The Conceptual Basis for SMPH Qualifying DALYs; Dallying with QALYs: How are we to Evaluate Summary Measures in Population Health?

This paper is concerned with various issues surrounding the validity of summary measures of population health (SMPH). As it argues that economic orthodoxy cannot answer this question the paper considers the prior issue of the criteria with which to evaluate SMPHs. It is argued that as there are multiple objectives or uses for SMPHs then it is likely that more than one such measure will be needed and specifically that the SMPH most appropriate for measuring the burden of disease will not necessarily be the same as the SMPH used for economic evaluation. With respect to the criteria for judging SMPHs, it is argued that the use of the Bale of Ignorance, favoured by Murray et al, 1999, is unsatisfactory. Alternative criteria are proposed and discussed. One criteria is that SMPHs should embody ethical values consistent with stable population values. This raises the prior question of the relationship between observed population values and social objectives. It is suggested that the investigation and analysis of this relationship should be described as 'empirical ethics'. Four examples of this in the context of SMPHs are discussed. First, it is shown that population values may be context specific. Secondly, it is shown that the measures of utility employed to derive SMPHs are all 'contaminated'; that is, they do not measure, empirically, what it is believed that they measure. Thirdly, evidence is presented that the preference number for utility attached to different health states will systematically differ when it is elicited from patients and the community. Fourth, there is an illustration of the way in which empirically derived utility values may be used to derive implications which may encourage the revision of stated values after deliberation (the 'strong interval property'). It is concluded that there is a need to recognise the importance of 'empirical e the techniques and rules which will govern its use and a need to apply empirical ethics to a potentially large range of social problems.