With COVID-19 vaccinations now widespread, it's once again worthwhile to compare the coronavirus to the seasonal flu virus. Here's some very rough arithmetic for people who have been vaccinated:
In past years, about 30 million people have gotten the flu annually. Of that, 10 million have been vaccinated. So if you've been regularly getting the flu vaccine your crude risk of getting the flu has been about 1 in 33 each year.
About 30,000 people die of the flu each year, 10,000 of whom have been vaccinated. So if you've been vaccinated the odds of death are 1 in 33,000.
At the current rate, 50,000 people are being infected with COVID-19 daily. That's 9 million people in a six-month season, or 1 in 36. The Pfizer vaccine is 95% effective, cutting the risk to 1 in 720 for those who have been vaccinated.
As far as we know, the death rate from coronavirus is zero for those who have been vaccinated.
Here's the summary:
Have I done anything wrong here? Maybe! But it looks to me—at the current time—as if people who have been vaccinated are in considerably less danger of getting and dying from COVID-19 than they routinely are every year from the seasonal flu.
Your logic is poor. There's no good reason to believe that deaths are equally distributed over vaccinated and unvaccinated people, and there is, in fact, scientific evidence that being vaccinated for the flu reduces the probability of death.
https://www.cdc.gov/flu/vaccines-work/vaccineeffect.htm
"A 2017 study showed that flu vaccination reduced deaths, intensive care unit (ICU) admissions, ICU length of stay, and overall duration of hospitalization among hospitalized flu patients."
Add to that the fact that more vulnerable populations are more likely to get vaccinated, and a dozen other reasons you can't just play with the numbers as Kevin has done and draw meaningful conclusions.
In fact you could jump to the bottomline conclusion: It appears that no one who has been fully vaccinated has died of Covid yet. That means that for the vaccinated population *so far as we know*, Covid is less of a death risk than flu, snakebite, or lightning strike. No need to play with numbers beyond that.
But "it appears" and "so far as we know" are doing a lot of work here. Get the shot, but don't go crazy. Keep up safety practices until the numbers go down and stay down.
Yes and yes. KD’s math is wrong — 1/33 presumably is 10M out of 330M, but that would hold only if everybody got vaccinated, which obviously isn’t true because there’s at least 20M who got the flu and didn’t get vaccinated, plus presumably many others who did t get vaccinated and didn’t get the flu — and what Krowe said in all respects.
But, since the number who were vaccinated against influenza is less than 330 million, as you say, then the chance of a vaccinated person getting flu is more than Kevin's estimate. So his estimate is a cautious lower bound on what the chance really is, and his point still stands that flu vaccination does not protect against illness and death from influenza as well as the covid vaccination does against illness and death from covid.
I think vaccinated deaths frlm coronavirus are at 2 not 0.
The population of the US is about 330M. About 30M people on average become sick with the flu. About half the population gets vaccinated for the flu. The flu vaccine is roughly 50% effective.
Assume exposure to the flu is random across the population. N people are exposed. N/2 people have been vaccinated, and of those half get sick. That's N/4 people. Of the unvaccinated, N/2 people get sick. That agrees with your 10M vaccinated people get sick.
But, out of 170M people who were vaccinated, 10M got sick. That's 1 in 17 get sick.
I can't find an estimate for vaccinated people who died, but did find "The CDC said 80 percent of the children who died of flu last season were not vaccinated." If we assume that applies to the entire population... if 30,000 people die, and 6,000 of them were vaccinated, the odds of a vaccinated person dying from the flu is 6/170,000, or about 1 in 28,000.
It's interesting that we end up with one of our ratios being approximately the same.
If 9M people get infected in a population... There were N people who were exposed. About N/2 were vaccinated. About N/40 people were vaccinated and got sick. About N/2 people were not vaccinated and got sick. Roughly .43M people were vaccinated and got sick, and about 8.57 people were not vaccinated and got sick.
But now we have to model how many people got vaccinated in this hypothetical scenario. And it's just not clear what your assumptions are. To get a figure of 1 in 720, 309M people would have been vaccinated.
Iirc flu vaccines range from about 30% to over 80% effective, but they skew high, especially now that they’re 4-in-1 instead of 3-in-1, so I think the average of 50% is materially low
It seems fairly incredible to me that a vaccine would be 100% effective. Even at 99.99% effective (which seems quite astounding all by itself) ... vaccinate 350M people and 35k of them would be at risk should they be exposed.
