuming that 45% is a fair approximation, it means that 44.8M / 55% = 81.5M Americans had the flu virus that season, of which 36.6M were asymptomatic. I.e., the spread reached 81.5M / 325M ~= 25% of the population before it ended. This lower percentage than usual herd immunity levels assumed for SARS-CoV-2 is likely to be due to already present immunity in the population, as a combination of vaccine and previous years’ infections.
If we assume that the asymptomatic cases were spread proportionally in the given age groups, we get the following IFR values:
|0 - 4||6 687 895||115||0,0017%|
|5 - 17||13 659 275||528||0,0039%|
|18 - 49||26 232 845||2 803||0,0107%|
|50 - 64||24 068 967||6 751||0,0280%|
|65 +||10 810 345||50 903||0,4709%|
|Sum||81 459 327||61 100|
As a simplification, we do not adapt the different IFRs to different spread levels, even if there are some differences. Applying these IFRs, linearly approximated, to each year of age, unsurprisingly yields a total number of flu fatalities of around 215000 for the entire US population (i.e. if the full population actually got the flu virus that year). The COVID-19 number is approx. 9 times higher. This ratio will be different for other countries since they have different population mixes.
Since the spread of the flu is much lower in reality (typically less than 30%), the calculations estimates a number of approx. 65000 real deaths. (This is close to, but does not precisely match CDC's estimate exactly since the original age intervals for IFRs are too wide and within the years of age within each interval, and the spread level actually varies a little bit for each age group. However, as an evidence it is sufficient.)
The deaths to compare the 65000 figure to for COVID-19 is approx. 1.3 millions. In other words, 20 times as many.
COVID-19 is in general many times more lethal than a bad flu virus infection, since:
For the youngest part of the population, however, the chance of dying if you get infected by the SARS-CoV-2 virus can be less than the chance of dying if you get a flu virus. The pivot point depends on the "type" of the flu virus and how it affects various years of age. When using the 2017-2018 virus, the pivot point was at approx. 35 years. This chance of dying at these young ages is so tiny, for both viruses, that it is practically close to nil compared to the older parts of a population.
Also, getting a bad flu virus is actually harder than getting the SARS-CoV-2 virus this season due the lower spread percentage of the flu. Based on that one could actually claim that the chance of dying is higher for SARS-CoV-2 vs. the flu for all ages.
The www.CoronaSim.eu table includes the flu figures for all countries based on the above calculations. The “Fatalities, realistic max” column assumes a flu spread of 30%, which would be a really bad flu season. Similar settings for calculating YPLL and IFR_hsa are used for the flu as for SARS-CoV-2.
As can be seen from the CoronaSim.eu table, the flu is less lethal than COVID-19 for all populations of the world, with clear margins.
Trump was basically just wrong.
The graph below shows the US IFRs for various ages: