Thread by @BillHanage

I was wondering earlier today if change was afoot in the UK due to B.1.617.2, but @jburnmurdoch's thread is better than anything I could put together. It deserves your time. I do think it misses something out however. Hospitalizations are climbing much earlier than I'd expect 1/n

https://twitter.com/jburnmurdoch/status/1397995388267810818

First it certainly looks like B.1.617.2 has become readily established, and is sweeping in multiple different places suggesting that it is more transmissible even than B.1.1.7, itself no slouch. IMO we can't put this down to founder effects/networks 2/n

(I still don't understand why given what *looks* like high transmissibility it has an MRCA so far back, but it's clearly capable of causing outbreaks like these. That's enough to take it seriously) 3/n

One thing @jburnmurdoch rightly emphasizes is that vaccination will make this round with the virus unlike prior ones. Huge numbers of people are protected already. But importantly not everyone, with younger age groups still to receive shots 4/n

That's why I am surprised, frankly, to see hospitalizations ticking up already. Previously we saw cases climb in younger age groups, often for a while, before the virus got into older networks and hospitalizations ticked up. At the moment they're close to lockstep 5/n

Numbers are small and the situation may change. But I am still surprised. I cautioned less than two weeks ago that it might take a while for the situation to become clear. While it's still not perfectly so, it's much clearer 6/n

https://twitter.com/BillHanage/status/1393963350854356997?s=20

Now the UK has had a pretty fine vaccine rollout. The great decline in cases since the winter has been chalked up to that. But that is wrong. Vaccines have protected many older age groups from more severe disease, but the case numbers have been coming down due to NPIs 7/n

This is why, given recent relaxation of NPIs, an uptick was expected. The UK Health Secretary is quite correct here. However the part later on when he says that vaccines 'sever the link between cases and hospitalisations and deaths' is exaggerated 8/n bbc.com/news/uk-572752…

This is a modeling paper, based on the properties of B.1.1.7, which suggests the consequences in the UK of removing all NPIs even with higher levels of vaccination than this could be quite grim 9/n thelancet.com/journals/lanin…

Intuition for how that might be the case is that vaccine uptake is imperfect. As we've surely learned by now truly uncontrolled outbreaks can infect lots of people quickly. And the *vast majority will be fine* but a small fraction of a very large number can be a large number 10/n

If you'd asked me what I expected a few weeks ago, I'd have said a mild outbreak in which seasonality and vaccination combined to limit the impact of relaxing NPIs. Now I am not so sure. I want to be clear that I *don't* think we are back in Dec 2020 11/n

And yes, it looks like vaccine efficacy is maintained. So getting shots into younger arms will help 12/n

https://twitter.com/JamesWard73/status/1398012064942235648?s=20

I'm not going to speculate on what might explain the increases. Surely some combination of transmissibility, active case finding, shifting contact networks and the changing nature of disease in different age groups but one thing is clear 13/n

It's not over. The future *is* brighter than the past in places with vaccination (too few), but we need to remain aware and be ready with other interventions, to prevent a bump in the road becoming a serious sting in the tail 14/end