HN Explorer

Ask HN: Best data source for Covid outcome by vaccination status?

17 points by TechBro8615 ↗ HN
I’m looking for time series data that I can use to model vax and unvax as separate populations to calculate the rates of case/hospitalizations/death within each. Does such a data source exist?

Basically looking for data similar to that used in this study:

https://www.cdc.gov/mmwr/volumes/70/wr/pdfs/mm7037e1-H.pdf

9 comments

[ 4.9 ms ] story [ 34.5 ms ] thread
halfjoking
Compare these numbers with a report by the UK government (page 19 and 20):

https://assets.publishing.service.gov.uk/government/uploads/...

I don't understand how there are 1613 vaccinated deaths versus 722 unvaccinated deaths. This is a period covering Feb through Sep, so it went from the low vaccination levels at the start of the year to high levels now.

If you look at the CDC numbers in their PDF you would believe Biden's claim "this is a pandemic of the unvaccinated." 91% of the deaths from Apr4–Jul17 are from the unvaccinated. The number of overall cases among vaccinated are a small fraction of the total, where the UK numbers that's not nearly the case.

Both countries use the same Pfizer vaccine. Can someone explain the discrepancy?

angelzen
The UK vaxx rate goes above 90% in high risk older populations. Note how 1,565 of the 1613 vaxx deaths are in the >50 age group. Simplifying a bit, the absolute numbers indicate:

* The death rate in >50 and vaxx population is 5x smaller than the death rate in >50 and unvaxx population.

* The death rate in <50 population is an order of magnitude smaller, with a negligible impact when naively computing age-independent rates.

There is a report I also linked in my other comment that computes rates by age group. See tables at pages 13-16 and charts at pages 17-18.

What I cannot wrap my head around is that covid case rates for >40 age group are higher among vaxxed than unvaxxed, and 30-39 roughly on par. Only in <30 there is a significant 2x-5x reduction in the number of cases. Weren't the vaccines supposed to be effective in preventing infection?

Prediction: As time moves on, we're going to see the definition of 'fully vaccinated' shifting from 2 to 2 + k boosters, and we're going to pretend that 2 + k - 1 boosters are 'unvaccinated'.

https://assets.publishing.service.gov.uk/government/uploads/...

hndamien
My understanding was that they were supposed to be effective in preventing hospitalisation, not infection.
angelzen
There is a worrying strain of scapegoating lurking around, asserting that unvaxxed are infecting us all, and/or are a reservoir of dangerous mutations. There are ruthless politicians perpetuating such messaging and enacting discriminatory policies based on vaxx/unvaxxed status. If the UK data shows that vaxx/unvaxx makes little difference in infection rates, then the scapegoating is Not Good.
LifeIsThermal
The 95% efficacy for Pfizer was for infection prevention. Hospitalisation in younger than 65 is extremely rare, so in this group the vaccine have practically no effect. The data from Israel even indicates a slightly higher rate of hospitalisation among vaccinated in this group. At least that was the case a couple of months ago.
angelzen
The UK report also says on page 6: "With the delta variant, vaccine effectiveness against infection has been estimated at around 65% with Vaxzevria and 80% with Comirnaty" which corroborates with your information, 80% vs 95% but same ballpark. And yet their data later in the report seems to contradict the vaccine effectiveness against infection numbers. They seem to know what they are talking about, I wish they had addressed this apparent contradiction in the report itself.
jobigoud
It would be interesting to know the rate of infected but not counted, and death attributed to something else, in both cases.

There could be a correlation where people that don't trust the vaccine also tend to not trust the hospital or medical professional and end up infected but not registered. And the opposite, people wanting to get vaccinated being more likely to get tested at any symptom.

logicalmonster
One problem I have with the statistics I’ve encountered is that they often seem to categorize people who were vaccinated less than 14 days ago as unvaccinated.

This makes it harder to dive into vaccine efficacy and safety relative to unvaccinated.

It would be great for anybody seriously trying to compile this data to distinguish between vaccinated, unvaccinated, and not fully vaccinated for both 1 and 2 shot vaccines.