delta taking off

With July and the rescheduled re-opening of everything in the UK approaching, let's do some back-of-the envelope calculations again to see what may happen in July and August, seeing that the UK govt. seems to think the vaccination programme is a magical solution to everything, including the exponential spread of the delta variant.

As per today's figures, there have been 32.2 million second doses, meaning that in two weeks time, when those second doses have all done their magic, there will still be 34.6 million people not completely protected. Some will be partially protected by a first dose, but, on the other hand, in some 10% of the doubly-vaccinated people the protection won't work, so let's assume these two imperfections cancel out and we will have 34.6 million unprotected people on Saturday July 10th (which happens to be the day before some football game planned to take place in Wembley Stadium).

Cases are currently increasing at more than 50% per week - earlier this month the figure had peaked at 66% and dipped to 31%, so 50% is a good midpoint between these extremes and easy to calculate. Going up by factor 1.5 each week means factor 2.25 in two weeks, so a doubling time of less than two weeks.

As per today's govt. figures, there were 98,460 new cases in the last 7 days, which gives us 14,065 for our 7-day rolling average of daily cases. Times 2.25 makes 31,650 new cases on the Saturday before the final.

Let's carry on in 2-week steps assuming (generously) 200,000 second doses are given every single day (2.8. million in two weeks):

24. 7. 71,208 daily cases, 31.8 million still unprotected.

7.8. 160,217 daily cases, 29 million still unprotected

21.8. 360,490 daily cases, 26,2 million still unprotected

4.9. 811.100 daily cases, 23.4 million still unprotected

18.9. ? 1.82 million daily cases, 20.6 million still unprotected

OK, so that last step in September is the one where I'm reasonably optimistic that it won't happen - with more than 2/3 of the population protected, the virus won't find another 1.8 million susceptible people it can infect. Especially because millions will already have had it. But before that, I wouldn't bet against the delta variant or expect any miracles from the vaccines.

Hundreds of thousands of cases in late August appear possible, and even with the mortality reduced to 1/1000 thanks to the vaccine protecting most of the previously vulnerable groups, this translates to hundreds of daily deaths. As we had in January, and last year in the first wave.

I made an assumption that may prove to be too optimistic, namely that the vaccine remains as effective as it is now. Letting the virus run free while conducting mass vaccinations against it is essentially an experiment in virus evolution. Given hundreds of thousands of opportunities to mutate, it may very well find a way around the vaccine. And then we will be back to square one. (It doesn't have to become completely resistant, even if it reduces vaccine protection from 90% to 45% we're facing a major disaster.)

To me the lesson is you shouldn't try to race an exponential function, unless you have exponentially growing power yourself, which would be unlikely. You need to catch and confine it while the case numbers are manageable, which the govt has failed to do for the third time in 18 months. As I was writing this post, the health secretary has handed in his resignation, but not for dismally failing to protect the health of the population.

first checkpoint 10.7.: I predicted 31,650 new cases for today. The 7-day rolling average falls a bit short with 30504, but the daily figure is above the prediction, with 32367. Second doses haven't come close to the 200k per day rate I assumed, so protection is generally less than anticipated.