Hacker News Reader

Comment by mlyle | Hacker News Reader

> How does that make sense?

Looks like the vaccine creates a broader set of antibody responses than natural infection does. If you want the population to be protected against the next wave of illness, this is the way to go.

> (against a strain of SARS-CoV2 that no longer exists).

That's the funny thing here. Getting boosted with B.1.1.7 doesn't just cause antibodies to develop towards B.1.1.7, but instead for your beta cells that make things against things-kinda-like-B.1.1.7 to mutate until they're creating all kinds of similar antibodies (preferentially drifting towards things that are B.1.1.7-like).

View on HN

> Looks like the vaccine creates a broader set of antibody responses than natural infection does.

That was the theory, but it is not being borne out by the actual data. See below.

> If you want the population to be protected against the next wave of illness, this is the way to go.

The data coming out of the UK and Israel does not support this claim. Their populations are already vaccinated and boosted at percentages much higher than those in the US, but those vaccinated and boosted people are still getting infected and are still being hospitalized and dying in significant numbers.

> That was the theory, but it is not being borne out by the actual data. See below.

I'm not sure where you're pointing "below", but I've read a lot of papers and titrations of neutralization across many variants of different subpopulations.

> The data coming out of the UK and Israel does not support this claim. Their populations are already vaccinated and boosted at percentages much higher than those in the US, but those vaccinated and boosted people are still getting infected and are still being hospitalized and dying in significant numbers.

Israel's had about half the COVID death per capita of the US in 2022, and is probably better at reporting deaths as COVID-related. The UK has about 65%. Boosted subpopulations have fared much better within all of these 3 jurisdictions than the population at large.

> Boosted subpopulations have fared much better within all of these 3 jurisdictions than the population at large.

Not lately, according to the data the UK has been reporting. The UK has about 90% of its population vaccinated with at least one dose, and about 70% of its population have had boosters. In last week's UK report, nearly 90% of the new COVID hospitalizations were vaccinated people, and over 90% of the COVID deaths. And 87% of the new COVID hospitalizations and 78% of the COVID deaths were boosted people. When the percentages of hospitalizations and deaths of people in a given category are equal to, or greater than, the percentage of people in that category in the population at large, we would not normally say that category of people has "fared better than the population at large".

Of course, as of this week, the UK will stop providing the data that I have just described. So we won't be able to see how these various populations continue to fare. That doesn't look to me like a government that is confident in the benefits of vaccines.

> Not lately, according to the data the UK has been reporting

Original research or are you repeating the well-known debunked tripe despite knowing better?

The unadjusted rates in population subgroups are in table 14.

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

Every single age subgroup has a significantly lower risk of death when vaccinated. The risk of hospitalization is also lower, but less dramatically. This is despite cases being recorded in the vaccinated at a higher rate over this interval (which is probably an artifact of the data rather than being real).

If you look at the raw sum without considering the subgroups, your number is confounded by the fact that the vast majority of the deaths are in people 80+, who are also vaccinated at a higher rate than the general population.

This is about the best known statistical "paradox" that there is-- https://en.wikipedia.org/wiki/Simpson%27s_paradox

Another confound is that we really don't know the number of unvaccinated people. Counting errors get magnified. If you think you have 1M people in a locality, and you know you have vaccinated 930,000 people... you get a huge swing in rates if you actually have 980k vs 1020k people.

> Original research or are you repeating the well-known debunked tripe despite knowing better?

No, I'm looking at pp. 42 and 44 of the report you linked to.

> If you look at the raw sum without considering the subgroups, your number is confounded by the fact that the vast majority of the deaths are in people 80+, who are also vaccinated at a higher rate than the general population.

Actually, the figures I gave for percentages of new hospitalizations and deaths were for people 80+. (I had misread the tables thinking that that row was totals.) You are correct that the vaccination percentages for this age group are higher than for the general population, but they're still about the same as the percentages of new hospitalizations and deaths for this population.

I have not tried to check the rates given in table 14 based on the numbers in the previous tables.

> Another confound is that we really don't know the number of unvaccinated people.

The report you linked to seems to be pretty confident in the percentages it gives of vaccinated people. That implies that the UK government believes it has accurate counts of the total population (or at least accurate enough that any uncertainty in the count doesn't significantly affect the rates they are showing).

> Actually, the figures I gave for percentages of new hospitalizations and deaths were for people 80+. (I had misread the tables thinking that that row was totals.) You are correct that the vaccination percentages for this age group are higher than for the general population, but they're still about the same as the percentages of new hospitalizations and deaths for this population.

Well, then you really screwed up, because there's a table in the report that shows for the 80+ subpopulation, the unvaccinated died at a significantly higher rate.

Or, using your table:

(23+175+1112)/1415 deaths in age 80+ had some level of vaccination = 92.57%, vs ~96% of the age 80+ bracket having some level of vaccination. If a group has ~7.4% of the deaths but is ~4% of the population, I'd say they're at excess risk!

> (or at least accurate enough that any uncertainty in the count doesn't significantly affect the rates they are showing).

It makes a massive difference for this number as the vaccination rate nears 100%. If it's 95% of the population that's vaccinated, the risk is ~50% higher. If it's 96%, the risk is ~80% higher. If it's 97%, ~150% higher. In any case, the conclusion holds up for any sane denominator here, but it's hard to have any confidence in the exact odds ratio.

> I'm not sure where you're pointing "below"

This earlier comment, maybe: https://news.ycombinator.com/item?id=30973197

> I'm not sure where you're pointing "below"

The latter part of my post.

How significant are we talking here? How does their death rate compare with those that chose to or could not be vaccinated?

> Looks like the vaccine creates a broader set of antibody responses than natural infection does.

That's not true. It's narrowly targeted at the S-protein, whereas natural infection will result in antibodies to the S and N proteins.

> That's not true. It's narrowly targeted at the S-protein

Granted, because that's the immune reaction you want, as targeting the N protein will marshal an immune response but not be neutralizing.

Natural infection does not produce as broad of a neutralizing response to different variants of the S protein.

It is still based on the Wuhan strain. Like getting a 2018 flu shot in 2022.

Reply on news.ycombinator.com