At this point, I really don't think we have any idea of the actual effectiveness of the vaccine (and even then, there are still at least 3 levels of "effectiveness": prevents symptomatic infection that would otherwise have occured, prevents serious illness, prevents death). Even if we know what the exposure risks are (and I'm not sure we actually do), it would still be impossible to guess the likely deaths/serious illness/illness of vaccinated people at the present time.
??? We don’t have any idea? I think many thousands of highly trained and experienced scientists and statisticians would like to have a word ...
Since there have been no confirmed deaths among the vaccinated, we presumably know that it is very effective in preventing deaths, though the actual death rate may turn out to be higher than zero.
The math is rough, but there are a few ways to show that for vaccinated people, covid is less dangerous than flu.
With all due respect KD I'm calling bullshit on this post/interpretation of the data..
sincerely,
Avid follower and cat lover
Ditto I’m afraid. And many of the comments as well.
Any comparison with the common flu will suffer from a great deal of uncertainty since the statistics for it largely depend upon extrapolation from estimates. Unlike with how we get the Covid numbers, comparatively few people ever get tested for the flu in any given year. The numbers are arrived at by a number of mostly observational factors like hospitalization rates with respiratory illness and then extrapolated up by some formula the CDC has.
The nature of the flu and its vaccine also makes any comparison shaky. There are always countless flue variants floating about, and the vaccine each year is then formulated to provide resistance to just the three or four most likely variants to become problematic for the largest population. That makes 100% effectiveness against all variants impossible. Sometimes a flu variant even unexpectedly becomes dominant for which that year's vaccine wasn't formulated.
Although there are now a few known Covid variants, at least so far the vaccines seem to be effective against them so we shouldn't be surprised by any higher effectivity rate compared to the flu vaccine.
We should also consider this: The clinical trial data from the US that have been publicized likely cover almost exclusively the "original" variant. Although we really don't know that for sure because for a long time nobody in the US was looking for variants. This deliberate ignorance seems to be the reason that all variants we are talking about now were discovered somewhere abroad. Doesn't mean that all of these originated abroad. The US is a large country with lots of cases. It stands to reason that variants must have originated in the US as well.
You are basically right that the US has been a laggard in genotyping the viruses, so we are more in the dark about variants than others.
But there has been some work in this area here. And the variants that have been recognized to have arisen here have, so far, all gone extinct.
Agreed the data is patchy and that much may have been missed.
Good exposition.
I think you didn't need to spend time trying to extrapolate; just compare the efficacy rates of the vaccines and call it good.
Something to think about: COVID-19 surged in winter even though Influenza was practically nonexistent, which likely points to how much more transmissible COVID-19 is compared to Influenza. I'm not knowledgeable enough in math science to derive an equation to tease out a relative R0, but, I imagine that Influenza might be a good baseline, and going forward all other respiratory viruses might be measured in an Relative Reproduction Rate to Influenza, or RI0.
...don't mind me interchanging COVID-19 and SARS-CoV-2, BTW.
About that zero death rate: In all these trials the number of deaths was around 150. Which--with simple arithmetic--would give a death rate of 2/3 of a percent or less ("less" would cover anything from zero to 2/3 %; of course statistical uncertainty would add an error bar to this value). We have had several trails with several vaccines; all of which had not deaths of vaccinated people. So we can assume that the real figure is most likely significantly below 2/3 of a percent. But with millions of patients it is not going to be zero. It may well be below the present death rate from the flu though. Which raises the question: If we had an mRNA flu vaccine rather than a chicken egg vaccine: Would the death rate from the flu then be lower?
A key difference between flu & COVID-19 is that COVID-19 is a single virus (and its derivatives) whereas the flu is a wide group of viruses. If you don't target the right virus, it doesn't matter if you are using an mRNA vaccine or a traditional vaccine. (having said that, I will be curious if the mRNA vaccine does turn out to be slightly better... certainly, the ability to target the specific part of the virus that gives it its transmissibility seems like a good thing)
Evidently the mRNA vaccines can be developed immediately, whereas previous types of flu vaccines take much longer. This is why many people were predicting much longer times to develop vaccines for covid. If the safety-testing procedure can be shortened, there should be much faster response to different flu strains.
