• Could the marginal health care dollar be put to better use?

    By e-mail, Brad Flansbaum focused my mind on this question long enough for me to respond as follows. It seems right to me at the moment, but it’s certainly worth debating.

    1. Spending on health is not without value. It does improve our lives. (See Cutler.)

    2. Yet in the US we spend a lot to get that value. So, price per QALY (or something) is very high. (See Aaron’s series on spending and his other on quality.)

    3. Just staying within the realm of health,* the price per QALY on another “service” might be a lot lower (like nutrition, exercise, healthy habits).

    If God were jointly designing all health-related systems and functions of society and government, He’d look at the marginal cost/QALY over all possible ways to spend the next dollar and pick the smallest. It’s not always going to be on health care services and it probably isn’t given what we’re already spending for those and what we’re getting for that spending.

    That doesn’t make health care services worthless, just worth less. 🙂  Physicians and hospitals are doing good, but not as much good as might be done with the resources expended. As a health economist, I’m doing even less good, improving the health very, very few and by hardly anything. I probably make a lot of people worse off. (Stop reading. Go brush your teeth or take a walk. There, that’s better.)

    This is a reason to advocate for lower spending on health services, but only if the dollars saved are put to better use. Tax cuts are not obviously better for health, for example.*

    * One can certainly argue that health is too narrow. Some broader notion of “welfare” would widen the field to considerations of substituting anything for health services if they were more welfare improving. Maybe doubling the speed of the internet produces more welfare per dollar spent at the margin than health care, for example. However, to keep things simple, I’m just thinking about health here, not welfare in general.

    Comments closed
    • Austin
      Thanks for the response. For readers interested in more clarification, the question (or observation) that generated discussion was this citation from Feb ’11 HA:

      In Canada–Ontario for this study, despite availability of universal access, lower SES individuals still have poorer outcomes. Why?

      Issue may not be health insurance–keeping in mind 10% of premature death is attributed to medical services, the rest is secondary to behavior, genes, etc.–but lack of spending and emphasis on population health (read: parks, recreation and space, stress reduction, institutional biases and disparities).

      The value of some health services available through insurance then, is potentially modest at best,and the cost/QALY for many services needs reappraisal. A bus with seat belts may be a better value than an implantable defibrillator. Regardless, Austin is far more authoritative on the market dynamics behind that dysfunction than me.

      On the QALY comparison front, I have also referenced this Am J Pub Health classic, and it is worth a look:


      • @Brad F – Thanks. Not a critique of your contribution to the conversation, just by way of moving it forward: As you know, mortality is no where near a complete view of the benefits of health care and, because it is a long-range effect for the vast majority of people, it is very hard to measure a mortality effect of health care for all but the sickest individuals or in special circumstances (see .

        So, QALY is good, as are other measures. I was hoping the QALY paper covered more than health care. There must be something out there that provides guidance of the effects of a wider range of types of investments. What would God read to do the work I suggested in the post? 🙂

    • We already know most of the answer, just not how to achieve it. Get people to stop smoking, stop drinking so much, decrease drug use, eat better and exercise more. None of those really come under the purview of medicine per se. Bad working conditions and hours are also mostly not something that medicine can do much about.

      All that said, I think that we know even less about how to cure social ills than we do medicine, or maybe we just care less. I think that we should concentrate on trying to get as good a ROI as possible in medicine. Try to get costs down, improve access while maintaining or improving quality. The rest of the first world does it so we should be able to do it also. Trying to address these issues as part of health reform will sink the effort IMO.

      ” What would God read to do the work I suggested in the post?”

      Didnt God invent public medicine journals for this kind of stuff?


      • @steve – I’m sure God has inspired some great work. There’s likely a key paper or two that plausibly ranks the ROI of the things you mention and others. I’m just not familiar with this type of literature, nor do I have time to pursue it. But if someone knows of that paper (or two or three), I’d love to see them.

    • This is probably not adding anything you all don’t already know. But there is a reason why conscientious health services researchers working in cost-effectiveness analysis qualify their work by noting that CEA can only be a metric of efficiency, not equity/fairness/justice. The late ’80s/early ’90s example of Oregon is often cited here. Essentially a cost-effectiveness framework was used to set priorities in the Medicaid budget. The proposal was scrapped early on in part because maximizing QALYs leads to the recommendation that tooth-capping and TMJ splints should be given higher priority than appendectomies. Sometimes small benefits for lots of people can get you a bigger QALY-for-your-buck than large (even life-saving) benefits for fewer people.

      If this sounds problematic to you, you’re not alone. Then again, virtually everyone to whom this seems problematic is also fine with speed limits that produce small bits of convenience for lots of people at a cost of thousands of people dying each year. Reduce the speed limit by 10 MPH, save lives. Reduce it by 10 MPH more, save more lives. At some point, we are not willing to give up the convenience enjoyed by lots of people to relatively fewer (but still many) lives of traffic victims. So what, philosophers wonder, is the difference between speed limits and health care?

      One difference is suggested by rescues like the Chilean miners. A lot of reflection is required to understand why we won’t hesitate to pay millions for such a rescue, but we’ll haggle over much less money to be used for organ transplants. Nevertheless, many feel that there is something having to do with the proper response to ongoing vulnerability that makes rescuing the identified victims a priority. But that just pushes the problem back one stage, since the victims of higher speed limits will be identifiable some day.

      All this is to say that there may be something special about health care–just as there seems to be something special about appendectomies–that makes it worth paying for, even if cost-effectiveness is not always on its side. I know Austin has written a lot about what that something might be, but I also know that he agrees that much, much more work needs to be done. In any case, thanks for keeping the conversation going.

    • A classic paper that touches on these issues from a different angle is Geoffrey Rose’s “Sick Individuals and Sick Populations,” which I’ve uploaded here:

      Although Rose does not make reference to cost-effectiveness analysis, his discussion is related, since small reductions in the risks for those with moderate risks can produce more health benefits than large reductions in risk for the most disadvantaged. If we were solely interested in boosting population health outcomes tied, say, to hypertension, we may therefore choose to focus resources on compliant well-to-do white men in the suburbs instead of high-risk black men in the inner city. Small reductions in the risks faced by those at moderate risk can have a bigger bank-for-your-buck, since a population at moderate risk can contribute more deaths than a smaller population at much higher risk. Rose’s point is subtle, but among the most important there are in population health, where the unreflective dictate is often: Maximize health outcomes!

      • @Paul – Thanks for the paper. It’s in my virtual pile. These are thorny issues. In all manner of rationing, our gut response is driving so much by expectations and norms. Sometimes they make sense given the obligations and expectations of the actors (e.g., an army physician rationing resources to the effect of sacrificing one soldier with severe complications in order to save five others). Sometimes they don’t.

    • (The document I link to above includes an introduction by Michael Marmot. The Rose pieces follows that.)

    • @Austin–As one who teaches bioethics, I know as well as anyone how thorny they are. A very nice book that I use with my students is Peter Ubel’s “Pricing Life.” But I would not dream of recommending it and making your pile bigger.