9 comments

[ 4.4 ms ] story [ 32.4 ms ] thread
(comment deleted)
Just making sure this paragraph from the middle of the article gets more visibility:

    The Health Ministry said the rate of serious cases was
    far higher among unvaccinated Israelis aged 60-plus, who
    constituted 151.5 people per 100,000 in serious
    condition on Monday; among the vaccinated the figure was 
    19.3, and among the partially vaccinated 40.9.
Serious conditions appear ~8x more often per-capita among unvaccinated compared to vaccinated populations.
How did you ended up with 8x figure?
151.5 / 19.3 = 7.8497

I rounded to 8.

(comment deleted)
Maybe I am wrong but shouldn't you normalize (in a Bayesian way) to the percentage of vaccinated people in the sample? I.e. we should look at the probability of contracting serious COVID for a vaccinated person vs a unvaccinated one.

Since 60-plus people are very likely to be vaccinated the effect of the vaccine could be much lower than reported. It's the textbook case of confusing conditional probabilities and while physicians surely know what they do it is simple for published data to be in a form that is misleading for the average reader.

EDIT: My bad, the data is already this way BUT if you look at https://datadashboard.health.gov.il/COVID-19/general (Hebrew) you see how the effect of the vaccine has decreased

Probability of contracting COVID for an unvaccinated 60+ y.o. was 15x than a vaccinated one in February, now it's more like 3x.

Probability of getting a serious case for an unvaccinated 60+ y.o. was 30x than a vaccinated one in February, now it's more like 8x.

I don't know for sure what they meant, but that's my interpretation of the numbers (and ambiguous phrasing) given in the article. Do we have any better/more detailed sources?
Thanks for the clarification in the edit! I would guess delta is the main cause of the reduced effectiveness?
This is impossible. They told us the vaccine would stop spread so we could all go back to normal. They must have made mistakes testing these people, or tabulating the data, or something.