Sorry, my bad, I'll correct my post. (I somehow used the number in hospital and not the daily admissions. The remainder of the post will still stand).
The comparison I'll make, since it's like-for-like, is to compare a peak in confirmed cases to the corresponding peak in hospitalisations, using the 7-day averages from the
dashboard.
Pre-vaccine: admissions 28th Dec, 302, cases Dec 23rd, about 10,000, gives 3%.
Post-vaccine: admissions Jul 27th, 83, cases Jul 19th, about 9,000, gives 0.9%.
So that in itself is a clear difference, but it's not the whole story. Even if vaccines only gave the 3x protection from hospitalisation that one might naively assume from those %, we'd expect to see 3x more unvaccinated than vaccinated going into hospital. From various reports, we know it's far higher than that, but I'll leave that point to one side. Instead, we can take the naive route and say that of those 83 statistically at most about 21 (a quarter) were vaccinated. Therefore remaining unvaccinated 62 could likewise be reduced to about 15, if they were vaccinated, thus more than halving the overall number being admitted - which makes that 0.9% actually 0.4%. But as said, reports from various countries suggest it's far better than that, with the vast majority of admissions being unvaccinated.
The number of deaths appears to have been reduced by an even greater amount, so much that it may be virtually impossible to identify the peak from the latest wave (currently averaging 4 per day). However, it's too early to say. In Jan, it was about 100 a day, roughly 1% of confirmed cases, and plain to see.
The other elephant in the room, which I'm sure you'd note if I didn't, is that 'confirmed cases' isn't accurate; it's way below the real number, and that affects the % calculated above. There are limits on how far out it can be though, given the cross-checking we have with antibody surveys and randomised testing. Every time I've checked on this with the UK data (since adequate testing was in place) it's been less than double. Comparing the RIVM June 2020 figure from antibody testing of 1.5% to the 3% I got above is about double, which is plausible if testing in NL is less extensive than in the UK. (Of course, since it's a survey, the estimate of how many had had COVID has its own limits on accuracy).
The RIVM study was most useful back when there wasn't sufficient testing. Now testing is in place and giving consistent results, and we have two peaks of cases of similar size that can be compared with the only major difference being the vaccine, that's our best data. The amount of daily testing varies a bit over time (far more in NL than UK), but is actually at about the same level during the peaks I used (if anything it's less now than in Dec/Jan, meaning more cases are missed, and current admissions are an even smaller %).
