Thursday, October 04, 2012

QALYs, DALYs, and Complete Lives

Persad et al's 'Principles for allocation of scarce medical interventions'* offers interesting criticisms of existing theories of just allocation, and proposes a new account that they call "the complete lives system".  I recommend checking out the whole paper, here I'm just going to explore one strain of it.

By way of background: "QALYs", or Quality-Adjusted Life Years, are self-explanatory.  "DALYs", or Disability-Adjusted Life Years, sound like they should be the same thing but (confusingly) also build in an instrumental component, discounting the "unproductive" years of the very young and elderly in favour of the (young-ish) working-age population.

Persad et al. criticize the DALY system on the grounds that "DALY allocation wrongly incorporates age into the outcome measure, claiming that a year for a younger person is in itself more valuable. Priority for young people is better justified on grounds of distributive justice." (428)

I find this criticism confusing.  Firstly, at least as they describe it, the DALY system is explicitly based on the idea that a working-aged life-year is instrumentally rather than "in itself" more valuable.  Is this "wrong"?  It seems factually plausible, though one could reasonably ask whether such non-medical factors should enter into medical decision-making at all.  (In principle, it certainly seems reasonable to prefer that public resources go to help as many people as possible, whether directly or indirectly.  But such a policy might have a high risk of poor implementation or other unintended consequences, rendering it unwise to implement in practice. I remain agnostic on the practical question -- the answer's not obvious to me either way.)

Moreover, it seems extremely plausible that some years of life are intrinsically more valuable for us than others.  (I previously noted this as a respect in which QALYs are insufficiently discriminating.)  Persad et al seem to recognize this, at least implicitly, when they explain how their own view would take age into account.  After noting a general principle of favouring the young on prioritarian grounds (they haven't yet had a "fair innings"), they add:
Consideration of the importance of complete lives also supports modifying the youngest-first principle by prioritising adolescents and young adults over infants... [A]dolescence brings with it a developed personality capable of forming and valuing long-term plans whose fulfilment requires a complete life. As the legal philosopher Ronald Dworkin argues, “It is terrible when an infant dies, but worse, most people think, when a three-year-old child dies and worse still when an adolescent does”

Here, intuitively, the adolescent's death is worse for his own sake -- a datum that may be explained by the fact that he has a developed personality, would have thwarted plans and projects, etc., in contrast to a cognitively undeveloped infant with nothing yet to lose.  But I would similarly argue that the twenty years of life from age 15 - 35 are likely of greater value to the individual than the twenty years from 70 - 90.  After all, the former span is likely to involve much more experiential diversity, is more central to the development of one's values and sense of self, and may also involve more central life projects (from relationships to career choices) than one is likely to be undertaking in very old age.  In short: The former twenty years are utterly transformative.  The same can't be said of the latter.  They may still be very valuable, to be sure; but they can't compare to the significance of more transformative decades of life.  I think our intuitions reflect this: A life cut short at 70 rather than 90 seems not as radically and detrimentally truncated as one that's cut short at 15 rather than 35, even though the same number of life-years are at stake in either case.

So while Persad et al. object that "DALY allocation treats life-years given to elderly or disabled people as objectively less valuable", in fact this seems like a point in favour of the view.  It's simply true that those life-years are less valuable, both instrumentally and in terms of welfare value: the value of those years to the individual who would live them.

If anything, the "Complete Life" perspective should help us to appreciate this fact.  When we look at our lives as a whole, we can see that some periods are more centrally important than others.  It's vitally important (for one's welfare) to live a full, complete life.  It's a nice bonus, but ultimately less important, to achieve additional life-extension beyond what is necessary to live a "complete life".  This isn't just a matter of distributive justice, as Persad et al seem to think.  It's a fact about welfare, or the value of the life(-year) for the person living it.

(But I do wish we had a less misleading name than "DALYs" for including age-based adjustments.  Something like "Welfare-Adjusted Life Years" would make more sense!)



* Persad G, Wertheimer A, & Emanuel EJ. (2009) 'Principles for allocation of scarce medical interventions' Lancet 373(9661): 423-31.

7 comments:

  1. The paper: https://dl.dropbox.com/u/13100539/ht/allocation-principles.pdf

    ReplyDelete
  2. One must recall what DALYs are designed for: population level health policy. But QALYs can be generated for any particular age.

    ReplyDelete
  3. It is not accurate to say that DALYs is the name "for including age-based adjustments." The WHO, which invented the concept of the DALY, did age-weight DALYs, but DALYs needn't be age-weighted. The essential feature of DALYs is simply that they are a measure of a health gap--a gap between the number of healthy life-years lived and some number of healthy life years that is chosen as a normative ideal. If 80 healthy years is the standard, then someone who lives to 79 in perfect health and then dies will have generated 1 fully DALY. Whether that DALY is discounted via age-weighting is a separate choice. As I understand it, anyway.

    ReplyDelete
  4. The DALY measure is part of the Global Burden of Disease (GBD) methodology for measuring population health (it also serves other ends). Efforts to improve GBD are unabating, and its leadership (which has moved from WHO to the Institute for Health Metrics and Evaluation, a Gates-supported research group in Seattle) tries to respond to criticisms of every element, including the DALY. This includes a continuing discussion of age-weighting. The current status of age-weighting for the DALY will be announced in the special issue of The Lancet devoted to GBD 2010 (the most recent iteration). It may change further.

    ReplyDelete
  5. Watch for the special issue (in mid-December)of The Lancet that will be devoted the latest iteration of the Global Burden of Disease project. It will give an account of the current status of age-weighting and the DALY. The debate continues and further commentaries could change its future course.

    ReplyDelete
  6. I agree with the general practice of DALYS, but want to point out one are where measuring them may break down. Imagine you have the choice between extending the life of one of two people by ten years. One person cannot walk, the other person can walk with a cane. It seems like at first glance we will get more welfare by extending the lifespan of the person who can walk with a cane.

    But imagine that the person who cannot walk has the life-goal of being a scholar, while the person who walks with a cane has the life-goal of being an athlete. In this case it seems like we will generate more welfare by extending the life of the person who cannot walk, since their disability does not impair their primary life-goal nearly as severely.

    ReplyDelete
    Replies
    1. Yes, absolutely. I'd say they're better thought of as our best practicable correlate with welfare, rather than being a direct measure of welfare, for just the sorts of reasons you point out.

      Delete

Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)