#15,726
Nearly two weeks ago Denmark's Statens Serum Institute (SSI) estimated the B.1.1.7 variant would likely become dominant in their country by mid-February, even though as of their last report (Jan 17th) they had confirmed only 283 cases.
Only a fraction of positive tests are currently sequenced, and so the numbers must be extrapolated. That said, the percentage of B.1.1.7 cases continues to increase each week.
With the famous caveat that `All models are wrong. But some models are useful', we've a new analysis from Denmark's SSI that looks at various scenarios (Rt, Social Distancing, vaccine rollout) and attempts to model the number of COVID (and covid variants) cases by early April.
The focus is on B.1.1.7, but it should be noted that on Sunday, the SSI Reported their 1st Detection Of 501Y.V2 (South African) COVID Variant. How much of an impact this, and other variants may have on these models isn't known.
Additionally, these models utilize a lot of assumptions which may - or may not - hold true. The range of outcomes in these models is therefore pretty broad, with a 10-fold difference in the number of variant cases by April between the best-case and worst-case scenarios.
First the (translated) press release, followed by (translated) excerpts from the 17-page report.
Forecasts for infection rates with a focus on the development in cluster B.1.1.7
The expert group for mathematical modeling of covid-19 has published a new report showing a number of forecasts for the expected development in the infection rate in the coming months.
Last edited January 19, 2021
The forecasts are to be used as a data basis for the future management of the epidemic.
The prognoses are simulated in a so-called population model (A SEIR model, where individuals in the population are assumed to be either susceptible to infection, exposed, infected or immune). The model has been adapted to take into account the current level of activity in society during the current closure. The model also takes into account the more contagious virus variant cluster B.1.1.7 as well as the roll-out of the vaccination program.
The forecasts show how crucial it is to slow down, especially the more infectious variants. For Denmark, it will especially be the English variant, B.1.1.7.
"The forecasts for the number of infections at the beginning of April are strongly dependent on the relative contact number for cluster B.1.1.7, which describes how much higher the contact number is for cluster B.1.1.7, compared with the other circulating virus variants," says Camilla Holten Møller , who is a doctor, Ph.D. and leader of the expert group.
The forecasts have been developed on the assumption that the level of activity in society will remain unchanged throughout the period up to April, corresponding to the current restrictions and measures remaining unchanged, and that the vaccination program will be rolled out as currently described by the National Board of Health.
“In order to get closer to the future scenario, it will be crucial to determine the contact number for cluster B.1.1.7 with greater certainty. We expect to be able to do this when the data basis for cluster B.1.1.7 is continuously improved in the coming period, where affiliation to cluster B.1.1.7 is examined for all positive samples, ”says Camilla Holten Møller.
From week 3, all positive samples are expected to be examined to determine if the cluster B.1.1.7 virus variant is present. Link to the expert report " Forecasts for infection rates with a focus on the development in cluster B.1.1.7 ".
Conclusions:
The relative contact number for cluster B.1.1.7 in relation to the contact number for the other virus variants is estimated to be 1.36 (95% CI [1.19; 1.53]), which can be converted to the current contact number for cluster B.1.1.7 is 1.16 (95% CI [1.01; 1.30]). Even with the current significant restrictions and measures are the contact number for B.1.1.7 thus remains above 1.
The prognosis for the infection number in early April is strongly dependent on the relative contact number for clusters B.1.1.7. The two scenarios below are based on a contact number of 0.85, an assumption that the level of activity in society remains unchanged throughout the period, similar to the current ones restrictions and measures are maintained unchanged, and the vaccination program is rolled out as current described by the National Board of Health. It is clear that the forecasts are significantly dependent on it relative contact numbers for cluster B.1.1.7.
- If the relative contact number for B.1.1.7 is 1.36, corresponding to currently found in Denmark, infection rates of approx. 400 cases per day in early April.
- If the relative contact number for cluster B.1.1.7 is 1.53, the scenario will be decisively different with an increase in the number of infections to about 4,000 new cases daily in early April.
Simulations of the development in cluster B.1.1.7 show that cluster B.1.1.7 is expected to account for 50% of the circulating virus variants around mid-February. If the relative contact number is at the low end of the confidence interval, however, it is not expected until the end of March.
The report's forecasts are subject to major uncertainties due to a relatively small data base for cluster B.1.1.7, as well as stochastic fluctuations in the spread of cluster B.1.1.7, which also means that the width of the presented confidence intervals may be too narrow. It is thus absolutely crucial that determine the relative contact number with greater certainty, in order to reduce the uncertainty in the forecasts.
The scenario where the relative contact number for cluster B.1.1.7 is 1.36 must be seen in the light of the fact that it is calculated assuming an unchanged level of activity in society, corresponding to the retention of those in force restrictions until April.
In summary, it can be said that B.1.1.7 even with the current restrictions has a contact number above 1.0. The future course of the infection rate is strongly dependent on how high the relative contact number for B.1.1.7 develops, with the current vaccine roll-out and activity level in society.
In order to get closer to the future scenario, it will be crucial to determine the contact number for cluster B.1.1.7 with greater certainty. The contact number is thus estimated again when the data basis for cluster B.1.1.7 continuously improves in the coming period, where affiliation to cluster B.1.1.7 examined for all positive samples.
While hard data on the transmission of B.1.1.7 is still sketchy, the anecdotal data from the UK strongly suggests that B1.1.7 may have what it takes to achieve global dominance in 2021. On Friday, the CDC's MMWR: Emergence Of SARS-CoV-2 B.1.1.7 Lineage — United States, Dec 29, 2020–Jan 12, 2021, issued a similar warning, stating:
The modeled trajectory of this variant in the U.S. exhibits rapid growth in early 2021, becoming the predominant variant in March.
Which assumes, of course, that another - even more transmissible variant - doesn't come along in the meantime.
