Tuesday, November 18, 2014

Health Improvement vs. Treatment

Appeals to quality-adjusted life-years (QALYs) in medical resource allocation decisions are naturally supported by a broadly utilitarian view of the role of health institutions, i.e. as having the purpose of improving social welfare (via health improvement) as much as possible.  But is that the right view to have? My colleague Mary recently pressed me on an intuitive alternative conception of healthcare as aiming at treating localized health problems rather than yielding global health benefits to patients.  Might that be a better view?


When contrasting the views, two major points of difference seem noteworthy. The first is that the "Treatment" view depends upon (and reinforces) the commonsense distinction between treatment (restoring you to "normal" health) and enhancement (as going beyond "normal" health), whereas on the "Improvement" view, we simply consider the benefit to be gained from a given health intervention, with no need to compare it against any assumed baseline of normalcy.

The second point is that the Treatment view will be less "discriminatory" than assessing QALY benefits, in that it will tend to treat other aspects of the patient's health status (including disability status, age / life expectancy, etc.) as irrelevant to allocation decisions.  After all, if the aim is just to treat local conditions of ill-health that arise (this disease, those infections, that cancer, etc.), this goal is just as well achieved by curing the disease of a 90 year old patient with Alzheimer's as it is by curing that same disease in an otherwise healthy 20 year old.  Either can be restored to a health state without the disease in question, and no further comparisons need or should be made (according to the Treatment view) about the respective values of those health states, or how much better off each patient is made by receiving the treatment in question.

One immediate challenge that arises for the Treatment view is how to make sensible allocation / prioritization decisions.  Presumably nobody wants to treat curing a cold as just as important as curing cancer, for example.  But what is the basis for considering the latter more important, if not that the treatment does more good (or prevents a greater harm) for the patient?  I guess if the aim is to make allocation decisions in a way that disregards individual factors other than the condition being treated, one natural solution would be to opt for allocation by generic QALYs, where this is understood to mean the QALY gain that could be expected from treatment for a "generic individual" with the condition in question. (The "generic individual" might in turn be understood as an otherwise healthy 40 year old, say. Though I'm not really sure what the basis would be for choosing that particular age over any other.  Perhaps a better view would be to use the age of the actual patient, thus allowing for age-based "discrimination", and simply using the generic-QALY construct to avoid differential treatment based on disability (etc.) status.)

While that seems a possible view, it's a little unclear to me what principled motivation could be offered in its support (i.e. besides simply wanting to avoid the implications of the standard QALY approach).  Deliberately ignoring pertinent information (namely, how much individuals stand to gain from proposed treatments) strikes me as rather perverse.  It is like how "common sense" bioethicists typically seem to prefer indiscriminate/coarse-grained allocation methods over more fine-grained ones -- throwing away information in a way that seemingly only harms our attempts to make our allocation practices more closely approximate the normative facts.

When different interests are at stake, why would you want to treat the cases indiscriminately?  Given that health is a mere means to what really matters in life (our "interests", broadly construed), why would we want our health institutions to narrowly focus on providing "treatments", independently of how much the individual patients in question can actually be expected to benefit from them?  It seems incomprehensible.

The only sense I can make of it is a broadly "rule consequentialist" justification -- if you do not trust our institutions to make decisions that are even on the right track -- if their attempts at fine-grained allocation could be expected to actually do more harm than good -- then that would be a practical reason to favour a more coarse-grained, indiscriminate approach.  But of course such merely pragmatic reasons are no reason to think that the QALY approach is in any way unjust in the way that most of its critics seem to believe.  It is, on the contrary, to recognize that while QALYs would be ideally just, the dangers posed by our flawed institutions mean that justice must take a back seat to more prosaic considerations of institutional competency.

5 comments:

  1. "While that seems a possible view, it's a little unclear to me what principled motivation could be offered in its support (i.e. besides simply wanting to avoid the implications of the standard QALY approach). Deliberately ignoring pertinent information (namely, how much individuals stand to gain from proposed treatments) strikes me as rather perverse."
    OK, without wanting to come across all Catholic (I'm not, I'm a good atheist!), isn't a principled motivation for this kind of view respect for the sanctity of human life? This would seem to require treating everyone as in principle equally worthy of a potential cure for a disease, independent of whatever other fortunes/misfortunes have beset them elsewhere in their lives. And this would seem to require what you describe as "Deliberately ignoring pertinent information (namely, how much individuals stand to gain from proposed treatments)" - from the perspective of a sanctity of life view, this kind of information just isn't pertinent.

