Thread by @VictimOfMaths

COVID case rates are still hard to interpret (because holidays), but I think the broad patterns are still worth looking at.

Changes in the past week here show cases rising in all age groups, but fastest in older, more vulnerable age groups.

(the latest day's data is promising, but I have low confidence that it tells us much about whether cases are actually falling or rising, so interpret with caution)

This is what those changes convert to in terms of absolute case rates. A really clear Omicron wedge, from people in their 20s up the age groups.

For line graph fans (or heatmap-o-phobes) here's the same data in line form. Under 14s clearly doing their own thing (at least until the new term appears in the data in a week or so), but everyone else heading upwards.

Moving back to case growth, i.e. the slope of the previous graph, the picture is messy and there is a lot of variation between ages, but there are some clear patterns...

In the under 20s, things are *relatively* stable, although cases are rising slowly.

The new school term is likely to wreak havoc with these lines though, I'd guess.

How worrying you think that is will depend very much on your perspective.

In younger adults case growth has settled down a lot, although cases are still rising, since the initial Omicron explosion in mid-December.

For people in their 40s and 50s, after a sharp initial rise when Omicron first hit, growth has levelled off, but cases are still rising pretty fast.

I'm honestly not sure how to explain/interpret the recent divergence in a group that has been tightly bunched previously.

Things get more worrying in 60-79 year-olds, where the Omicron wave has plateaued at much fast rate of growth, with cases doubling every week.

The picture is similar in the over 80s, but, if anything, slightly worse. Cases more than doubling every week.

Now, of course, this is case growth, not case rates. Case rates themselves are still much lower in older age groups than younger adults. But the gradient risk is steeper still, so this is a big concern for hospital admissions in the coming weeks.

There is also an important geographical element, with cases seemingly falling in younger ages in London, and rising rapidly in older ages in the North West.

https://twitter.com/VictimOfMaths/status/1478671596625801218

I sincerely hope that the latest data isn't a blip and that cases do start heading down across the country soon. But I have low confidence in the recent data suggesting we might be there already. The effect of the holidays makes everything very uncertain.

So I'll just have to sit and wait. And draw graphs.

R code for these graphs is here:
github.com/VictimOfMaths/…