solarbird: (Default)
[personal profile] solarbird
New paper in The Lancet. Read it.

My takeaway of interest: detection rate is down to 5% of cases worldwide. That's an absolutely dramatic decline. You get slightly higher with semi-aggressive testing, but to get legitimately higher, you really have to work at it. A few places are, most places aren't, but let's put speculation to the side as one should and go with numbers we actually know.

Now let's go back to the path to r0 << 1, remind yourself of that if you like.

1,049,560 12-and-over cases remaining/anticipated in Washington State remains a valid number from its date. I'd aimed at 80% resistance from boosters, that was probably a little optimistic, and 70% or less would've been a little more accurate, but let's not change the number now.

5% of 1,049,560 is 52,478. 10% is 104,956. We're doing a lot of testing, let's say we bump it up real high (and that some UW data that isn't strictly about this but is kind of reflective about this) indicating a 10% detection rate. Much higher detection than typical, but we have some fairly aggressive testing here. For, you know, the US.

Two-week average for cases is 14,085/day. Times eight days (since the route to r0 post) gives us 112,680 detected cases, vs. a detection of 104,956.

That should just about compensate for that overly-optimistic 80%.

Now let's throw in some worse numbers for the under-12 set, which wasn't included in the above at all.

Population data says there are 1,044,827 under-12s in Washington State. No way to know how many have already had COVID before the path to r0, so let's assume none. And let's assume 100% case rate and be really on the high side. 10% is 104,482 detections remaining, without reinfection. Given the testing at schools, that might be low, but... we can't know. We're on the tenuous edge of numbers as it is. But despite that, I'm just going to say it:

If these numbers are right, we should be at peak right now.

If I'm right, we're about to be heading down in case count like the wrong kind of rocket but the absolutely right kind of epidemiology.

And as if on cue, the statewide two-week average looks like it has levelled off, and King County - as you'd expect, given our vaccination + booster rate - is already very well past peak.

If these numbers are right, we're done by March, folks. Not even mid-March. Beginning of March.

We. Are. Done. By. March.

Not to zero. But as an epidemic. Barring, as usual, a new and unpredictable variant that resets resistance. Read the Lancet paper, they talk about that, saying that COVID isn't disappearing, but it's about to become the flu.

Get your shots, be ready for occasional outbreaks, but as a true pandemic, here, in Washington State? Assuming all the assumings?

You do have to keep riding this out, but it's all downhill from here. And everybody else is going to be only a couple of months further along.

Barring a vastly new variant...

...it looks like it actually is gonna be a pretty nice spring in Cascadia.

Date: 2022-01-21 03:02 pm (UTC)
From: [personal profile] talkswithwind
This is great news for me, but I have enough immunocompromised people in my social sphere to see this as the conditional win it is. The future is coming, but it will be contingent to a greater degree on your antibodies working than in the before times. I supposed it will be 'officially' over once we start seeing "no masks" signs on businesses/courthouses in the name of security.

Regretfully disagreeing

Date: 2022-01-21 09:28 pm (UTC)
kathmandu: Close-up of pussywillow catkins. (Default)
From: [personal profile] kathmandu
I'm expecting the new-case graphs to form repeated dramatic zig-zags like that.

These graphs aren't so much showing what the virus does as showing what people do. Since immunity starts wearing off in two or three months, as soon as people start acting reckless again, the new-case numbers will rise again. It can keep rolling through the population, exploding every time it reaches a rich node of unguarded connections, indefinitely.

June 2025

S M T W T F S
1 234 5 67
891011121314
15161718192021
22232425262728
2930     

Most Popular Tags