COVID-19: February 2021 Update 2

Where are we now?

The graph below shows the retrospective 7-day average of the daily deaths from COVID-19.

The time axis runs from 1st October 2020 (when there were last less than 100 deaths per day) to May 2021 (when we might again anticipate having less than 100 deaths per day).

Click for larger version. See text for details.

The dotted purple line shows an extrapolation of the data with a halving-time of 21 days – the same as occurred after the first lockdown. So this corresponds to an expectation of how the daily rate of death might fall if the vaccine has no effect.

If the vaccine is working, we might expect the rate of daily deaths to fall faster than the dotted purple line. Such an effect might just be becoming visible, but it is not yet a clear signal. I’ll say more about what we might expect below.

Vaccinations

The blue line (which should be read against the right-hand axis) shows the percentage of the UK population who have received the first dose of a vaccine.

The dotted blue lines show extrapolations of (a) all the data and (b) just the data from the last week. If things continue as they are now then we might expect roughly 66% (two thirds) of the entire UK population to have received a first dose of vaccine by the start of May – just 78 days away as I write.

Even though I am very confident that vaccinations will eventually reduce the rate of COVID deaths, it is difficult to anticipate precisely when this effect will show up.

The graph below shows the number of weekly deaths for the first four weeks of 2021 (link for ONS spreadsheet) categorised in 5 year bands by the age of the people who have died. Clearly it is mainly the elderly who are dying.

Click for larger version.

According to a spreadsheet downloadable from this page (link) by the 7th February 2021 more than 90% of people aged over 75 had received a first dose of vaccine.

If we allow a few weeks for the inoculation to take effect, we should expect ‘shortly’ to see deaths amongst this 75+ age group fall dramatically.

With the “eye-of-faith”, one might already imagine that the effect is detectable in the first graph (see detailed graph below), but I don’t think the data is quite strong enough at this point to confirm the protection.

Click for larger version. Close up of the first graph showing daily deaths.

To find out how long you will need to wait for the vaccine use this online calculator.

Admissions

The graph below shows the retrospective 7-day average of the number of daily hospital admissions for COVID-19.

Using just the power of my eyes it looks like daily hospital admissions are falling with a halving-time very similar to the 21 days observed in lockdown#1.

Click for a larger version: See text for details

It is difficult to anticipate the effect of the vaccine on hospital admissions, but if it reduces serious disease, then we might expect to see hospital admissions fall below this extrapolated line.

Cases

The graph below shows the retrospective 7-day average of the number of positive COVID-19 tests per day.

Click for a larger version: See text for details.

It is curious that since approximately day 390 (25th January 2020) the number of positive cases has fallen faster than the expected 21-day halving-time. The halving-time appears to be about 17 days. I cannot explain this.

The Independent SAGE group noted on 5th February that there had been a technical change in the way tests were counted on the 27th January 2021 (marked as blue dotted line in the graph). But this change in slope – which is in any case using data from the previous 7 days – had already changed by that date.

The increased rate of decline is to be welcomed: if it continues then by 1st May the daily rate of new cases would be just 50% of the dotted red-line projected in the figure above.

When can we go back to normal?

So things are getting better and there is the prospect before us that the ‘virus may soon be beaten’.

But opening up too early, before the vaccination campaign is substantially complete, will allow the virus to spread widely again – mainly amongst younger people – and give it many chances to mutate. This would open up the possibility that the winter of 2021-2022 might be not so different from our current winter.

Personally I would set a target date based on the capacity of test, trace and isolate to actually function. Outside of lockdown, this is the only social measure guaranteed to work irrespective of viral mutations.