Abstract
After criticizing existing systems of health measurement for their unargued commitment to evaluating health states in terms of preferences or well‐being, this essay argues that public rather than private values of health states should help guide the allocation of health‐related resources. Private evaluation of health states is relative to a prior individual choice of specific activities and goals, while public evaluation is relative to the whole range of important activities and goals. Public evaluation is concerned with securing a wide range of choices as well as with success given one's choice. A reasonable simplification from the public perspective is to focus on just two features of health states: the subjective feelings attached to health states and the limitations that health states imply on the range of important activities that individuals can pursue. Focusing on just these two dimensions permits the construction of a parsimonious classification of health states with regard to what matters most from the public perspective. This classification, which resembles those in the HALex and the Rosser and Kind Disability and Distress Index, might best be built on top of existing health‐state classifications, by mapping the health states they define to activity‐limitation/feeling pairs.
To assign values to these pairs, I propose relying on deliberative groups to make comparisons among the pairs with respect to the relation ‘is a more serious limitation on the range of objectives and good lives available to members of the population’. A ranking according to this property, is not a preference ranking, because it is not a ranking in terms of everything that matters to individuals.
Working back from the weights attached to the activity‐limitation/feeling pairs, one can impute weights for the health states in other classification systems that were mapped to those pairs. If those weights coincide roughly with current weights, then one legitimizes current weights and provides a vehicle for their public discussion and possible revision. If those weights do not coincide, then one has both an argument for revising current views of the cost effectiveness of treatments and policies and a method to carry out such a revision. Copyright © 2009 John Wiley & Sons, Ltd.