Validité théorique de la Méthode des Choix Discrets : le cas du Traitement Hormonal substitutif de la Ménopause

Discrete Choice Experiments (DCE) consists in providing individuals with fictive scenarios in which characteristics of the good to be evaluated are varying, and asking them to indicate their preferences. Under some behavioural hypotheses, DCE allows to explain individuals' trade-offs between those characteristics. Since Hormone Replacement Therapy (HRT) presents risks, benefits and monetary costs we aim at exploring the potentialities of DCE regarding individual preferences elicitation, by an application to HRT. In this paper, we study the theoretical validity of the method by testing standard behavioural hypotheses: internal consistency, transitivity, and stability of the preferences, no lexicographic preferences, no framing effects. Methods: Eleven pairs of scenarios were derived from the following HRT attributes: climacteric troubles, osteoporosis fractures, colorectal cancer, breast cancer, cardiac risk, thromboembolism risk, and monetary cost of the treatment. Women's demographics and medical background were also collected. Two versions of the questionnaire were developed and randomly administered allowing to test framing effects due to the presentation of probabilities. Theoretical validity was tested by descriptive statistics. Results: 462 women aged 45 to 65 and from Lyon, France, sent back their questionnaires. Internal consistency, transitivity and stability of preferences were observed in large proportions of women (95%, 97% and 88%, respectively). Only 12% expressed lexicographic preferences. No framing effect was detected on data quality. Conclusion: since behavioural hypotheses were verified, it will now be possible to estimate a utility function and to discuss women's preferences and willingness to pay for HRT.