Responses to standard gambles: are preferences 'well constructed'?

Standard gamble (SG) is commonly used to elicit preferences in order to assess health related quality of life. There has been little qualitative research exploring how respondents answer such questions. An SG study was designed to elicit values for the health states associated with anti-hypertensive medication, stroke and cardiovascular disease. This paper describes a qualitative study that was carried out alongside the SG exercise in order to document the thought processes respondents bring to bear in formulating their responses. Data were generated using 'think aloud' techniques and semi-structured interviews. Values were generally well-constructed: responses were thoroughly considered, and respondents made complex trade-offs and arrived at a point of indifference. However, some respondents incorporated inappropriate information into their choices, redefining the hypothetical 'Option B' resulting in problems interpreting the probabilistic information. Consideration of non-health factors was commonplace, in particular the impact of choices on others. We discuss these findings in terms of the use of qualitative methods in health economics and the wider discourse surrounding the theoretical underpinnings of health state valuation. Copyright (C) 2003 John Wiley Sons, Ltd.