Base! How low can you go?

I'm all about that base

Now it's clear to even the worst doomers that XBB.1.5 didn't cause a huge wave, and that (from its growth rate) XBB.1.16 is unlikely to either, the new line is that it's the rising baseline we need to worry about. The theory goes that repeated COVID infections might mean that our immune systems might be collapsing (or so we are told), and that being able to draw sloping upward straight lines connecting the troughs on graphs like this should be worrying us.

However that largely depends on what definitions you use. For example in England, the total COVID beds occupied have indeed not gone particularly low lately (even if nothing like as high as the two big Wave 1 and 2 peaks), and might appear to be running at something like the autumn 2021 level.

But to a large extent that is a function of the big changes that occurred with omicron. The fact that a higher level of (generally milder) cases were circulating in the community meant that at that time we moved from just under 80% of COVID beds occupied being "for COVID" to under a half. This has continued to drift down a little bit, and now only about a third of beds occupied are actually “for”. (This is taking "for" data from the primary diagnosis supplement).

Put it another way: we've gone from a ratio of 4 “for COVID” patients to every “with”, to a ratio of 0.5, so plotting the total number doesn't capture what it did before.

You can see this effect by adding a blue line for primary diagnosis beds ("for COVID") to compare with the overall red figure ...

... and zooming in on the bottom right corner, you can see that it's pretty hard to draw an upward sloping line across the two years of troughs like the original poster did. It doesn’t feel wildly glass-half-full to also notice that the omicron peaks have been coming down in height pretty consistently since the start of 2022.

In fact IMO the picture is more:

1. a rise to a consistent level of about 4,000 with the arrival of delta (July 2021)

2. biggish spikes up and down for BA.1, 2 and 5 (Jan, April, July 2022)

3. smaller ripples (with falling peaks) for the nine or so months since then

While the June 2022 trough was indeed lower than say the late January 2023 one, I think this is at least in part due to effects of seasonality and behaviour, as well as to the precise timings of the arrival of variants.

In fact, I don't think it's crazily optimistic to hope that the current trajectory will continue down for the next couple of weeks say. This has the potential to take us to the lowest level of primary beds seen for two years, given summer weather, not terrifying variants on the horizon, incoming spring boosters for the most vulnerable and so on.

Of course we did go lower at times before that, in Summer 2020 and 2021, but it's worth remembering that each time it took months of lockdowns and school closures to get there, compared to the levels now which are essentially due to immunity alone.