Thread by @jburnmurdoch

NEW with @clivecookson:

Two models from leading epidemiologists put Covid-19 R number below 1 in many parts of UK.

Models from @LSHTM and @MRC_BSU/@PHE_uk put R below 1 in London, South East and East. May be slightly above 1 in Scotland, N. Ire

Story -> ft.com/content/a3145d…

Notes:

1) ⚠️ Uncertainty ranges! ⚠️ We’re deliberately not plotting central estimates or indeed any hard numbers here, because the models produce ranges not precise numbers.

2) Why do the two models look different?

LSHTM (blue) based mainly on cases data, hence steady evolution.

Cambridge/PHE (pink) includes data on mobility levels & lockdown, giving a more sudden step-change.

These are methodological differences. Neither necessarily more "true".

3) Regional differences.

Some reports today that UK R number could be creeping back up towards 1. As you can see, the concept of a UK-wide R number is misleading.

Like outbreaks themselves, the reproduction number also varies from place to place.

4) Things can change quickly. R may be ticking back up due to change in lockdown guidelines this week.

This is very possible, and it’s both wrong and dangerous to assume R crossing below 1 is the end. What goes down very much can go up.