While the government’s decision to remove most lockdown measures in England was widely expected to result in a large wave of infection and disease, the number of new cases of Covid-19 has been falling over the last five days. Many hope this could mean that we’re past the peak. Yet the reality is more complicated. This is the first time an epidemic has taken place in a highly vaccinated population without control measures in place, so we are in uncharted territory. There is considerable uncertainty about what the next two months hold.
The big questions are how high the current wave will get and how long it will last. The number of people in hospital and dying of Covid-19 is directly linked to the number of infections. It’s impossible to accurately predict when we’ll reach the peak of infections, or how long it will take to come back down from this (if I had a pound for every time I’m asked “are we there yet?”, I’d be able to give away a lot of money).
Unfortunately, we will not be able to know if we are at a peak until we have passed it, probably by a few weeks. There are several reasons for this. Though a peak is measured by the number of infections, we don’t observe this number straight away. We only see the outcome of infections. One or two days of low numbers can easily be followed by another day or two of high numbers, so it takes at least a week to establish a trend.
Moreover, a change in one data stream might not be echoed in others. Reported, diagnosed cases are the earliest data collected during a pandemic, but these lag a week behind infections. A drop in cases might indicate a fall in infections, but it might also indicate there are problems with testing. If the prevalence of infection continues to rise, it’s likely that we may see problems with people getting tested, who therefore don’t appear as positive cases. And even if there are tests available, people might not use testing as much, or in the same way, in the future.
Hospital admissions are a more reliable data stream. These lag a week from cases. Even then, a change in hospital admissions might be related to a change in who is getting infected. Younger people, for example, are less likely to end up in hospital. Office for National Statistics (ONS) and Real-time Assessment of Community Transmission (React) data, which is collected from random samples, takes a couple of weeks to be processed and published. Once these show falling numbers of infections, we will have confirmation that prevalence is falling.
So when we have a day with lower cases, and people ask “are we there yet?”, I will say wait a week. If cases have fallen during the week, I will point to rising hospital admissions. But if they fall, and ONS and React data show a fall in prevalence the following week, then we will be able to say that the day with lower cases three weeks earlier was just after the peak, but actually the peak in infections was a week before that.
An even bigger challenge is that the rate of increase of infections (the R number) is determined by two things: contact rates and immunity. At any peak of a wave, R = 1. As immunity increases, due to infection and vaccination, R falls. As we change our behaviour and make more contacts, R increases. It’s a balance. There might be enough immunity to keep R below 1 if we are meeting three people per day, but not if we each meet five people every day.
If nobody had immunity to Covid-19, and everyone was still socialising at pre-pandemic levels, the R number for the Delta variant would be about 6. Currently it is 1.4. Recent contact surveys suggest that we are having about half the social contact now compared with pre-pandemic levels.
It’s quite likely that we will reach a peak, and that, as behaviours change and social contact increases, we will have another peak to climb. And, again, the delays will matter, because we can’t immediately see the effects of increased socialising on case numbers. If the removal of all restrictions in England on 19 July has a big effect on transmission, we won’t be able to see this until the week beginning 2 August. So in the next couple of months it’s unlikely there will be a rise to a single peak and then a fall. It will be more like going over a range of hills, with many peaks in different places, than it will be climbing a single mountain.
The epidemic is going to go up and down in the coming weeks. The number of cases will jump around a lot. It’s only by looking backwards that we can see clearly where we have been. So we will never be able to answer “are we there yet?”. We will only be able to say when we were there. And that might have only been temporary.
The consequence of all this is that our individual risk of infection is going to be higher in the coming months than it has been at any point during the pandemic, because the number of infections will likely be higher than at any other point. Vaccinations will prevent many deaths and keep many people out of hospital. But the risks to individuals still exist. If one thing that is clear from all this, it’s the importance of getting vaccinated.