SCALE RECALIBRATION IN WILLINGNESS TO PAY FOR FINGER SPLINTS IN PATIENTS WITH RHEUMATOID ARTHRITIS

Purpose: To test whether adding alternatives induces scale recalibration in willingness to pay (WTP) evaluations.

Method: Rheumatoid arthritis patients (n=47) evaluated two different splints, designed to stabilize and align the small joints of the fingers. Splints were made of either silver or thermoplastic. In a randomized crossover design, patients used their first type of splint for four weeks and assessed that first splint (M1), used no splint for two weeks, used the second type of splint for four weeks and assessed that second splint (M2a), immediately followed by a re-assessment of the first splint (M2b). Differences between geometric averages were statistically analyzed using standard t-tests with log-transformation.

Result: At the second evaluation, 24 (51%) of the patients preferred the silver splint, 21 (45%) the thermoplastic splint, and 2 (4%) were indifferent. Average WTP was similar for the silver splint and the thermoplastic splint (€ 13 versus € 12, p=0.63). Average WTP was significantly higher for the preferred splint than for the non-preferred splint (€ 19 versus € 10, p<0.001). Compared to the first valuation (M1), the re-valuation of the first splint (M2b) suggested scale recalibration: the change in valuation among those who preferred the first splint was significantly different from the change among those who preferred the second splint (+28% versus -58%, p=0.03).

Conclusion: Between the first and second evaluation, respondents obtained no new information on the first splint they used. Nevertheless, in the second evaluation, they changed their valuation of the first splint in such a way that the contrast with the second splint was increased: WTP for the first splint was increased if it was preferred and decreased if it was not preferred. This finding can be explained by response-shift theory, range-frequency theory or differentiation-consolidation theory. Regardless of which theories apply, we conclude that WTP measurements can be susceptible to similar distortions that affect quality of life measurements. This challenges the monetary interpretation of WTP measures, at least for substitutable goods.