Measuring treatment preferences of people with type 2 diabetes, an application of stated-preference methods

Stated-preference methods are increasingly used to obtain patient-preference information for use in healthcare decision making. However, there remains a paucity of literature transparently reporting on the application of good research principles through all phases of stated-preference study design. In part one of this dissertation, we applied a novel 5-step framework to the development of a stated-preference instrument that could measure the treatment preferences of people with type 2 diabetes. The developed choice experiment contained 6 attributes (A1c decrease, blood glucose stability, low blood glucose, nausea, additional medicine, and cost). In part two of this dissertation, we applied good research practices to the implementation of the developed discrete-choice experiment. Members of a United States online panel with type 2 diabetes completed a web-enabled, self-administered survey that elicited choices between treatment profile pairs. 552 participants (51% male) completed the survey. We found that patients with type 2 diabetes value both the benefits of their treatment, and the harms and treatment burden associated with treatment. In part three of this dissertation, we sought to assess the impact of educational attainment on treatment preferences of patients with type 2 diabetes. 231 participants had completed high school or less, 156 participants had completed some college, and 165 participants had a college degree or more education. Participants with college or more were willing to pay more for A1C decreases than participants who had completed some college or high school or less. In part four of this dissertation, we conduct a targeted literature review to identify tests for assessing validity and reliability of a DCE. We identified four domains for the validity of a DCE: measurement validity, preference reliability, decision processes, and choice rationality. These domains consist of 14 components that can be identified using 24 possible tests of validity and reliability. Using treatment preferences for type 2 diabetes as a case study, we demonstrated good research practices for the design, implementation, analysis, and evaluation of a stated-preference study. The applications of our approach will help other researchers conduct high quality stated-preference research for use in healthcare decision making processes