Thread by @chrischirp

THREAD: I know many people are thinking "but how can this new variant be scary when overall case numbers are flat?"

Here's how.

(Numbers that follow are illustrative but based on case data & sequenced proportion data from Sanger: covid19.sanger.ac.uk/downloads) 1/6

Cases of B117 ("Kent") variant have been falling every week, while cases of B.1.617.2 (1st sequenced in India) have been growing. Total cases are B117+B.1.617.2.

We've had two epidemics at once. During early days for B.1.617.2, its growth is entirely masked by fall in B117 2/6

As proportion of cases that are new variant grows, overall cases start to creep up - but not by much. Not at first.

In this example, it's an 11% increase week ending 22nd May. 3/6

As soon as new variant becomes dominant, cases start increasing much more quickly - *assuming* it keeps growing at the same rate.

According to Saturday's PHE tech report & Sanger data, we've probably just reached the tipping point in England between B.1.617.2 & B.1.1.7. 4/6

The growth in cases is then quite rapid. Who knows if this will happen here - but if it does it should be obvious. And we should see it in next couple of weeks.

B.1.617.2 has showed a few signs of slowing growth past few days - but even so we should still see something. 5/6

At the moment, total cases are almost perfectly in line with this illustration. Cases in England rose 15% in week ending 23rd May.

If we *don't* see increases over next 2 weeks, I think it's fair to be less worried about B.1.617.2. 6/6