An Empirical Comparison of Discrete Choice Experiment and Best-Worst Scaling to Estimate Patient Preferences in...

Infertility has become a growing public health problem in China. It is important to accurately and easily obtain patient preferences. This study aimed to obtain and compare stated preference results from the perspective of infertility patients. By assessing the validity and acceptability of both methods, it provides lessons for empirical research and practical application.

Discrete choice experiments (DCE) and profile case best-worst scaling (BWS-2) are methods used to identify and weigh various criteria affecting the order of priorities. We conducted a face-to-face survey of female patients with infertility aged between 20-45 years. The survey included socio-demographic information, preference questionnaires and completion of evaluation questions. Attributes included live birth rate, pregnancy rate, degree of participation in treatment decision making, maternal complications, neonatal complications and program cost. Conditional logit models were used to analyze attribute level weights and relative importance was calculated separately.

A total of 330 valid questionnaires were collected. The preferences of patients experiencing infertility were quantified through two stated preference research methods. The findings indicated that patients exhibited a preference for treatment options that were highly effective, exhibited minimal side effects, were patient-centered, and were cost-effective. The BWS-2 and DCE preference weights demonstrated high consistency, with only slight difference observed in the ranking of individual attributes within the order of relative importance. In the view of the patients, the DCE questions were perceived to be less challenging to comprehend and were therefore preferred to be completed.

The BWS-2 and DCE exhibit identical validity and highly consistent preference results. In the context of specific research questions, the selection of a method or the combination of methods must be suitable to the purpose of the study in order to ensure that the utility gained is maximized. Further research is required to corroborate these findings.