(Note: I have enabled anonymous comments. If the requirement to register discouraged you from chiming in, your concerns have been addressed!)
The main plenary session at the IHEA conference involved a discussion on end of life (EOL) care. EOL care is a huge issue in the United States (and other OECD countries as well) given that a disproportionately large portion of total health care expenditures are spent on interventions and services rendered to those who are within a few months of death. Many health policy experts agree that such spending is inefficient: why spend this much money on those who are going to die anyway? Perhaps this money is better spent elsewhere. Also, some argue that prolonging life is actually reduces patient happiness by extending pain and frustration. There is a huge movement to incentivize the use of palliative care and hospice in lieu of medical interventions for individuals near the end of their innings.
Implicit in the efficiency argument is a cost benefit calculation. Essentially, those who argue that current EOL spending is wasteful note that: (life years saved)*(value of life) < expenditure of EOL care. That is, the benefits accruing from a dollar of end of life spending (particularly on non-palliative or hospice interventions) are less than the costs.
It is (relatively) straightforward (but certainly not easy - lest I invoke the wrath of health finance and cost-effectiveness researchers!) to calculate life years saved and expenditures. What about the value of life? This number is derived in a variety of ways. The main idea is to use individual and institution behaviors that reveal his/her/their internal valuation of life. Here is an example: lets say we have a population of individuals who can choose to work in two sectors, one safe and the other dangerous. Some individuals will choose riskier jobs (higher risk of death or injury) in exchange for higher wages, or a compensating differential. Here the value of life can be computed by dividing the wage premium by the difference in mortality risk between the two jobs. Clearly, this can be generalized across multiple employment sectors. Other means to calculate the value of life include using legal settlements (in the case of injury or wrongful death)
and the amount individuals are willing to pay for safety devices of various sorts.
(Aside: Note that we simply do not ask individuals how much they value their own lives or how much they would pay for an additional value of life. Every time I ask someone why this can't be done, I get some answer to the effect of: well, how can anyone give a reasonable answer to that question? Fair enough. But I wonder if this renders estimates from revealed preference studies useless. If individuals cannot adequately get their head around the value of their own life, how are the implicitly valuing this when rationally deciding to take the risky versus safe job? Let me know if you have any insight on this.)
The main thrust of the plenary talk, given by Tomas Philipson of the University of Chicago, is that the value of life estimates used to declare that EOL care is inefficient may not be appropriate for the situation at hand for a variety of reasons. The big picture point is that value of life estimates are derived from observations on healthy people, and those at the end of their innings may be drastically different in the following ways:
-Individuals at the end of life might value an additional unit of survival time more because they have less of it left. This is the old diminishing marginal utility argument. Everyone loves the first hot dog, but each additional hot dog confers less and less utility (maybe this doesn't apply to Kobeyashi).
-Individuals at the end of life may derive additional utility from hope, thus valuing each additional unit of life more than a healthy person would. This is the Michael J. Fox/Chris Reeve phenomenon: people want to live longer in hopes that new technology might be developed that can save/improve their lives. Another way to think about it is that people will want to live longer if there is some hope for survival.
-There may be an additional social value to life: family members may derive utility from having their loved one stay around longer (unless they are in a hurry for bequests).
-Trade-offs between quantity and quality of life may not be as well-defined as originally thought. There are two offsetting effects: there is greater value to extending a healthier life. However, the amount of consumption forgone rises as the quantity of life declines.
What does this mean? Philipson contends that, if these theoretical ideas hold, those at the end of life may actually value an additional unit of life much more than a healthy individual would. This would fundamentally alter the cost-benefit calculation used to determine whether EOL spending is efficient or not.
Needless to say, this created quite the stir among the audience. I would say the reaction was a mixture of shock, indignation and awe. I talked to some of the attendees afterwards and people seemed equally split between lauding Philipson for his audacity and fresh insight into an old problem and damning him for his arrogance and his implicit belief that neoclassical economics could yield insight into "softer" phenomenon such as hope.
All of this is really interesting to me, especially in the context of this book I've been reading: The Chicago School: How the University of Chicago Assembled the Thinkers Who Revolutionized Economics and Business by Van Overtveldt. Here is a passage from the book about Gary Becker, a Nobel Laureate in economics who is best known for his work on various aspects of human capital and extending economic analysis to topics as diverse as the use of time, educational investments, divorce, marriage, polygamy, intergenerational transfers, altruism, crime, kidney transplants, etc:
"The nontraditional orientation of Becker's research has regularly stirred up emotions and given rise to comments such as this one from Robert Solow [ed: another Nobel laureate in economics]: 'There are some things that should not be analyzed as if they were subject to being bought and sold. Let's be philosophical. Call them primary things, things that are intrinsic to themselves. There are all kinds of social circumstances that are fundamentally different from economic circumstances.'"
If we follow Van Overtveldt's thesis, Philipson's talk had all the elements of the Chicago tradition: the belief that neoclassical economics works and can be applied to a wide range of phenomena, the intense desire to question and search for the truth, and the resulting controversy from the efforts.
I myself found the talk to be interesting and the points to be thought-provoking. I do have a few questions though. First, suppose that we are able to calculate a value of life that is specific to those at the end of life. What does that mean? If we find that the benefits of EOL care exceed the costs, it might imply that people are rational. But so what? Could that money still be better spent elsewhere? Second, I wonder what Philipson has to say about the burgeoning health services literature on hospice, which tends to suggests that people are often happier with palliative care than with intensive medical efforts. If people elect to prolong their lives in a rational manner, can they simultaneously be happier not doing that? Does the shift involve a change in preferences (something like coming to peace with one's fate) or is it a question of information (people have stable preferences but do not have information about alternative modes of care?
No comments:
Post a Comment