Objective: To test the validity of the techniques used to calculate quality-adjusted life-year (QALY) values based on utility functions, using a real population dataset.
Design: Using the standard gamble technique, we gathered the preferences of a population sample of 189 individuals on a combination of probabilities concerning four simple health states (no physical disability, limp, walk with crutches and need a wheelchair), and three life-year spans (5, 10 and 15 years). Each of the four assumptions of the multiplicative model was tested based on the results of the experiment.
Results: The utility of the health state "limp" was high at 0.89 and that of "walk with crutches" only slightly lower at 0.85. However, of the 189 individuals, only 57 are not in contradiction with the assumption of mutual utility independence since they strictly preferred (15 years, "need a wheelchair") over (10 years, "need a wheelchair") and (15 years, "need a wheelchair") over (5 years, "need a wheelchair"). For these 57 individuals, the results of this study do not fit the assumptions underpinning the multiplicative model.
Conclusion: This work suggests that the techniques used as a basis from which to calculate QALY values are flawed. In particular, the underlying assumptions of the multiattribute utility model do not correspond to behaviour patterns observed in a real population. It therefore appears that use of the QALY technique should be questioned in healthcare decision-making settings.