Intended for healthcare professionals


How should different life expectancies be valued?

BMJ 1998; 317 doi: (Published 24 October 1998) Cite this as: BMJ 1998;317:1155

Existential model may be better than scale that uses quality adjusted life years

  1. Tim Benson (tb{at}, Chairman.
  1. Electronic Point of Care, Wembley, Middlesex HA0 1SU
  2. Institute for Medical Technology Assessment, Erasmus University, Rotterdam PO Box 1738, 3000 DR Rotterdam, Netherlands

    EDITOR —Problems with the valuation of life expectancy and quality adjusted life years1 can be traced to the underlying philosophical paradigm. The quality adjusted life year paradigm is based on 19th century classical utilitarianism. The greatest good is perfect health, which is valued at 1; being dead (and the event of death) has the value of 0. Other states are given values of <1; states judged worse than death may be given negative values.2

    This century, philosophers such as Martin Heidegger and Sir Karl Popper have put forward ideas that differ from those of the utilitarians and suggest an alternative model of how to value health. Heidegger's premise is that we are unaware of things when they are normal; our conscious concern is with the abnormal. For example, we are not aware of using a door handle when we enter a room unless it is broken.3 We are not consciously aware that our body is healthy; we become aware of our body only when we are ill, injured, or dying. This existential approach directs attention to aspects of health that we are aware of, such as distress, disability, and impending death.

    Valuations when scale based on quality adjusted life years (QALYs) and existential model based on Heidegger's and Popper's premises is used

    View this table:

    Popper exhorts us to minimise misery and misfortune, not seek to maximise good. These are not symmetrical. One person's suffering cannot be traded for another person's happiness. There is an analogy here with Popper's premise that the task of science is to eliminate false theories,not to attain ultimate truth.4

    Deaths at different ages and in different circumstances have different consequences and should be valued differently. Once one is dead one has ceased to exist (at least for direct health care). This distinction, between the event of death and the state of being dead, is ignored in the literature about quality adjusted life years.

    The table compares the valuations when the QALY scale and an existential model are used. It shows fundamental asymmetry. The task of commissioners of health care is to allocate resources in order to minimise the overall consequences of morbidity and death for their population. An existential model provides a direct indicator of these consequences. Unfortunately, the scale that uses quality adjusted life years creates a utilitarian distortion.


    Diminishing marginal utility and discounting future effects have similar consequences

    1. Werner Brouwer, Research fellow.,
    2. Ben van Hout, Senior research fellow.
    1. Electronic Point of Care, Wembley, Middlesex HA0 1SU
    2. Institute for Medical Technology Assessment, Erasmus University, Rotterdam PO Box 1738, 3000 DR Rotterdam, Netherlands

      EDITOR —Waugh and Scott propose that health effects should be tripled or doubled when total life expectancy is below 6 months or 12 months, respectively, when the duration of lifetime left is taken into account in economic evaluations.1 Economic theory may be of help in the issues that they raise.

      Firstly, the principle of attaching more weight to benefits gained when life expectancy is short corresponds to the economic principle of diminishing marginal utility, reflecting the idea that giving an additional sandwich to someone who has little to eat is preferable to giving it to someone with a lot to eat. This principle implies that giving an additional quality adjusted life year to a person with a quality adjusted life expectancy of 20 years is less valuable than giving one to a person with a life expectancy of only 3 months. This notion is already used implicitly: lifesaving lung transplantation, with huge costs per quality adjusted life year, is considered worthwhile, whereas prevention programmes aimed at people with high cholesterol concentrations, with much lower costs per quality adjusted life year, are not considered cost effective. The same notion may explain the acceptance of high costs in the last year(s) of life, when potential health gains and life expectancy are often low. Waugh and Scott's proposal to triple or double health effects is as arbitrary as is making no adjustment, and more research is needed to find the appropriate weights.

      Secondly, correcting for diminishing marginal utility may partly solve the fact that people with a short life expectancy may be more willing to accept a poor quality of life than people with a longer life expectancy. This relates to one of the principles underlying quality adjusted life years —that of constant proportional trade off; this means that equivalence between 10 years in health state A and 5 years in health state B implies equivalence between 10 months in A and 5 months in B. Again, more research is needed to indicate how the concept of quality adjusted life years should be adapted for situations involving short life expectancies.

      Finally, Waugh and Scott mention discounting future effects. Discounting and diminishing marginal utility are two different subjects, with similar consequences, but from completely different backgrounds. Thus they should be treated separately.