February 4, 2021 - 4:30pm

We have vaccines. Britain has done that remarkably well. But vaccines alone are not enough: we also need a strategy.

Britain also has one of those, and it’s admirably straightforward: vaccinate healthcare workers, plus the oldest and most vulnerable first, and then work down the age groups.

There’s a really solid rationale for that. Deaths from Covid are hugely concentrated among older people. Roughly speaking, someone diagnosed with Covid in their 80s is about 10 times as likely to die as someone diagnosed in their 60s, and about 100 times as likely to die as someone diagnosed in their 30s. The over-80s and people in care homes account for about 60% of total deaths in the UK.

But someone pointed out a potential wrinkle to me. In healthcare, most of the time, we don’t assess the value of an intervention by the number of lives saved; we talk about “quality-adjusted life years” saved, or QALYs. This makes sense. In the end, everyone dies; so has a drug “saved a life” if it delays death by a year? Who knows. But a drug that delays death by five years on average is probably better, all else being equal, than one which delays death by one year.

We haven’t really talked about QALYs when we talk about vaccinations, though. The life expectancy of a British citizen at age 80 is about nine years (a bit less for a man, a bit more for a woman). The life expectancy at age 60 is about 24 years. If we naively assume that people who die of Covid are representative of those groups (they won’t be, but hopefully they’re both unrepresentative in the same way), then vaccinating an 80-year-old becomes only about three times as effective as vaccinating a 60-year-old.

When you factor in other considerations — that younger people are more likely to spread the disease, and that the bulk of hospitalisations and intensive-care admissions are among people between 50 and 70 — the initially overwhelming case for vaccinating the elderly first becomes more marginal.

In practice, I don’t think this ought to change the strategy. For one thing, although QALYs lost are more evenly distributed than lives lost, they’re still heavily weighted towards the older age groups. And for another, having a simple, unambiguous system for assigning priority is a good idea — if we were to use some complex algorithm to establish “vulnerability”, it would be open to being gamed, usually by pushier, better-off middle-class people like me (as with special educational needs support in schools, where wealthier parents are more likely to demand and pay for assessment).

And another excellent reason is that it would be simply politically unacceptable. “It’s worth letting 2.5 80-year-olds die to save one 60-year-old,” and “800 people may have died today, but our vaccination strategy saved about 4,000 quality-adjusted life years,” is a very hard message to get across from the podium. It’s just too callous-sounding.

But I think it is an (other) excellent reason to remember that the game isn’t over when we’ve vaccinated the very elderly. Letting the disease rip through the 50-to-70-year-olds once the older groups have been vaccinated might bring those headline figures down, but the number of years of life lost (not to mention the spread and the hospitalisations) will still be awful.

ADDENDUM: It’s worth noting (I discover since writing this) that the JCVI in fact did look at QALYs in its assessment. Thanks to Kevin McConway and Kit Yates for pointing that out.


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

TomChivers