In the year and a bit since then, there haven’t been many opportunities to write straightforwardly upbeat things. So it is with great relish that I write that the vaccines are now unambiguously saving British lives.
Like what you’re reading? Get the free UnHerd daily email
Already registered? Sign in
One (so far not peer-reviewed) study looking at 5.4 million people in Scotland — that is, basically all Scottish people — found that vaccinated people were 85% less likely to end up in hospital than unvaccinated ones. The Oxford/AZ vaccine and the Pfizer vaccine were statistically indistinguishable; despite ridiculous and confected European concerns about the Oxford vaccine, it is stopping people from dying just as well as the Pfizer one. And in Scotland the AZ vaccine has been mainly used in older people, to great effect.
Another smaller, UK-wide study looking at infections in healthcare workers found that the Pfizer vaccine was about 70% effective at reducing infections 21 days after a first dose, and 85% shortly after a second dose. Further analysis suggests that even if you do get infected you’re about half as likely to get seriously ill or die, suggesting that it’s somewhere up in the 80-90% region against severe disease.
We’re not mucking around trying to reverse-engineer pharma industry press releases and proxy immune responses any more; this is the vaccine out in the field, doing what it was made to do, and doing it well. One of the most heartening graphs showed that the number of severe cases in the elderly was dropping at a far faster rate than cases in younger age groups; when we came out of Lockdown 1, they stayed stubbornly high for a long time. There was some slight concern in the Scottish data that the response started to weaken a bit once you got past 28 days from the first dose, but it wasn’t clear and it didn’t seem to appear in the other study, so I’m not very worried about it.
When you add in that cases are dropping day by day – the seven-day rolling average is down to around 11,000 positive tests a day, the lowest it’s been since the beginning of October – I think we are allowed to be cheered.
All that might make you think that the road ahead is clear — but there are still reasons to be cautious.
The chief scientific adviser, Sir Patrick Vallance, in a briefing to journalists ahead of the release of the roadmap, said that it is important, as we come out of lockdown, to do it in stages: open up a few things, wait for a few weeks, and watch how R changes; the basic scientific thing of changing one variable at a time.
Schools are the big unknown. It’s nearly a year since schools were closed in the first lockdown, and despite them being opened and closed again since, we still don’t really know what impact they have on transmission. I think it’s pretty clear that they do drive R upward somewhat, but it’s not clear how much. I get different answers depending on which scientists I ask.
In the roadmap, we will open schools on 8 March (plus a few other things which are probably fairly inconsequential from a transmission point of view), and then see what impact that has. Vallance suggested we should leave four to five weeks between new measures, in order to see what the impact is: under the roadmap, we’re leaving just three weeks before further measures are introduced on 29 March, which disease models presented to SAGE suggest would be “unlikely [to] allow the effect of step 1 to be clearly discernible before step 2 takes place”.
The measures introduced on 29 March seem unlikely to be major causes of transmission: the reintroduction of “the rule of six” outdoors, outdoor sport starting up again. And they will coincide with the start of the Easter holidays. More major changes will come at least five weeks after the schools reopen. But still, this isn’t being overly cautious: if anything, it’s sailing somewhat close to the wind.
You might, reasonably, say that it doesn’t matter that much whether schools raise R that much, because the vaccines will stop people dying. But not everyone will be vaccinated, and the vaccines are not perfect, so a huge outbreak will lead to a gigantic surge in cases. The SAGE models suggest that if we were to reopen fully all at once, we would end up with the NHS overwhelmed far more than it was even at the peak of spring 2020.
Slower relaxation, obviously enough, leads to slower growth in cases. And given the uncertainty, especially around schools, coming out slowly is obviously wise. I understand the wish to come out of lockdown quickly: by one measure, we currently have the most restrictive measures in the developed world. But I’d suggest that’s probably at least partly a result of coming too quickly out of previous lockdowns, not a reason to come out quickly now.
There’s another concern, which is that the more infections there are, the more chances there are for the virus to mutate. There are already somewhat vaccine-resistant strains out there. “If we see big outbreaks and exponential growth,” says Rupert Beale, leader of the Cell Biology of Infection group at the Crick Institute, “you might get a more complete vaccine-escape mutant. It’s relatively unlikely, but each new infection is like giving the virus a lottery ticket.” It only needs to be lucky once.
So the roadmap seems sensible: finally Boris Johnson has learned how to pass the inverted marshmallow test.
A couple of points, though. First, it is baffling that it is “irreversible”, as briefed a few weeks ago. If we reopen schools, they drive R way above one, and we find ourselves in an unexpected new surge, then the sensible thing to do will be to close schools again. It will be awful for a lot of people, but better than a huge second outbreak and the need for a much longer lockdown.
Second, there are missed opportunities here. The number of cases is falling, and that’s a good thing. But as Adam Kucharski of the London School of Hygiene and Tropical Medicine points out, it means that we have only a short window to find things out about the virus. When there are very few cases in the community, it becomes impossible to tell what effect your interventions have: if you tested a vaccine on a population with no disease, then you’d get zero cases in your placebo arm and in your treatment arm. The same applies if you are trying to see the effect of, say, lateral-flow testing.
So now, while there are still a reasonable number of cases, would be an excellent time to do some randomisation. Try lateral-flow tests in schools in a randomly selected 50% of boroughs. Or try new ventilation techniques in half of schools but not in others. Then measure the impact. It is very unlikely that this will be the last we see of Covid, in either this form or some variant; knowing more about how it spreads, what we can afford to leave open and what we definitely need to close will help us when numbers come back up again.
Third, there was something that Vallance said which caught me up short. We can’t expect the vaccine to be entirely successful, he said, because even if it’s 80% effective, we’re only vaccinating 80% of the population – adults – and 80% of 80% is only about 65%.
Which is of course true, but it made me think: perhaps … we … ought … to … vaccinate … children?
It’s just a thought, but 100% of 80% is 80%. The vaccines are amazingly safe in adults, and I don’t know of any reason why they’d be less safe in children. I was worried that there’d be some Joseph Heller-style situation where we can’t do safety trials on children because we don’t know that it’s safe to do so, but it turns out that Pfizer have already recruited participants. We’re going to have to live with this disease for a long time. Vaccinating people on their 16th birthday forever is not ideal, especially if it’s still circulating in schools. It might not be the most immediate priority but I’d like to know if it’s safe.
I sincerely doubt that this will be the last time I write “coronavirus” in my life, or even the last time this week. But the vaccine news does make me think that, if we do come out of things cautiously, then perhaps this time next year, I’ll at least be writing it less often.