QALYs, standard gambles, and the expected budget constraint.

This paper shows that when quality adjustment factors for use in constructing QALY indices are established through answers to standard-gamble questions or similar methods, the assumptions that are made by respondents about the financial consequences of changing probabilities of illness and death are critically important for the use of QALYs in cost-utility analysis (CUA). It qualifies Meltzer's [Journal of Health Economics 16 (1997) 33] result that the cost per quality-adjusted life-year (QALY) for life-saving medical interventions should include the future consumption of those who would otherwise not have survived, by showing that its validity depends on how the QALY index has been established. The paper also shows that, contrary to a widely held notion, allocation of a fixed health care budget through CUA does not generally result in a second-best efficient allocation. Another finding is that failure to specify carefully what respondents to standard-gamble questions are supposed to assume about the financial consequences of ill health may result in a bias against providing care to older individuals.