A few weeks ago, as the government was preparing for a great reopening on 21 June, I wrote a cover story for The Spectator with some bad news: a third wave was coming, I argued, and it could be even bigger than the second.
It jarred with the mood. Covid cases were falling and a great many people desperately wanted this to be the end of it. My model, the Bristol University PCCF (Predictor Corrector Coronavirus Filter), showed otherwise: the biggest wave could be yet to come. But with a vital difference: hospitalisations and deaths would be much lower than at the beginning of the year when we were largely unvaccinated.
All of that has come to pass, with one exception. The third wave, powered by the far more contagious Delta variant, looks like it will peak at a million infections, which is far lower than I had feared. It is both good and (potentially) bad news. My original argument was that Delta is so infectious that by this summer it would reach everyone susceptible, vaccinated or not, but at least there would be no one left to infect in winter. Now that Covid is in retreat, does it have enough strength for a fourth winter wave?
Neil Ferguson, the Imperial College scientist behind the lockdown strategy, says he is ‘positive’ the pandemic will be mainly behind us by ‘September-October time’. But in contradiction to Prof Ferguson’s newfound optimism, these months have now been pinpointed by Sage’s modelling subcommittee as ‘a particularly risky point in the trajectory of the epidemic’. They identify the threat of a spate of flu and other respiratory illnesses caused by a reduction in our immunity against them after we were repeatedly locked down. They might well be right.
Will there be a Covid-19 wave in December? Based on my experience tracking the course of England’s epidemic, I would say this is extremely unlikely. Daily new cases are like the flow of water into a tank, where, to continue the analogy, the outflow corresponds to the number of people who recover each day. The inflow may drop, but as long as it exceeds the outflow, the tank level, which now represents active infections, will rise. This has been happening with Covid for the past ten days, but very soon the level of active infections will fall too. This may occur within the next week or so. It may have happened already.
No matter what happens with policy on reopening, masks or vaccine passports for nightclubs, it seems likely that Covid patients will never again occupy more than 5 per cent of NHS hospital beds. There is no longer a scenario — from Sage or anyone else — suggesting that hospitals will be overwhelmed. We may never again see triple-digit Covid deaths. This ought to inform government policy. (What is the scenario they are trying to avert?) The situation is better than I bargained for at the beginning of June and also better than my estimates a month later.
What has done it? Has Covid been forced back because everyone is at home, having been pinged? It’s true that the 19 July reopening has been muted; people are still cautious, as might be expected with a ‘pingdemic’ in full swing. But this does not explain why Covid, which started surging in the last lockdown, should be falling now. My model points to an intriguing possibility: that the decline in active infections can only mean that England is about to reach the herd immunity threshold for the Delta variant. This means that around 86 per cent of England’s adults and children must now be immune.
This would not mean a sudden halt. It is inevitable that more people will get infected as the epidemic winds down — this is what causes the overshoot on herd immunity — so population immunity will rise higher still, likely reaching over 90 per cent eventually. It should provide a pretty good barrier against a fourth wave — the ‘wall of vaccinated people’ that Chris Whitty described. To this, we can add a wall of people immune thanks to recovery from infection: my model suggests 27 per cent. The categories will largely overlap, due to the UK strategy of vaccinating the already recovered.
In some ways, my PCCF model was too pessimistic. It estimated that at most 75 per cent of England’s population would have immunity to Covid-19 by now. So I was pleasantly surprised by the early downturn in new cases. I imagine the Sage modellers were positively shocked, as they have tended to go for pessimism rather than a best--estimate interpretation of the vaccination data. They produced nine scenarios for hospital beds occupied after the 19 July re-opening. The actual figure came in lower than all of them.
The size of the discrepancy looks too big to be explained away easily, and I wonder if there is more immunity than has been allowed for among children. The ONS survey shows that 92 per cent of adults have antibodies, but children are not sampled. While it is known that very few children get seriously ill with Covid and many experience few if any symptoms, might there also be a substantial fraction who do not transmit either?
The extra immunity that the nation so clearly possesses is a great boon, whatever its cause. Our immunity is robust against not only the Alpha and Delta variants, but also the once-feared Beta variant, with recent data from Canada showing 82 per cent protection against hospitalisation and death from a single AstraZeneca jab. Covid may not disappear entirely from our lives, but we can now resume life secure in the knowledge that the effects of the third wave are much less than before, and confident we will not need to face a fourth.