How bad will the third wave be?

‘We see no reason to go beyond 19 July,’ the newly appointed Health Secretary Sajid Javid confidently declared on Monday. His comments follow those of the Prime Minister who has described 19 July as the ‘terminus date’ for lockdown restrictions. But has Javid grasped the realities of the situation? The Indian (Delta) variant rise poses a whole new set of questions. Will it be blocked by the 'wall of vaccinated people' that Chris Whitty described as our shield? Has enough been done to keep hospitalisations under control? And will we see a rise in the UK’s death toll, which stands at 128,000?

Important new data has arrived since I last wrote for The Spectator on 12 June outlining the predictions of my model, the Predictor Corrector Coronavirus Filter (PCCF). We now know that two vaccine doses enormously decrease the risk of serious disease and hospitalisation: Pfizer by 96 per cent, AstraZeneca by 92. But we have also discovered that the Delta variant is more infectious than I first thought. Public Health England thinks it is 64 per cent more transmissible than the Alpha strain. I originally thought it was 50 per cent higher.

So how does this affect the PCCF? In short, none of the new data substantially changes anything. My fundamental conclusions are the same as they were a few weeks ago: we are in for a summer of very large case numbers, but with far fewer deaths than in the previous waves.

The third wave has been with us for several weeks already; it will continue because people have been mixing a lot more since restrictions eased on 17 May. Someone with the Delta variant would pass it on to five others in a fully susceptible population, and this figure is set to rise by 50 per cent on 19 July. The number of cases is, therefore, likely to increase very rapidly. While we should remember that most cases are not severe, with about a third suffering no symptoms, this rapid growth will nevertheless leave us with more infections than we had in January. But, crucially, because the vaccine programme has ensured the elderly are well protected, deaths will not follow.

In January, all age groups were getting Covid equally. Now, most cases are amongst under-25s. Only one in 10 cases occurs in over-65s. The crucial fact is that Covid is ‘like a bad cold’ for most young people, as Professor Tim Spector, the director of the Zoe Covid Symptom study, has suggested. 

(Note: The death rate for those under 25 is too low to be distinguishable on this scale: it peaks at four per day.)

But while the death toll may be much lower, rising infections still raise the risk of hospitals being overwhelmed. Is this still the case? Fortunately, my modelling suggests not.

Hospitals are getting better at dealing with Covid patients. Whereas between November and January 30 per cent of hospitalisations ended in a fatality, only one in five did so of those admitted between mid February and June. Part of this improvement lies with better treatments. But the rate of hospitalisations also reflects the realities of the situation on the ground. When cases are lower, doctors are able to send more patients to hospital because they feel the system can cope. In fact, hospitalisations rise not only because of increasing cases, but also because of doctors directing more patients to hospitals in the belief they have sufficient breathing room.

Despite the vaccines having reduced the risk of serious illness amongst the elderly very substantially, most hospitalisations will still come from those aged 65 and over. This is because the chance of serious disease remains so much higher for the oldest group than it is for younger people.

It is inevitable that the coming third wave will lead to more Covid deaths. The PCCF calculates 8,500 more people in England will succumb between 19 July and the end of the year. However, by that time over 95 per cent of the population will have immunity. Such a figure must be put into context: during the same period in 2020, 6,564 people died ‘due to’ pneumonia and flu, according to ONS stats.

Given the strong performance of the vaccines against all Covid strains seen so far, the epidemic may then be regarded as finally over after this wave. Despite fears to the contrary, there will be no winter wave as there will not be enough susceptible people left to infect. While it is possible that a new strain could arise capable of reducing the protection conferred by the vaccines more than either the Alpha or Delta variant have managed, the reduction in effectiveness would have to be substantial. At the same time, any new variant would need to be highly infectious to give it a chance of taking hold and becoming the dominant strain.

If the transmissibility rate was higher than I have predicted, would it pose a serious risk? The short answer is no. The situation would not really even change if Delta were twice as infectious instead – the PCCF shows that that would cost only a further 500 lives.

Not proceeding with Step 4 and staying as we are in Step 3 until the spring of next year would reduce the peak of infections at the expense of drawing the third wave out into 2022. While infections would peak at a lower level, the PCCF model suggests the number of deaths would only fall by around 2,000. The eventual immunity of the population as a whole, moreover, would also drop by around 7 per cent. This would reduce our protection against new strains.

Because of the levels of mixing currently taking place, changes from the virus’s present level of transmission are somewhat academic. Whether we move the 19 July easing earlier or later, this third wave will still come due to the current degree of mixing. We should confront this wave now in summer, not autumn or winter, since the pressure on the health service is lower.

The vaccines have forced the virus to switch its focus onto the young, who are much better equipped to resist it. This time a year ago, this wave would have infected us all equally. Now, we are in a very different place. Although the pandemic is not over, the sting in its tail has been drawn. The wall of vaccines is working.