Read on Twitter
The new release of the randomized ONS infection survey shows the advantage of the variant B.1.1.7 strain has evaporated in regions (London, East, Southeast) where it reached greatest prevalence.
[R_0 for variant cases declined without a decline for other case R_0].
1/
![The new release of the randomized ONS infection survey shows the advantage of the variant B.1.1.7 strain has evaporated in regions (London, East, Southeast) where it reached greatest prevalence.[R_0 for variant cases declined without a decline for other case R_0]. 1/](https://pbs.twimg.com/media/EsVmoE2WMAMdK4q.jpg "The new release of the randomized ONS infection survey shows the advantage of the variant B.1.1.7 strain has evaporated in regions (London, East, Southeast) where it reached greatest prevalence.[R_0 for variant cases declined without a decline for other case R_0]. 1/")
[R_0 for variant cases declined without a decline for other case R_0].
1/
This is a very intriguing finding which seem to argue against simplistic theories about B.1.1.7 being "more transmissible".
If infectiousness was simply higher across the board for B.1.1.7, than when interventions or immunity cause B.1.1.7 cases to plateau or decline, ... 2/
If infectiousness was simply higher across the board for B.1.1.7, than when interventions or immunity cause B.1.1.7 cases to plateau or decline, ... 2/
it should be accompanied by an (even steeper) decline in wild-type cases, which we do not see here.
This does not argue that there is no biological difference or advantage to B.1.1.7, but instead that increased transmissibility/infectiousness would not explain this data. 3/
This does not argue that there is no biological difference or advantage to B.1.1.7, but instead that increased transmissibility/infectiousness would not explain this data. 3/
On the other hand, for example, B.1.1.7 could have a *susceptibility*-mediated advantage. For example, if some fraction of the population is more susceptible to B.1.1.7, this could cause early growth giving the impression of higher transmissibility, but this advantage would...4/
be expected to decrease as more of this especially-susceptible population was depleted through infection.
(It is even possible for this effect to occur without B.1.1.7 having an advantage at all in a population with no previous infections.
) 5/ https://twitter.com/WesPegden/status/1346298493162450944
(It is even possible for this effect to occur without B.1.1.7 having an advantage at all in a population with no previous infections.
) 5/ https://twitter.com/WesPegden/status/1346298493162450944
Of course we only have a few high-B.1.1.7-prevalence regions. It will be important to watch for confirmation of these trends in other places.
The data so far look reassuring but also scientifically intriguing, being incompatible with the mainstream theory of B.1.1.7.
6/
The data so far look reassuring but also scientifically intriguing, being incompatible with the mainstream theory of B.1.1.7.
6/
Addendum: let me highlight some counterpoints raised to the decline in the variant advantage in high prevalence regions in the ONS data.
One is that the same trend has not been observed in the sequencing of cases presenting for nonrandom testing.
7/ https://twitter.com/TWenseleers/status/1352652470795841537
One is that the same trend has not been observed in the sequencing of cases presenting for nonrandom testing.
7/ https://twitter.com/TWenseleers/status/1352652470795841537
Another is a question about whether the ONS is using the best proxies for distinguishing B.1.1.7 from other strains.
In the thread below there is some discussion of this, and the extent to which it could explain the ONS trends.
8/ https://twitter.com/theosanderson/status/1352637938048372736
In the thread below there is some discussion of this, and the extent to which it could explain the ONS trends.
8/ https://twitter.com/theosanderson/status/1352637938048372736
Here's a nice thread on this 2nd possibility with some discussion underneath.
Basic hypothesis is that other probe dropouts could be causing the ONS survey to miss B.1.1.7 strains.
OTOH, these are usually rarer than the dropout used to *detect* B.1.1.7. https://twitter.com/theosanderson/status/1352669515583279105
Basic hypothesis is that other probe dropouts could be causing the ONS survey to miss B.1.1.7 strains.
OTOH, these are usually rarer than the dropout used to *detect* B.1.1.7. https://twitter.com/theosanderson/status/1352669515583279105
