September 24, 2020 - 11:55am

I’m a subject in the Oxford/AstraZeneca Covid vaccine trial (As I keep on droning on about). It seems to be progressing well, insofar as I can tell as Guinea Pig #9385. And other vaccine candidates are also coming along — Johnson & Johnson’s vaccine has entered phase 3 trials, the fourth to do so.

There is a real and significant problem for developing vaccines, though, which is that not very many people have the disease. That sounds like a good thing — it is a good thing — but if you’re testing a vaccine, it’s a problem.

Here’s why. In Oxford’s case, they’ve given 5,000ish people the real vaccine and 5,000 a control. Then they watch to see whether the real-vaccine group gets Covid less than the control group. But if the disease is very rare, you’ll hardly see any cases in either group, so it’ll be really hard to compare; I go into the issues in more detail here. Oxford is running trials in other countries, such as Brazil, where the prevalence is higher, but it’s a real problem.

In animal trials, of course, scientists deliberately infect (“challenge”) the animals with the virus, to see if the vaccine protects them. In general, that’s frowned upon in human studies: infecting people with a dangerous disease, knowing that half of them will not even have the protection of the vaccine and that the vaccine may not work anyway, is high-risk and usually vetoed by ethics boards. But a few people, notably the group 1DaySooner, have been arguing that the current circumstances demand it, and that under carefully controlled conditions for fit and healthy volunteers, the risks are acceptable.

Now the FT is reporting that 2,000 volunteers recruited through 1DaySooner are going ahead with these “human challenge” trials in London, under the auspices of Imperial College London, another team working to develop a vaccine.

Full disclosure: I signed up with 1DaySooner back in June, but by the time they got back to me I was on the Oxford trial, and under the Oxford terms I’m no longer allowed to take part. But it seems a good idea to me. The risk of death to reasonably young, healthy people is enormously low; the risk of medium — or even long-term — health consequences are higher.

But, as 1DaySooner itself says, about 5,000 people die every day from Covid-19; if human challenge trials can speed up the development of a vaccine by a single day, and that vaccine can prevent just 25% of those deaths, then it would save 1,250 lives. That hugely outweighs the likely cost to the participants, so, from a cold utilitarian perspective, it seems a good bet. And I could easily imagine that the trials would speed the process up by more than a single day.

For some people, there’s something deeply uncomfortable about the idea of deliberately infecting healthy humans. It may make sense from a utilitarian point of view, but other systems of ethics balk at it (although I probably haven’t got room to go into the full Kant-vs-Bentham argument here). I understand that. But the human challenge trials have the potential to save a lot of lives, so I’m cautiously in favour of them.


Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers