Version 2 2017-10-23, 07:12Version 2 2017-10-23, 07:12
Version 1 2017-06-05, 03:14Version 1 2017-06-05, 03:14
journal contribution
posted on 2017-10-23, 07:12authored byRichardson, Jeff
This paper re-examines the question of the appropriate unit of measurement in cost utility analysis and the technique that will measure it. There are two main themes. The first is an examination of the use of utility' as an object of measurement. It is argued that, as it is often conceived and measured, utility' is not an appropriate unit. The second major theme of the paper is the selection and evaluation of a unit of measurement if the usual application of utility theory is rejected. Because of its appeal to utility theory, the standard gamble has been widely accepted as the gold standard for measurement. More specifically, the standard gamble is believed to measure utility directly while other techniques simply measure value' or are pragmatic devices for replicating the results of the somewhat complex standard gamble. In the first part of the paper it is argued that the standard gamble cannot be supported in this way because the theory from which it is derived is unsatisfactory. Three issues are considered. The first is the historical ambiguity in the concept of utility' itself. The second is the validity of the distinction between utility and value which is used to demonstrate the superiority of the standard gamble. Third, there is a consideration of both the evidence and the normative argument used to defend utility maximisation as envisaged by the proponents of the standard gamble. In the second part of the paper the view that there can be no unambiguous unit of output is discussed and rejected. It is suggested that this view may have arisen as some of the measurement techniques in CUA have been adapted from psychometrics where the object of the measurement is often imprecisely defined. Four criteria are suggested which appear to satisfy social objectives and the practical requirements of CUA. These are that there is a social consensus; that an increase in the number of the units corresponds with a socially desired outcome; that the units have a meaningful interval property; that they are readily understood and that it is possible to map the various health outcomes from an intervention into the unit. These criteria are applied to five of the major measurement techniques currently in use including the standard gamble. It is tentatively suggested that Healthy Year Equivalents are probably the best available units of output but that these should be derived directly from the time trade off or equivalent technique. In the final section of the paper there is a consideration of some of the recent criticisms of cost utility analysis and particularly those associated with equity and distributional objectives. It is argued that critics have implicitly demanded more from a unit of output than is reasonable. When its inevitable limitations are recognised cost utility analysis in its present form fulfils an important role in the analysis of resource allocation.