I’m just a guy who helps his kids with their Algebra lessons. And I always tell them the same thing that my teachers told me (and it drove me crazy when I was their age). Show your work.
And I can’t see a reason why that wouldn’t apply here. There have been a few questions that make me feel like showing a little of my work would help.
First - how am I calculation populations? With this from the National flu and COVID-19 surveillance report
They have a data file (which is a massive Excel spreadsheet that contains this monster table (it goes many many columns to the right). It has vaccination information by age and week. But it also has population data - “People in NIMS Cohort”. It’s not all broken up into the same age brackets as the Vaccine Surveillance reports, but I think most people can imagine a little Python and I can combine the columns I need combined so that these age brackets look like what I need.
To be sure this is what they are using for their reports, I tested my rebuilt table against their case numbers to generate case rates for the vaccinated and unvaccinated which matched for the most recent week (prior reports actually had slightly different population numbers in the NIMS cohort, but since this sheet updated all of them, I decided to go with this number and use it to recalculate rates. In this step I also confirmed that Unvaccinated is simply the NIMS Cohort size - Number Vaccinated (at least 1 dose). This is the most inclusive measure of vaccination and it includes all the categories in the Surveillance Reports. This Unvaccinated population I then checked against case rates and they matched, so I have high confidence the population sizes match the ones that PHE and now UK HSA have been using.
So then what? Well I dug into the Flu and COVID-19 Weekly report for the Roche S Age Corrected Seroprevalence numbers and hand entered them into a spreadsheet going back to week 30. With those percentages, I can calculate the total number in each age group with S protein immunity. I then subtract the same Number Vaccinated (at least 1 dose) from this number. I’m left with those that have Protein S antibodies but aren’t Vaccinated. If they aren’t vaccinated, then they are in the “Unvaccinated” population. So this remainder is our “Prior-Infected” group and the “Unvaccinated“ population minus the “Prior-Infected” population is the “Non-Immune” population.
At this point it’s all theoretical - it could be 5% of the “Unvaccinated” it could be nearly all of it. It turns out - it’s closer to the latter.
Only in the youngest group do we start with even a quarter of the total population being Non-Immune, but that quickly dwindles down to roughly 10% of the population. Fear not the population isn’t shrinking by too many deaths - that’s people getting “the jab”, (as the British insist on calling it even in their most serious communications) and being removed from this Cohort. Notice it is the only group to notably shrink during these 10 weeks. (30-39 does too if you look very closely).
Yes - the vast majority of the “Unvaccinated” population is excess S Protein immunity by blood donation survey (17+ age - which is why I don’t include the Under 18s in any of my charts).
Now it’s been suggested that back tracking N protein immunity into the Vaccinated cohorts would give us an idea how much natural immunity is affecting the case rates on the Vaccinated side - and it’s a good idea, but I didn’t have time to do that today (mostly due to needing to hand jam all those Seroprevalence values). But I can give you a back-of-the-envelope estimate - 3/4 or more of the N antibodies belong in the the Unvaccinated but Prior-Infection group based on their size vs the entire population and the percentage of the population with N antibodies. This would mean that a VERY small portion of the vaccinated population has had COVID-19 and while it will have some effect it will be significantly smaller - possibly an order of magnitude less.
Stay tuned as I show more work tomorrow including example case numbers in a few scenarios.