Thread by @chrischirp

THREAD on exponential growth, Omicron & what we might see in England over next few weeks.

In Gauteng province SA, cases (now almost all Omicron) have been growing at 25% a day for last three weeks.

Omicron rapidly increasing here, at roughly similar rate (so far). 1/12

Boosters might well slow it down, and vax + booster should reduce severe illness, but even a small percentage of large number can be a very large number.

And no reason to think Omicron can't sustain 25% daily growth here for at least a few weeks (*in absence of measures*). 2/12

Meanwhile, Delta is actually increasing slowly at the moment in England - Omicron might not replace it straight away, so much as co-exist and overtake it.

Sajid Javid said 276 sequenced cases of Omicron in England to date - all would likely have been infected pre 1 Dec. 3/12

There will be more than the sequenced cases. I've assumed 200 new daily cases of Omicron on 1 Dec & 25% daily growth.

For Delta, I've been v conservative and assumed 0% growth.

So where does that get you? 4/12

Well, you don't really notice Omicron for a few weeks (remember a new case will take a week or so to show up in a test result and a bit longer to show up in sequencing!).
But then it rapidly catches up Delta until at Xmas they are roughly equal and we're at ~90K daily cases 5/12

And the nature of exponential growth is that things get rapidly much worse - Omicron cases could continue to double every 3-4 days leading to well over 100K cases a day v fast.

Even if vaccines blunt much of the impact the sheer numbers could cause major problems. 6/12

This all assumes no changes in behaviour or measures - if people work from home, socialise less, wear better masks more often... all of it would slow it down and at this stage this makes a big difference. 7/12

E.g. slowing growth to 15% a day gets this.

Omicron would still eventally take over but would take another 2 weeks to match Delta - in which time we could have boosted another 5 million people and added protective measures. 8/12

The more we can slow it down, the better prospects we have as we learn more about it & vaccinate more and more people (inc hopefully children).

In these early stages, it can make a really big difference! 9/12

This is a TOY projection based on daily growth rates. Reality is far more complicated, with differential immunity at different ages to Delta & Omicron, different behaviours, school holidays approaching etc etc.

So take this mainly as an illustration of exp growth BUT 10/12

Over last 5 mnths, we've got used to things not doubling quickly. We're now back in 3-4 day doubling territory. That means decisions *cannot* wait for lots more evidence.

If Omicron as bad as feared, don't want to be at thousands of daily cases when that evidence comes. 11/12

The lesson of the pandemic has surely been that early action is better than late action.

I worry that we've *still* not learned this lesson. We should be acting to slow growth *now* as they are already doing in e.g. Ireland & Norway. 12/12

PS for those who like log scales, here is the main plot on a log scale.