The other thing that doesn't show the whole story is using a data set starting in March when there was a quite low percent vaccinated but comparing totals of unvaxxed vs. vaxxed deaths over the whole period, when fully vaxxed in March was only a few percent.

It'd be much more interesting to see a straight comparison over Oct-Nov, say, when the vaccination rate was stable and get the unvaxxed vs 1 shot vs 2 shot vs 2 weeks after 2nd shot vs booster numbers.

This is exactly simpsons paradox. It is a shame so few people have heard of it!

This same breakdown implies that vaccinating half of the currently unvaxxed would do a lot to reduce pressure on hospitals. That is, if you have 80% vaxxed, a mere 10% increase could halve the number of hospital beds needed. No wonder governments are going so hard on vaccinating the hesitant and antivax.

No, this does not imply that going from 80% to 90% vaccinated could halve the number of hospital beds needed - that reasoning would only work if the vaccine completely prevented Covid-related hospitalizations, in which case zero vaccinated people would be hospitalized and Simpson's paradox wouldn't apply. What it actually implies is that at much less than 100% vaccination, you hit a point where vaccinating the remaining population does very little to reduce pressure on hospitals because most people who're hospitalized have been vaccinated, despite the fact that the vaccines do reduce the risk of hospitalization. That is, the benefits of vaccinating the last few percent of the population actually diminish rather than increasing like they would do with a more effective vaccine.

> most people who're hospitalized have been vaccinated

Is this true? My understanding is that the opposite is true, most people who are hospitalized are not vaccinated by a large margin.

I haven't checked, but I think your understanding is correct. Even if it's not, here's a good explanation https://ourworldindata.org/covid-deaths-by-vaccination

>This same breakdown implies that vaccinating half of the currently unvaxxed would do a lot to reduce pressure on hospitals.

This ignores the demographic differences between the vaxxed and the unvaxxed. The elderly and those with conditions that made them high risk have a much higher vaccination rate as one would expect.

If you cut the number of vaccinated in half you won't cut hospitalizations in half because subgroups of the population being hospitalized are not present in the same ratios in the unvaxed.

> What is the breakdown of unvaxxed/vaxxed in hospital admissions and deaths?

85.6% of hospitalized in Denmark are either unvaccinated or "half vaccinated". See https://experience.arcgis.com/experience/aa41b29149f24e20a40... graph 2.

This means that if the vaccination rate grew from 90% to 100% the hospitalized number would drop to very manageable levels.

Even with a 90% fully vaccinated population, the remaining 10% are still enough to fill up the hospitals.

Edit: I think I misunderstood the graph. It is actually already normalized by the size of each population (Unvaccinated, Half-vaccinated, Fully vaccinated).

If you look at the raw numbers nearly twice as many of the hospitalized are vaccinated.

When taking the vaccination rate into account, the likelihood of hospitalization still seems to be a fifth after vaccination.

But even if the remaining population got vaccinated it wouldn't be enough to drastically reduce the hospital load.

> If you look at the raw numbers nearly twice as many of the hospitalized are vaccinated.

This doesn't mean what you think it means. Imagine if 100% of the population were vaccinated, what percentage of the hospitalized would then be vaccinated?

https://ourworldindata.org/covid-deaths-by-vaccination

the point is the beds are being taken up by the vaccinated. the idea was after vaccination these beds wont have to be taken up by the vaccinated suffering COVID and will be vacant for others suffering other diseases.

> We have more infected, the same number of admissions in hospital and a 15% higher death rate than all of 2020

Another really, really important piece of context is how human behavior has changed. In 2020, many people were very cautious about covid. For example, in liberal US cities like Seattle and Portland, many indoor activities were restricted pretty hard until 2021. People didn’t dine out in person, or didn’t attend the movies. In 2021, people have more or less returned to normal, less-cautious behavior. While masks are still worn, people are comfortable dining out and going to the movies and all sorts of activities.

My point is that this change in behavior is likely to “blame” for the increased cases. If people were behaving the same way a year ago as they are today, cases would have been significantly worse then.

In other words, vaccinations have allowed us to return to a mostly normal life while stopping cases from rapidly increasing. At the very least, anecdotally, in places with lower vaccination rates, a return to normal life has been accompanied by more covid scares.

> People didn’t dine out in person

I wonder how did they dine out :-)

Order out, take home? Sort of eating out. At least you are not cooking. :)

Your numbers are very similar to Ontario's, and in fact Canada in general as a whole.

But what's strange is out the US doesn't seem to reflect this.

There are states with fairly high vaccination rates where vaccinated people are ending up in the ICU at rates many times ours, and are only showing up as 2.5 or 3 times better protected than the unvaccinated for hospitalization.

I don't know how to explain this other than length of time since injection. But we're heading into 6-7 months here now in Ontario and I'm still not seeing this trend. Boosters will certainly help, I'm sure, but I don't feel it can be the sole cause.

It's so ridiculous just driving across the border... look at Vermont or New York compared to Ontario or Quebec in terms of daily case rates and hospitalization. (Or Maine compared to New Brunswick, etc.) The only way I can explain it is by masking, which is still rigidly followed here in most of Canada, but not as much in the US.

PS congrats on having perhaps the most competent governments in North America in terms of managing this disease. It's not the first time I've found myself browsing for real estate in Nova Scotia, but it certainly became more serious this past year...

My guess is that obesity plays a bigger role than many people realize. Perhaps obese people who are vaccinated are still at a substantially elevated risk? This also explains why Africa is doing pretty well in terms of low casualties: Young people and fewer obese.

Most developed countries have populations with high proportion of elderly people, just look at their population pyramids. Their population is generally older on average, as the country has long life expectancy, a low death rate, but also a low birth rate.

In Canada, we had a longer gap between doses, which may be a factor in extending efficacy. There is also a robust booster strategy, though it does not get much attention here. Not sure how it compares, though.

> from what I can tell the 2-shot people who die are generally older and/or infirm. So I believe the 2shot numbers are being thrown off by an atypical population...

People over 50 make up 1/3 of the population and 93% of covid deaths.

The overwhelming majority of people that die from this thing are older and/or infirm. It has nothing to do with an atypical population, that's just how this virus works.

The vaccine can only do so much. We need to stop pretending it's the "slam dunk" our experts advertised it as and get on with our lives.