Skip to content
Coordinates of ThoughtHome of Philosophy
Join
Question · Open

When the ventilators run out: lottery, first come first served, or most life-years saved?

Wanjiru Njenga Fellow
· 5 replies
BioethicsAnalyticUtilitarianism#lotteries#qalys#scarce-resources#triage
Follow

Its coordinates: 3.3.1 Bioethics and Medical Ethics

During the pandemic a doctor friend in Nairobi described a night when her hospital had two ventilators and five patients who needed them. Since then I've been reading about allocation rules, and I'm surprised how much they disagree.

The main options, as I understand them:

  • First come, first served. Simple and seemingly neutral, but it favours people who live near the hospital, have transport and know the system.
  • Lottery. Treats everyone's claim as equal. Feels fair. Wastes resources if one patient has a far better chance than another.
  • Maximise lives saved. Give it to those most likely to survive with treatment.
  • Maximise life-years, or quality-adjusted life-years. Favour the young and the otherwise healthy.
  • Priority to the worst off, or to essential workers, or to those who haven't yet lived a full life.

Each one encodes a view about what fairness is. As a statistician, I'm also uneasy about how confidently the survival predictions are made.

So, an open question: which principle should govern, and should it be one principle or a mixture? I'm especially interested in objections I haven't thought of.

LikeAnswer
Follow5 answers

Join the conversation: members reply here, and joining is free. Reading is always free.

Join free to reply Sign in

5 answers

  1. Omar Ragab

    Fellow

    From a hospital that faced this: in practice almost every protocol I've seen pass across my desk uses short-term survival likelihood first, then something else to break ties, often a lottery or age bands. Pure first-come-first-served is the default when no one has decided anything, which is the worst way to end up with a rule. The worst night I have had at work was a night with no protocol.

    Helpful · 1
  2. Dima Nasser

    Fellow

    I'd object to quality-adjusted life-years strongly. The quality weights rate a year lived with a disability as worth less than a year without it. So a protocol built on QALYs will, other things equal, give the ventilator to the non-disabled patient, and it will call this objective. Disabled people have been pointing this out for decades. Survival likelihood is less bad, but even then, scoring systems built on "baseline function" can smuggle the same judgement back in.

    Helpful · 2
  3. Javier Ferrandis

    Fellow

    In defence of life-years: there's a serious fairness argument for favouring the young, sometimes called the "fair innings" view. Someone of eighty has had a long life; someone of twenty has had a short one. Giving priority to the twenty-year-old isn't saying her life is worth more, but that she has had less of one. Several proposed allocation systems, such as one published by Persad, Wertheimer and Emanuel in 2009, combine that idea with others rather than relying on one principle.

    Helpful · 3
  4. Elin Hedqvist

    Fellow

    One practical thing, from what I've seen and read: whatever the rule, it must be decided before the night, by people who are not at the bedside, and applied by a triage team rather than the treating doctor. The bedside doctor's job is to fight for the patient in front of them. Asking them to also be the judge breaks something in them, and the doctors I've talked to say so too.

  5. Wanjiru Njenga

    Fellow

    Dima's point about QALYs is exactly the kind of bias I worry about in my own models, a value judgement hidden in a weight and then called objective. Where I've landed for now: survival likelihood first, life-stage as a tie-breaker, lottery for real ties, and Elin's rule about who decides, which seems to me the most important of all.

    Helpful · 1