Study Protocol: Comparison of Inconsistency between Time Trade Off and Discrete Choice Experiments in EQ-5D-3 L Health State Valuations

Abstract Background: The EQ-5D-3L is the most widely used multi-attribute utility instrument for describing health states. A popular method for valuing the EuroQOL five dimension (EQ-5D)-3L health states is the time trade-off approach (TTO) where quality of life is traded against length of life. However, TTO valuations can provide logically inconsistent values. That is, where a respondent provides a utility value for one health state that is lower than the score they give for a logically worse health state. More recently there has been a tendency by researchers to use discrete choice experiments (DCE) as opposed to TTO in health state valuation exercises; however, DCEs often exclude dominant choices by design. The aim of this paper is to explore the differences in the rate of logically inconsistency health state valuations between TTO and DCE methodologies. Methods: A representative sample of the Australian general population will be recruited from an online cohort. Of the 243 EQ-5D-3L health states, a number of health state sets, comprising of potentially logically inconsistent health state pairs, will be used for the valuation. Participants will be asked to value given health state sets using both TTO and DCE methods. Consequently, the proposed study is not a health state valuation exercise, but rather an evaluation of competing methods under controlled circumstances. Logical inconsistency will be assessed based on comparing quantitative health state valuations within the TTO and stated preferences from discrete choices within the DCE. The count of logical inconsistencies will be estimated at an individual level for both approaches and compared. The comparison of the two approaches will identify if there are significant differences between the number of logical inconsistencies produced from DCE and TTO methods. Discussion: A difference in logical inconsistency is only one of many criterions for selecting the best approach for conducting health state valuations. It is recommended to examine the strengths and limitations of each methodological approach both theoretically and empirically. This is novel research into important methodological concerns often overlooked up until now