Developing tools for the efficient design of health preference studies

Abstract

Preference information describes the relative desirability or acceptability of specified alternatives that differ across health states, interventions, or services. Studies that generate preference information are being designed to support patient-centered decision making across all stages of the medical product lifecycle, as well as in healthcare more generally. Ensuring high-quality preference research with the potential for impact requires transparent and thoughtful study design, a core aspect of which often includes the development of attributes. Good practices for attribute development in preference studies have started to emerge and demonstrate that developing attributes requires substantial time and effort. Resources to more easily and systematically identify potentially relevant attributes may support the accessibility, interoperability, and reusability of attributes, in turn improving the efficiency of preference study design and comparability of findings across studies. In this paper, we first describe the need for and potential benefit of tools that promote the purposeful re-use of attributes for preference studies. We next present a taxonomy for categorizing and describing attributes that could be applied to facilitate their identification. Finally, we apply this taxonomy to a prototype "attribute library," developed as a part of a Medical Device Innovation Consortium work group, to demonstrate the potential value of these resources to support the preference research community.