Hacker News Reader

Comment by WalterBright | Hacker News Reader

We've been vaccinating people since December. There ought to be statistical evidence now of whether and how much vaccinated people still get sick from Covid and its variants. Why do we still hear that epidemiologists have no idea about this?

For example, everyone being tested for covid or being admitted to the hospital for covid, should be asked if they've been vaccinated, when, and which vaccination.

70 million people in the US have received vaccinations. There ought to be plenty of data by now.

View on HN

We do have real world results, and they’re mostly very good. For example, analysis of the real world effectiveness of the Pfizer vaccine in Israel:

> The latest analysis from the MoH proves that two weeks after the second vaccine dose protection is even stronger – vaccine effectiveness was at least 97% in preventing symptomatic disease, severe/critical disease and death.

https://www.pfizer.com/news/press-release/press-release-deta...

The AstraZeneca vaccine looks not so good, especially against the South African variant, but the early real world results for the Pfizer and Moderna vaccines look outstanding, and the Johnson & Johnson vaccine looks solid.

Do we have any educated guesses on why we see a difference between the J&J shot and the AZ one? Is it just luck of the draw regarding different adenovirus vectors, or something deeper?

For one, AZ doesn’t use the pre-fusion stabilized spike protein in their vector, while the mRNA and J&J vaccines do.

Disclaimer no experience with vaccine development but...

Two problems with some types of engineered vaccines. The first is people can have existing immunity to some component of the vaccine. And the immune system wipes out the vaccine before it has a chance to generate strong immunity. Second people develop immunity to proteins in the vaccine that aren't present in pathogen you're trying to vaccinate against.

As an example I think one of the vaccines based on modified viruses uses a different virus in the first and second shot.

On the latter, you mean the sputnik vaccine.

Note however that Israel has closed its borders tightly, so Israel can't tell us much about the vaccines' effects on variants that haven't entered that country yet.

I've only seen reporting suggesting that every variant listed as a "variant of concern" from the CDC is present in Israel. So their data should be good.

> I've only seen reporting suggesting that every variant listed as a "variant of concern" from the CDC is present in Israel.

But not present in large numbers now, so we have to wait months before we'll know if they'll be able to spread in the vaccinated population or not.

> The AstraZeneca vaccine looks not so good

This study, from Scotland, looks at risk of hospitalization after first dose of either Pfizer or AstraZeneca vaccine. The Pfizer and AstraZeneca vaccines look equally good. If anything, in the raw numbers AstraZeneca looks a little better that Pfizer, but the difference is well inside error bars.

https://www.bmj.com/content/372/bmj.n523

> especially against the South African variant

There is this one study, a proper randomized controlled trial from South Africa with 2000 people. But it only studies the AstraZeneca vaccine. They only saw mild (15 vaccine recipients and 17 placebo recipients) or moderate (4 vaccine recipients and 6 placebo recipients) infections, so they can not say anything about severe disease, because no one got it.

And they only studied AstraZeneca. Unfortunately, this study has been giving AstraZeneca maybe an undeservedly bad reputation, as media goes around saying that AstraZeneca specifically is not effective against the B.1.351 variant, suggesting that other vaccines would be. But as far as I know, similar studies of other vaccines do not exist, so we just don't know yet if they would be, maybe, equally bad against the B.1.351 variant or not.

https://www.nejm.org/doi/10.1056/NEJMoa2102214

> analysis of the real world effectiveness of the Pfizer vaccine in Israel:

In Israel vaccinated population is not tested for COVID-19 unless admitted to hospital in serious condition.

In the past two month 22% of those who died because of COVID-19 have received both doses of the Pfizer vaccine.

MoH refuses to clarify further on testing policies and outcomes.

Public health England has been tracking that sort of information in their SIREN study.

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

This should be the top comment. I really wish this was addressed somewhere.

But with all the charts we see of case, hospitalization, and death counts, why isn’t someone charting data for case, hospitalization, and death counts following vaccinations?

You might see some study with possibly some data buried in it, but this should be part of the rest of the data being broadly shared so that it’s clear the effects it’s having.

There are many studies on the Israeli population post-vaccination, since so many Israelis have finished the two-dose course of the Pfizer vaccine. There are even some studies on the British variant there, since that is now responsible for >50% of infections. TL;DR: the Pfizer vaccine is extremely effective against the British variant as well.

There ought to be statistical evidence now of whether and how much vaccinated people still get sick from Covid and its variants. Why do we still hear that epidemiologists have no idea about this?

How much energy do you want to put into tracking so that you can draw better conclusions?

We have good evidence that people don't become symptomatic, because that is easy to track. But we haven't generally been constantly testing to see how many caught it and fought it off. Or how many could have spread it to unvaccinated people while they were asymptomatic. Which are both very useful for epidemiologists to know for modeling purposes.

The data from Israel is pretty good: https://ourworldindata.org/vaccination-israel-impact

How do you collect the data? How do you validate it? How do you make sure your treated and control arm are balanced?

Agree we should now have data to look at but don't? They can have their kayfabe but when it comes to me injecting stuff to believe it I have questions.

There are thousands of actually qualified professionals looking at that data, including the licensing bodies of every developed nation. The EU has recently gone back-and-forth on the AZ vaccine. which would seem to only make sense under the assumption that they are actually cognisant of possible risks and trying to avoid them (and alternative some might want to argue is their hatred of the British, but that cannot explain the ultimate resolution, since EU still very much hates at least the Johnson government)

If you trust them, get the vaccine. If you don't trust them, why would you trust any data they might release?

Reply on news.ycombinator.com