I believe to properly measure the risk you should perhaps adjust by:
- relative infectiousness of covid-19 vs flu: which is a factor of perhaps 2 - 4 depending on variants
Even when this correction is added however, covid-19 after vaccinations appears, for the duration of immunity, to be less dangerous than the flu
We still have two issues:
- whether the effectiveness numbers of covid-19 vaccines hold up under large scale monitoring
- how long the immunity lasts
Even so as you note, once vaccinated as a minimum covid-19 now becomes comparable in risk to the seasonal flu.
And since masks have worked so well to control seasonal flu this year ... 😉
How does that work?? The risk of the regularly vaccinated getting flu would be VaccinatedAndFlu/Vaccinated no?
Flu vaccination rate seems to be very roughly 50%:
https://www.cdc.gov/flu/fluvaxview/coverage-1920estimates.htm
So Kevin's probabilities for flu seem to be about 50% low, assuming they are based on total population of 330 million.
You can tell the math is wrong because the conversion to X in YYYY chance makes no sense when compared with actual death.
The US has about 310 million people, and according to some quick research in pre-covid times about 2.8 million die per year from all causes.
So is less than a 1% chance of dying ( I got .009032), but call it 1%, from all causes.
That is a 1 in 100 overall chance. It seems to me that the risks, when added up, need to equal 1 in 100.
Kevin tried to extrapolate among the entire population, but that only leads to possibility of error. Obviously, a 90 year old probably has more than a 1 in 100 chance in dying from anything, including the flu.
Seasonal flu cannot possibly be 1 in 33,000. If it were, that is the largest ad for getting vaccinated I could imagine.
There are 330 million people in the US.
10 million people who are vaccinated get the flu.
That does lead to the 1 in 33: but is perhaps misleading since not everyone gets the flu vaccine.
It looks like perhaps 50% of the population gets the flu vaccine, so that is about 165 million people.
Of those, 10 million using Kevin's numbers get the flu or 1 in 16.5 (the 50% error perhaps referred to elsewhere).
Again by Kevin's numbers 10,000 people who have been vaccinated die of flu so that is 1 in 16,500 chance of death from flu if you have been vaccinated.
It appears that Kevin would like to compare an approximately 6 month flu season with a six month covid-19 period.
This is perhaps a bit misleading, however.
Using Kevin's numbers 30 million people catch the flu while in a 6 month period in the US and only 9 million catch covid-19 in a six month period catch covid-19 so his numbers suggest that covid-19 is 1/3 as infectious as seasonal flu and this does not seem to match the epidemiological data.
A way to repose the question is to repose it as:
- what is my risk of dying for this flu
- what is my risk of dying for this covid-19
Without vaccinations everyone will eventually catch covid-19: effectively 100% probability of catching covid-19.
With the 95% effectiveness number about 16.5 million people will catch covid-19. Not quite twice Kevin's number for people vaccinated for annual flu catching the flu.
With an expected case fatality rate of 1% this yields 3.3 million deaths for covid-19 without vaccinations but effectively no deaths with vaccinations for individuals vaccinated against covid-19. So based on this there is less risk of death covid-19 if you are vaccinated than from flu. To make it equal to the seasonal flu we would need to have a death rate of about 1 in 30,000 for vaccinated people from covid-19. This is effectively so small that it will require long term monitoring to catch this.
Short answer:
- If vaccinated with the 95% effectiveness rates from test, you may be twice as likely to catch covid-19 ever as you might be to get sick form the flu
- if you are vaccinated with there is a nominal 100% protection against death, if vaccinated against covid-19 you are at less risk of dying for covid-19 than you are from dying from the flu.
And vaccinations at large enough scale can return us to a quasi-normal.
And, since Kevin says on average 30,000 people die of the flu each year, that's 1% of the deaths. Which makes sense considering the percentage of deaths (out of total deaths) from other causes, like cancer, heart attack, etc.
So it would be a 1 in 100 chance of dying from any cause, and a ........ ha, I'm wrong as usual! 1 in 10,000 chance of dying from the flu as opposed to something else.
Which makes 1 in 33,000 for vaccinated people make sense. Oh well.
For some definition of “we” and “zero” ... https://www.kgw.com/amp/article/news/health/coronavirus/vaccine/168-vaccinated-oregonians-test-positive-for-covid/283-3bbf23b0-6189-42d8-b5de-a0cd7a3f000e