Hacker News Reader

Comment by allturtles | Hacker News Reader

I don't think this is a particularly interesting stat without a baseline of what % of people fell into that category. Since this is computed over the time period January 2 to July 2 and mass vaccination didn't really take off until the spring, you wouldn't have a large number of people in that category until near the end of the study period.

View on HN

I think you need to read it again. Especially table 1.

In the data they have they found that:

1. For all unvaccinated people, they had 38,964 covid-related deaths and 65,170 of other, non-covid-related deaths of unvaccinated people.

2. For all vaccinated people 21 days or more after second dose they found 458 covid-related and 57,263 non-covid-related deaths.

It doesn't matter when these were recorded (unless we want to account for Delta, but there is not enough data for this here).

So unless we are seeing huge increase in non-covid-related deaths of vaccinated people, this data suggests that vaccination makes it many, many times less likely that you are going to die of covid as compared to you dying to other things.

Which I think is wonderful news to see in actual numbers from a large sample of data gathered over longer period of time.

If that's is not a clear indication to you then I don't know what is.

The ratio of Vax vs. Non-vax death proportions is 1.3%.

So,50X to 100X reduction in deaths versus all-cause

Yes although the vaccine was given first to the most vulnerable people (ie more likely to die from other things) so this still doesn't give a clear sense of the improvement. No one is doubting the effectiveness of vaccines (ok a bunch of idiots, but not me) but this isn't the most helpful stat to assess the relative improvement.

The data you mention provides extremely strong support for vaccine effectiveness in preventing death. The single data point in GGP post is quite useless, as GP points out.

The elderly were prioritized in getting vaccinated, and they have a higher baseline death rate. So it's still not clear what it means that the # of non-covid deaths in (1) and (2) is roughly the same.

I'm not disputing that the vaccines are effective at preventing death, but that doesn't mean that we can just point to any stat that looks good and celebrate without considering what it actually means.

Wouldn't that just point to the vaccine being even better than the stats suggest (by just looking at ratio of deaths of vaccinated / unvaccinated) since the vaccinated group would be the more vulnerable in the first place.

Also there are more people vaccinated than not now so that would also skew the data to make a direct comparison look worse.

> So unless we are seeing huge increase in non-covid-related deaths of vaccinated people, this data suggests that vaccination makes it many, many times less likely that you are going to die of covid as compared to you dying to other things.

Unless it's happening within that 21-day window, which is the claim I've been seeing elsewhere for around a month now.

From Table 1:

In unvaccinated people 37.4% of all deaths was from COVID (38,964/104,134)

In fully vaccinated people 0.8% of all deaths was from COVID (458/57,721)

This is all amazing news! However we still don't know how much worse the vaccines are against the Delta variant, since most deaths during the period were not from the new variant. It can't be too bad though.

As stated in the report, “ the characteristics of the vaccinated and unvaccinated populations are changing over time, which limits the usefulness of comparing counts between the groups.”

It is still useful to be able to answer "What was your chance of dying of covid compared to all other types of deaths based on your vaccination status".

Because if you can't do that you can compare data even with the status and vaccination rates changing over time.

So even if we lump together all different cases at different times saying that person in this class of people had less than 1% of chance of dying to covid vs. another class where you had 30% of chance of dying to covid is a clear indication that one class was much more resistant to dying than other.

About half the (total, not adult) population was vaccinated by end of May: https://www.bbc.com/news/uk-57313399

Yes, what is the the overall mortality rate for vaccinated vs. unvaccinated. Ballpark estimates looking at the vaccination rate over time, it seems like vaccinated people had almost double the mortality rate of unvaccinated.

https://coronavirus.data.gov.uk/details/vaccinations

Vaccinated are less likely to die of Covid-19, but more likely to die? Not very reassuring.

Depends on the demographics of who gets vaccinated and who doesn't. If the elderly get vaccinated at a higher rate, it seems normal that vaccinated people would have a higher non-covid mortality rate.

It is also not surprising if healthy people are less likely to get vaccinated.

Wouldn’t the original study depend on demographics as well? If the young get vaccinated at a higher rate, it seems normal that unvaccinated people would have a higher covid mortality rate.

Well, if you think about who got vaccinated first, it should be pretty obvious why that is the case, don't you think?

Aye. Stratification required here.

I don’t see any mortality data for vaccinated vs. unvaccinated in your link?

UK had strict age-banded rollout, starting with the eldest. By the end of this study, I had just had my first dose (I'm 27, and I booked it as soon as possible). It's not at all surprising that older populations were more likely to die.

Reply on news.ycombinator.com