Empirical comparison of discrete choice experiment and best-worst scaling to estimate stakeholders' risk tolerance for hip replacement surgery

Objectives Empirical comparison of two preference elicitation methods, discrete choice experiment (DCE) and profile case best-worst scaling (BWS), regarding the estimation of the risk tolerance for hip replacement surgery (total hip arthroplasty and total hip resurfacing arthroplasty). Methods An online survey was constructed, following international guidelines, and consisted of socio-demographic questions and two randomised sections with 12 DCE and 8 BWS questions. The survey was sent to a general population who can be faced with choosing between THA and TRA (males between 45-65 years old) in the US. After an intensive literature search, the following attributes were selected: probability of a first and a second revision in seven years, pain relief, ability to perform moderate daily activities, and hospital stay. In addition, survey respondents rated the difficulty of each method and the time to complete each section was monitored. BWS and DCE data was analysed using conditional logit analysis. The maximum acceptable risk (MAR) for a revision was estimated for four different hypothetical hip replacement scenarios. Results The final data set consisted of 429 respondents. The MARs estimated for four hypothetical hip replacement scenarios differed between both methods, ranging from 0% to 19% difference for a first revision. BWS questions took significantly more time (401 s.) than DCE (228 s.) questions. And respondents found BWS more difficult to complete. Conclusions Both methods to elicit stakeholder preferences produce different results. Yet, both seem to be consistent in predicting risk tolerance if the benefits are changed. However, DCE seems to be more sensitive for a change in benefits and risks while the MAR estimates obtained through BWS have considerably lower uncertainty than DCE