The Case for Reflection : Do Respondents Really Have Well Articulated Preferences or Do They Need Time to Discuss and Reflect Before Making Their Health State Preference

Objectives There is interest in health economics around the values people ascribe when undertaking health state valuation techniques such as the PTO and TTO techniques. Part of the debate is around the fact that people may not have well articulated values, especially given the complex task of health state valuation. There is growing concern that instant responses to a survey question do not reflect people's preferences as well as considered, reflective responses, and that the sharing of information, experience and deliberation may be a mechanism that allows respondents to make a more informed decision. This study explores the effect of discussion and deliberation on health state valuations elicited using PTO and TTO techniques. Methods This study incorporated a planned health state valuation exercise which attempted to encourage reflection through a structured process of discussion in a group setting. Statistical analysis was undertaken to explore the difference between valuations elicited before and after discussion and deliberation. Following the planned health state valuation exercise respondents were asked to take part in a semi-structured interview. The interviews explored respondent's feelings with regard to the group process with particular focus on aspects of discussion and deliberation and its possible effects on their overall response. Results The process is evaluated in terms of the number of people who changed responses following discussion, and the impact of the change on the aggregated health state valuation. A total of 450 individual utility weights were elicited for both PTO and TTO techniques. Following discussion and deliberation a total of 60% of respondents changed at least one utility weight for PTO and 34% for TTO. This had an effect on the aggregated utility weight for some health states but not all. The interview data collected as part of this study suggested that individuals may not have well articulated views and that respondents valued both the opportunity to discuss their views with others and the time to reflect on their own and others responses to the questions. Conclusions A structured process of group discussion resulted in some individuals changing their responses. Whilst change in itself is a poor measure of reflection the qualitative analysis suggests that changes in responses were due to respondents having time to reflect and articulate their valuations. This study challenges some of the notions around the theory and process of preference elicitation