A comparison of conjoint analysis and time trade off elicited utilities for quality of life states associated with the treatment of genital herpes
Doward, L. C., Ratcliffe, J., McKenna, S. P., Meads, D. M., & Langley, P. C. (2002). A comparison of conjoint analysis and time trade off elicited utilities for quality of life states associated with the treatment of genital herpes. In , pp. 574–574. .
OBJECTIVES: To compare and contrast the utilities generated for quality of life (QoL) states from the recurrent genital herpes quality of life questionnaire (RGHQoL) using two main elicitation techniques: conjoint analysis (CA) and time trade off (TTO). METHODS: A questionnaire was designed for face-to-face interview including QoL states defined by six RGHQoL attributes representing aspects of QoL. A sample of patients (n = 207) diagnosed with recurrent genital herpes consented to participate in the study. Interviews were conducted with 192 (92.8%) of these. Respondents were asked to compare QoL states using a discrete choice CA framework and to complete a TTO exercise in which they were asked to indicate how many years spent in specific QoL states would be equivalent to 10 years with perfect QoL.
RESULTS: Some degree of concordance was observed between the CA and TTO elicited utilities. The same health state achieved the highest utility value regardless of the elicitation technique used. However, statistical comparison of CA and TTO elicited values for health states indicated statistically significant differences overall (Wilcoxon test, p = 0.037). CONCLUSIONS: Unlike CA elicited utilities, the utilities generated using the TTO technique are considered to have cardinal properties and hence have the advantage that they can be used within a quality adjusted life year (QALY) framework for economic evaluation. However, our results suggest that there was a much greater degree of intra-respondent variation in TTO elicited values than CA elicited values. Many respondents struggled to grasp the concept of TTO and provided inconsistent responses as a consequence. In contrast, the degree of completeness and consistency with the discrete choice CA exercise was high. Further research should explore the potential use of CA in developing utilities that possess qualities such that these can be used directly within a QALY framework.