Patient preference for a long-acting recombinant FSH product in ovarian hyperstimulation in IVF: a discrete choice experiment

STUDY QUESTION What factors or attributes of a long-acting recombinant FSH (rFSH) or daily-administrated rFSH influence women's preferences IVF? SUMMARY ANSWER Patients' preferences for rFSH products are primary influenced by the attribute ‘number of injections’, but a low ‘number of injections’ is exchanged for a high ‘number of injections’ at a 6.2% decrease in ‘risk of cycle cancellation due to low response’ and at a 4.5% decrease in ‘chance of OHSS’. WHAT IS KNOWN ALREADY Injections of long-acting rFSH have been claimed to be preferred over daily-administrated rFSH injections, but patient preference studies to underpin this assumption have not been performed. STUDY DESIGN, SIZE, DURATION A discrete choice experiment (DCE) was created to assess women's preference for long-acting or daily-administrated rFSH under varying attributes of efficiency, safety and burden. The selected attributes were the ‘total number of injections’, ‘chance of ovarian hyperstimulation syndrome (OHSS)’ and the ‘risk of cycle cancellation due to low response’. Questionnaires were handed out during information gathering sessions in one academic hospital and two teaching hospitals in The Netherlands between April 2011 and April 2012. PARTICIPANTS/MATERIALS, SETTING, METHODS Women at the start of their first IVF treatment were asked to participate in this patient preference study. Participation was voluntary. We analysed the data by using mixed logit models to estimate the utility of each attribute. MAIN RESULTS AND THE ROLE OF CHANCE Questionnaires (n = 125) were handed out with a response rate of 77% (97/125). Four respondents did not complete the questionnaire. Hence, there were 93 questionnaires available for analysis. All attributes significantly influenced women's preference. Overall, the lower ‘number of injections’ was preferred above the higher ‘number of injections’ (mean coefficient 1.25; P < 0.001), while an increase of 1% in ‘chance of OHSS’ or 5% ‘risk of cycle cancellation due to low response’ was non-preferred (mean coefficients −0.31 and −0.24, respectively, P < 0.01). The majority of respondents was willing to trade-off a lower ‘number of injections’ for a higher ‘number of injections’ when gaining a 6.2% reduction in ‘cycle cancellation due to low response’, or a 4.5% reduction in ‘chance of OHSS’. LIMITATIONS, REASONS FOR CAUTION The generalizability of this DCE is limited in time-span. Women may choose differently when they have previous experience with long-ac