BBCNews 18:02 on 20211223: someone infected with omicron rather than delta is up to 45% less likely to attend A&E and up to and up to 70% less likely to be admitted to hospital.
1. If 100 people of the many who tested positive for the delta variant in August 2021 turn up at hospital then, if the prevalence of omicron is the same in December as delta was in August, how many (on average) do you think ought to turn up at hospital?
2. If, say 40 of these 100 were admitted to hospital for covid-related illness in August, how many would you expect to be admitted in December?
In the finer detail that is called 'more' but is usually just repetition, the up to 45% became 30 to 45% and the up to 70% became 50-70%.
3. Repeat Q1 and give a range.
4. Repeat Q2 and give a range.
This is a matter of relative risk and the figures are very much averages. On 15th December 2021 it is thought that there are something like 100,000 omicron cases in the population and the case count is doubling every two days, perhaps between 1.8 and 2.4 days.
5. If nothing occurs to change behaviour, how many days before there are a million omicron cases in the population?
6. There is proposed to be no state intervention —additional regualtion—over Christmas and people are being left to make there own decisions. If everyone is assumed to continue with their same behaviour and if no regulation change occurs until the 4th of January, what might the omicron case count by then? What is wrong with this assumption?
There is a noticeable delay from knowing you have contracted omicron to perhaps being ill enough to attend hospital and the figures collected show that this is 10-12 days. So a surge in case count has a matching surge in hospitalisations around a week and a half later. Generally we hear about the new case count (new daily cases) and for delta an weekly average of 50,000 cases is matched by about 7000 in hospital. When we consider delta, we are given a lower relative risk of hospitalisation. Let us suppose that omicron growth stabilises at 100,000 new cases per day.
7. What does this suggest to you as a hospitalisation figure?
Comment: What an awful way of presenting information; the 100%, that essential for understanding, is quite well buried. Having said this was awful, I cannot think of a better way to put it without requiring a lot more detail, like an absolute risk, measured or perceived. I suspect that bald numbers would be better.
So suppose there were 100 people with delta who attend A&E, of whom 10X are admitted to hospital. Replace these delta cases with omicron it would seem that there are 55 that attend A&E and 3X that are admitted to hospital. So if omicron figures are a lot higher, let's say twice as big as an equivalent delta sample, then 110 are at A&E and 6X will be admitted. If omicron is three time as big, then 165 attend A&E and 9X are going to be in hospital.
Let's go back to that 50-70%, then. If the issue is to do with swamping the NHS and if we were very close to breaking the NHS at 21,000 in hospital (adjust the number to suit, but we learned some limiting numbers in early 2020) and if, on average, people occupy a bed for three weeks, then our limit per day is 1000 new hospitalisations. So in turn that means that the limiting prevalence of omicron (case count) is between 40% and 100% bigger than for the delta waves. 42% is 1/70% but these are crude numbers so I've rounded it off. So, while we watch the case figures go sky high, we know that in terms of 'protecting the NHS' the numbers we need to watch are for hospitalisation rates, which vary tremendously with the age group that is being infected. We also observe that something like ⅔ of the acute (covid) beds are taken by people who are not vaccinated (not even a little bit, and they commonly think that now is a suitable moment to be jabbed, when they are so very wrong; they have to survive first).
Another figure to watch is the staffing levels at NHS, which hits big problems when absence exceeds 20,000, needing external support (e.g., drafted soldiers, firemen, etc). Today's figure is just under 19,000.
A1. One would like to think 45% less, 45 fewer people, 55 people attend.
A2. 70% less likely, so 28 fewer, leaving 12 admitted.
A3. From 30 to 45 people fewer, so 55 to 70 people.
A4. 50-70% of 40 is 20-28 people fewer, so 12-20 people would be admitted.
A5. We only need to go up by a factor of ten, so more than six days but less than eight. Precisely 2ˣ=10, so x=log10/log2 or do decimal search, x = 3.3 doublings, 6.6 days. Using the 1.8 to 2.4 range , this works out as 2.6≤x≤ 3.9 or between 5 and 8 days.
A6. 15th Dec to 4th Jan is twenty days, or about ten doublings (8.3 to 11.1), so 100,000 would become 102 million (31.5 to 219), which is bigger than the national population. So omicron would much earlier run out of candidates to infect.
A7. Much approximation here, but this is twice the delta rate and 50-70% fewer will be hospitalised, i.e. 2x(30-50%) = 60-100% will be hospitalised, so we would expect (hope) that 4,200-7000 cases are admitted to hospital.
