Thread by @BillHanage

Very interesting preprint on epidemiology of B.1.1.7 (or the 'Variant of concern' - VOC) in SE England. Have not fully digested but there are some important take homes 1/n cmmid.github.io/topics/covid19…

Some have argued B.1.1.7's sudden rise is a 'founder effect' reflecting the VOC getting into a more densely connected network of hosts, and a number of superspreading events. This seems unlikely to me because there were lots of more common lineages available to be amplified 2/n

the preprint agrees. First, here is the increase in the proportion of cases due to the VOC in the SE (pink) compared with the rest of England (these are detected btw through characteristics of the test result). The gray box is the 2nd national 'lockdown' (👈🏼hate that word) 3/n

Here's mobility data (left) and mean contacts over time and by age (right). The SE in tier 4 are in pink. More contacts were not being made in those regions during November. There are other ways of measuring contacts, but this suggests a property of the virus, not the network 4/n

So what is going on? More transmission is obvs bad news but what's making it happen? The preprint considers higher infectiousness per contact, immune escape, a change in the susceptibility of children and a shorter generation time 5/n

Time for an #OptimismSpa. While the immune escape model looks like it fits quite well... 6/n

Greater infectiousness fits substantially better. This is important because it hints that vaccines may remain effective. It's also consistent with evidence of higher viral loads (which don't mean more infectious on their own, but fit the pattern) 7/n

This is increased susceptibility among children. Again a poor fit, but the preprint assumes children 0-19 are half as likely to be infected as adults, while most would say that's only true for the under tens (high schoolers are more like adults). So results might reflect that 8/n

Finally generation time. This matters, because you always have to remember that if something is increasing rapidly it's not just how many people get infected but how quickly ncbi.nlm.nih.gov/pmc/articles/P… 9/n

But shorter generation time does not fit well at all 10/n

What does this mean? Well it provides more evidence that this really is more transmissible. Don't get hung up on the exact figures. The take home is that it is harder to control. I think the consequences for children need to be thought about more too 11/n

If kids remain 50% as susceptible as adults, but the infectiousness is that much higher, that means they're more likely to get infected and transmit. With consequences for schools. Especially schools without masks or other mitigation 12/n

Relevant to that, look at this from ons.gov.uk/peoplepopulati… and note the uptick in the youngest age group. This was not seen when schools opened earlier in the year. It took weeks to increase. This urgently needs more work 13/n

The abstract "control measures of a similar stringency to the national lockdown implemented in England in November 2020 are unlikely to reduce the effective reproduction number ​R​t​ to less than 1, unless primary schools, secondary schools, and universities are also closed" 14/n

What to do? Well as often is the case @MackayIM has good advice

https://twitter.com/MackayIM/status/1340868953397084161?s=20

15/n

I might add that if you don't yet have the virus in your community, you want to delay it. If you're not looking for it, you ought to learn from the UK's excellent genomic epidemiology operation and get looking 16/n

One final somewhat optimistic point - there remains no evidence of more severe disease from this. Which is good. But vaccination, already important, is now more important than ever. As the preprint concludes 17/end