Over the last few weeks we’ve been bombarded with messages telling us that Omicron cases are doubling every two days. This has been accompanied with graphs showing a near vertical increase in cases, along with dire predictions of the whole of the UK, or perhaps even more, being infected by January.
I’m not sure that this is the case.
I think that what is happening is Omicron variant has been present and growing in the UK since the start of November1, however, because the symptoms of Omicron infection are different from Delta, people hadn’t tested themselves and thus case numbers for Omicron were very low. Then, the public was informed of the new symptoms and those that had been infected for some time suddenly rushed out to get tested. I believe that this is visible in the case data for Omicron:
What I think has occurred is up to around the 5th of December case numbers were very low — people didn’t realise they had a covid infection because the symptoms were ‘wrong’. Then, at that point the media started reporting on Omicron variant. This then encouraged some to get tested even though their symptoms were ‘wrong’. Next, over the weekend of the 11-12th December (note the lower case numbers — this is a weekend effect) the media started reporting heavily on Omicron, its symptoms and how it was going to devastate the country — the consequence of this was for many more people to decide to test themselves for ‘the strange new variant with strange new symptoms’, even though many had experienced symptoms for some time and they were in the recovery part of their infection. I suggest that it is that weekend’s reporting that resulted in the massive increase in cases over the start of last week.
I also note that the ‘cycle threshold’ for Omicron cases has been much higher than for Delta variant — earlier in the month the CT for Omicron was approximately 40 while for Delta it was around 22 (remember, this is an exponential scale — an infected person that tests positive only with 40 PCR thermal cycles will have less than a millionth of the viral particles of a person that tests positive after 22 cycles).
Figure 14 from this document. SGTF refers to suspected Omicron variant, while S+ refers to ‘others’, probably Delta.
I note that the upper limit for PCR cycles for Omicron is currently 30 — this might have an impact on the number of positive cases found…
If this hypothesis is true, then the increase in case numbers for Omicron should slow down over the next week or so, before picking up again as the testing starts to fully reflect current cases.
Note that I’m not suggesting that Omicron is over. Indeed, I think that we’re currently in a Delta wave (probably sub-variants AY.4.2, AY.4.2.1 and/or AY.4.5) and this will lead to the usual problems around the end of the month. My suggestion is that the Omicron wave will take a few more weeks to peak than the hysterical modelling based on a 2 day doubling time suggests. As far as I can tell, removing the effect described above gives a doubling time of approximately 4-5 days (ie, similar to previous variants), and we should hit the Omicron peak in mid to late January. Given that Omicron has achieved near complete vaccine escape and that the temporary protection offered by the boosters will have worn off by January, I expect this wave to be rather brutal.
I think that the above also explains why data suggests that Omicron is more deadly than Delta — as deaths from covid appear approximately three weeks after infection, really we should consider today’s covid deaths in terms of the case numbers three weeks ago… which gives a death rate of approximately 5% based on official numbers. If there were really far more cases three weeks ago then the mortality rate for Omicron is likely to be similar to that of other variants for a mid-winter outbreak (note that South Africa is in mid-summer, so you would expect a lower mortality rate for their Omicron wave. I also suspect that the ‘delayed testing’ hypothesis also explains their very fast rate of increase of Omicron cases — this might mean that their cases will plateau or even rise further from here).
Or possibly before — I’ve got a post coming up on early signals for the start of a new covid wave.
Hi,
I agree somewhat with your analysis of the Omicron data, although I suspect the "super cold" we were treated to in late fall may have been early signs of Omicron.
The question I have, given the emergence of data from Denmark (roughly 60% fewer hospital admissions and indeed a shorter stay for those in hospital) and the (granted, summer) evidence from SA that this is less virulent, what do you think will be causing the brutality of a late January peak?
The high case numbers themselves mean little, and the NHS has coped with all other issues so far - if the staff weren't forced to isolate after contact with Omicron (negating their vaccine status, rendering the whole freedom via vaccine argument useless) we wouldn't see any issue at all in the healthcare system, would we?