« Blood clots: is there a there | Main | Note on a Simpson's paradox in real-world vaccine effectiveness studies »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Michael Droy

Last line:
The NNT for deaths is 10000/18 = 1,250.

10000/18 = 555.5


MD: Thanks. Fixed.

A Palaz

So the NNTs are going to vary by age group

US reported death incidence by age group

<4 0.0001%
5-14 0.0002%
15-24 0.0017%
25-34 0.0073%
35-44 0.0210%
45-54 0.0596%
55-64 0.1466%
65-74 0.3595%
75-84 0.9021%
85+ 2.4335%

Also other factors in the mix such as avaliable successful therapy.

the reinfection poss ibility and length of immunity from vaccination.

A Palaz

But we need some the context too for year

Incidence. Incidence
All causes. Pneumonia
<5. 0.13% 0.0019%
5-14 0.02% 0.0005%
15-24 0.09% 0.0021%
25-34 0.18% 0.0072%
35-44 0.29% 0.0186%
45-54 0.54% 0.0518%
55-64 1.21% 0.1354%
65-74 2.53% 0.3254%
75-84 6.07% 0.7878%
85+ 18.01% 1.9121%

Georgette Asherman

If I hear another MD say "you still have a 5% chance of getting COVID with the vaccine" I will scream. Or if certain cheesy websites run articles about vaccinated people getting COVID.

An interesting statistics paper I would assign is the risk of accidents when traveling to the vaccine. In February I heard about elderly people driving in the snow 250 miles for vaccines. I felt they should have waited for something nearer. Even in New Jersey I will have to travel a total of 65 miles for the 2 trips. Friends of mine have to go south of Philadelphia.

Michael Droy

In the end the vaccine is treating Lockdown, not patients.
Just as Covid kills by an auto-immune over reaction, lockdown is an auto-immune over reaction to a limited threat.
Remove the threat (real or imaginary or just flu x2 for a 3 year period) and we can go back to normal.

Sure this is homeopathy - treating the patient's mind, not the body, but if it works it is worth it. And who cares if it is not scientific or if science is created for what is really just an end to stupid media relations.


MD: My view is the opposite. The vaccines are masking the effect of lockdowns. In the U.K., the vaccination campaign coincided with a lockdown, which they haven't eased till this week. Lockdowns reduce contacts, which prevents transmission.


GA: Great points. In my book, I have a quote of someone pointing out that there is a higher chance of dying driving to the airport than riding on the plane.

And yes, even in the worst cases, the case rate in the placebo groups of those trials was 2-3% so no one has 5% chance of catching Covid (within say 2 months). It is also amusing to see people assume that 100% of trial participants have been exposed to the virus. They can't seem to distinguish between a trial of a therapeutic in which everyone in the trial has the disease versus a vaccine trial in which we don't know how many have been exposed. Most of the uninfected people in those vaccine trials have simply not been exposed to the virus.

Michael Droy

Kaiser: Have you seen clear evidence that lockdowns have any effects?
I agree it isn't vaccines - as surely most spread comes from super spreaders and so out breaks go first through the younger non-vaccinated before elderly people get it later. In UK infections and deaths are both low, and given ever more testing infections are at extremely low levels even compared to last summer. (But if they stop lockdowns as a result, I'm happy to let them let governments/science take the credit falsely).

In the UK there is anti-body presence in some 50-55% of the population. That is surely why the natural Autumn rise winter peak and late winter collapse in infections must have happened in UK. Most of the European countries which have not fallen right off never really had a big first wave.

If it isn't enough for full herd immunity now, it certainly is enough to speed up the normal spring time collapse in infections. (the previous winter wave had a very late start).

And if we permit normal assumptions that natural immunity is a thing, or consider that T-Cell immunity might be a thing (particularly in children where no one understands why it ignores them), then that 50-55% anti-body presence might well mean something close to herd immunity.
I'm proposing this as an explanation for the collapse in UK infections, not as a guarantee of no 3rd wave or a rationale to not vaccinate.

GA: elderly travelling to get vaccines. I suspect in many cases any "legitimate" excuse to get out of the house has been grabbed. The further the better. I don't blame them.


MD: Your post illustrates the problem with every argument trying to explain the trends. In your view, it must be immunity. That requires both an assumption that immunity matters (which you explicitly did), plus an assumption that other factors such as lockdowns, masking, holiday parties and travel don't matter. Remember, the argument that there is no evidence that X matters does not justify an assumption that X does not matter. This fallacy is unfortunately also being peddled by "experts" who are saying that there is no evidence that vaccines cause rare blood clots as if that implies that they don't. In my view, a more reasonable approach is to assume all of those factors matter to some extent - and thus, a serious analysis must look at all factors in the same model, and not exclude them by assumption; or even worse, justify a single-factor analysis by an "all else equal" type stipulation.

The comments to this entry are closed.

Get new posts by email:
Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR, Wired.

See my Youtube and Flickr.


  • only in Big Data
Numbers Rule Your World:
Amazon - Barnes&Noble

Amazon - Barnes&Noble

Junk Charts Blog

Link to junkcharts

Graphics design by Amanda Lee

Next Events

Jan: 10 NYPL Data Science Careers Talk, New York, NY

Past Events

Aug: 15 NYPL Analytics Resume Review Workshop, New York, NY

Apr: 2 Data Visualization Seminar, Pasadena, CA

Mar: 30 ASA DataFest, New York, NY

See more here

Principal Analytics Prep

Link to Principal Analytics Prep