How should different life expectancies be valued?

EDITOR—Commissioners of health care have to decide which of the many possible uses of resources should be given priority, given that we cannot afford everything that might do some good for some people. One way of comparing possible investments is to consider cost per life year gained. Extending life from 3 months to 9 months at a cost of £4000 costs the same per life year gained as extending life by 5 years at a cost of £40 000—namely, £8000. Let's leave quality of life, and comparisons of the benefits of mortality and morbidity aside for the time being.

Our problem is this: should 6 months of life, when that is most of what is left, be valued equally with 6 months that is part of a much longer survival period, or should it be valued as part of the progression to complete recovery? Should the time that is left weigh more heavily the closer a patient is to death? We suggest, for debate, that if life expectancy, including the extension, is less than 6 months then any benefit in terms of life years gained should be trebled, and if it is less than 12 months the benefit should be doubled. This assumes that life becomes more precious the nearer to death one gets: a “duration of life left” effect. Do readers agree with this weighting?

Such weighting assumes adjustment for quality of life. There may also, however, be an effect of brevity of life. Some patients may opt for minimum treatment, such as best supportive care, rather than for a slightly longer life of poorer quality. A patient's attitude may be different, however, if an event such as a family wedding or graduation coincides with the potential extension. Patients who have only a short time left to live may tolerate a poorer quality of life than those who have a longer time, and work has shown that poor states of health become less tolerable the longer they last.1 The implication may be that if life is short there should be less downward adjustment for quality in calculations of the quality adjusted life year (QALY).

One assumption here is that the patient is aware of the approximate time left, which is more likely when the life expectancy is shorter. The “duration of life left” effect might also apply towards the end of longer life expectancies, which would have implications for the discounting of future QALYs.

We invite comments, as a letter either to the BMJ or direct to Norman Waugh.