    ReplyDelete
    Replies
    1. Hmm, perhaps. Though I think it's important to stress that the Improvement view doesn't claim that any person is less "worthy" of treatment, but just that some treatments don't do as much to help some people compared to others. Insofar as the "sanctity of life" is just a label for the view that quality of life doesn't matter, I worry that it's really just a re-statement of the problem (why think that quality of life doesn't matter?) rather than a principled motivation for that viewpoint. (But maybe the relevant notion of 'sanctity' moves others in a way that I just don't understand? Never have had much exposure to Catholicism!)

      Delete
  2. I think I may be missing something in what the Treatment view is supposed to be. I'm not sure why you would think that the Treatment view commits one to denying *any* consideration of benefits, as opposed to restricting considerations to benefits directly related to the treatment as such. (That is, localized benefits rather than global ones.) In a sense that's what Jean-Dominique Larrey invented the concept of triage for -- that is, the real concept of triage rather than what usually goes by the name of 'triage' in most places today. The whole point of the concept of triage was originally to make decisions about allocation and the like entirely on the basis of medical necessity in light of the specific end of restoring to health. This deliberately excluded any broader and more general considerations of the patient, but did admit consideration of the specific benefits involved in the restoration of health (among other things).

    I'm also not sure I understand your question: " why would we want our health institutions to narrowly focus on providing "treatments", independently of how much the individual patients in question can actually be expected to benefit from them?" Why wouldn't one simply answer this by pointing out that people in fact primarily make use of health institution for treatments, not to have the quality of their lives judged? What in particular authorizes health institutions in general, where not specifically given the right by the particular patients in question, to base their actions on assessments that are that general? (This is a general problem with institutions. To take an extreme example, but far from the only one, would it really be just, even in principle, if the government started allocating all government services on some standard of quality-adjusted citizenship years, rather than on some more specific criterion directly linked to both the actual services and what the citizens took the services to be for?)

    ReplyDelete
    Replies
    1. Hi Brandon, right, I didn't mean to suggest that "the Treatment view commits one to denying *any* consideration of benefits, as opposed to restricting considerations to benefits directly related to the treatment as such." Though I think it is potentially a challenge for the view to spell out exactly how that line -- of being "directly related to the treatment as such" -- is to be drawn (hence my idea of generic QALYs).

      "Why wouldn't one simply answer this by pointing out that people in fact primarily make use of health institution for treatments, not to have the quality of their lives judged?"

      Well, certainly judging quality of life (or life expectancy, or anything else) is not itself a purpose of health institutions. But plausibly we make use of the treatments offered by health institutions because they benefit us, and what we really care about is this benefit, not the treatment per se. And if making these intermediary judgments can help health institutions to benefit us more (via more efficient allocation of limited resources), then that sure seems desirable to me (and, I would expect, to anyone thinking clearly from behind a veil of ignorance).

      "would it really be just, even in principle, if the government started allocating all government services on some standard of quality-adjusted citizenship years"

      I'm having trouble imagining what exactly this would look like. But I suspect (if it's modelled on the QALY idea, and similarly aims to yield maximal benefits to those affected) that any objection I'd have to it would, similarly, be merely practical. But maybe I'm just not grasping what you have in mind vividly enough.

      Delete
    2. I'm not sure how patients' interests in how the treatments benefit themselves would have much to do with QALY assessment at all. Surely patients don't generally think of their benefits in those terms -- they generally think entirely of those benefits traceable to the treatments themselves. When people in Europe have done studies on this, for instance, they've always found that what patients are expecting from health institutions tend to be very different from what health institutions using QALYs would give. So the question just stays the same: What authorizes health institutions to use such general criteria where they are not authorized to do so by the patients themselves?

      The problem with talking about a more efficient allocation of resources in this context is that efficiency is only defined relative to goals. It's not possible to determine what is the most efficient allocation of limited resources in a health institution context until one has already determined what that allocation of resources is supposed to be for.

      (I'm also a bit puzzled about your veil of ignorance comment. I've been thinking for the past several minutes about how one could possibly get anything like QALY assessment under veil of ignorance conditions, and I can't think of any way that would work -- given a veil of ignorance, no one knows how they would be assessed in QALYs, so the optimal solution from a veil of ignorance seems to be to have our health institutions to focus on addressing immediate problems as they come up, to the extent that solutions are possible, practically manageable, and likely to be effective, and to leave more general considerations to more general institutions in which everyone has a say. This seems much closer to the Treatment view -- again, assuming I am not misunderstanding what is meant by it -- than to anything like QALY assessment.)

      None of this is a general argument against QALYs, of course; I just don't see how your response avoids begging the question against the Treatment view, assuming I am not misunderstanding that view.

      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.)