• Sessions
• Presentations
• Posters
• Rooms

Presentation

Session: Methodology 2
Room: Meeting Room 21
Time: Sat 08:30-09:45

Presenter: Dorte Gyrd-Hansen (Institute of Public Health – Health Economics Unit, University of Southern Denmark & Danish Institute for Health Services Research. )

Abstract

Objective: The main purpose of the present study was to test the extent to which outcome scope insensitivity is prevalent when performing an internal as well as an external scope test. Secondary purposes are to test for anchoring and adjustment bias when respondents are asked two consecutive contingent valuation (CV) questions and to test for payment vehicle bias by asking respondents to express their monetary valuations in annual versus monthly payments.

Methods: The study was conducted using an internet panel. The sample was obtained from The Nielsen Company’s online database in July 2007. Each respondent was randomised to three different versions of the questionnaire (A, B and C). In arms A and C respondents were initially asked to value a severe health state by way of a time-trade-off (TTO) exercise, which was conducted interactively. The health state was described using the EQ-5D descriptive system. The health state in question was 22222 (i.e. some problems on all five dimensions). Respondents randomised to arm B were also asked to do a TTO exercise, but the health state they were to value was less severe: 11221 according to the EQ-5D descriptive system corresponding to some problems with performing usual activities and some pain and discomfort. Subsequent to the TTO exercise all respondents were asked to value an intervention, which offered a reduction in risk of cardiovascular disease from 2 out of 1000 to 1 out of 1000 over a period of 10 years. The negative outcome associated with the risk involved moving into a chronic poor health state. The poor health state was described as equivalent to the health state that the respondents were asked to value in the TTO prior to the CV exercise. All respondents were subsequent to this initial CV exercise asked to value the same risk reduction, but in this case the negative outcome was death.
Results: Although our study passes the internal scope test, there is not a high degree of sensitivity to outcome. As many as 68% of respondents stated an identical maximum WTP in first and second CV valuation exercise implying that they value the interventions equally despite the fact that the health state presented in the initial CV question was deemed far better than death according to the TTO responses given by the same respondents (mean QALY loss per year in health state: 0.46). In contrast, the external scope test (comparison of response to initial CV across study arms) fared much better.

Conclusions: The preferences exhibited in the present study are characterised by a high degree of validity. Our results do, however, suggest, that this validity is based on the responses provided to the initial CV question and less so by the responses provided to the second CV. Amongst those respondents who do not express identical WTP, the responses to the second CV are affected by anchoring and adjustment bias due to influence of the initial stimuli in the first CV task.

Key Terms

Willingness-to-pay, ordering effect, internal scope test, external scope test, anchoring and adjustment bias

Authors:

Dorte Gyrd-Hansen (Institute of Public Health – Health Economics Unit, University of Southern Denmark & Danish Institute for Health Services Research) , Trine Kjær (Institute of Public Heath –Health Economics Unit, University of Southern Denmark) and Jytte Seested Nielsen (Institute of Public Health – Health Economics Unit, University of Southern Denmark)

Search with Google