The relative social-willingness to pay instrument: justification and initial results
journal contributionposted on 07.06.2017, 03:50 by Richardson, Jeff, Iezzi, Angelo, Sinha, Kompal, McKie, John
The paper presents the case and evidence for the use of a new scaling instrument for measuring the value of health states, the Relative Social-Willingness to Pay (RS-WTP). In Section 1 it critiques existing scaling instruments used in economics for measuring health-related quality of life, vis, the SG, TTO, PTO, VAS, WTP1,I In Sections 2 and 3 it compares them with the RS-WTP. It is argued that each of the existing trade-off instruments has a potentially serious upward bias. Empirical results are presented from a comparison of eight instruments (four postal versions of the RS-WTP, interview results for RS-WTP, TTO and two versions of the PTO). Subsidiary analyses include the validity and reliability of postal versions of the instrument, the effect of several edit algorithms, a change in the framing of the instrument, a change in the questionnaire preamble and format and the relationships between the two versions of the PTO and the TTO instruments. With respect to the main subject, correlations of 0.93 and 0.95 between mean values of RS-WTP, TTO and PTO scores represents strong validation of the new instrument and supports the case for its use as a stand-alone instrument. The RS-WTP produces values over the range of observations which are about 15 percent lower than the PTO and TTO. As the instrument has no obvious bias this further supports the argument that conventional instruments exaggerate utilities. This does not imply that RS-WTP values are (technically) ‘correct’ for use in CUA. Results have been obtained from a particular group (the public) using a particular form of question, and perspective, and the issue of what is appropriate is unresolved. The RS-WTP can accommodate changes in these elements. In the concluding sections the justification for an additional scaling instrument is reviewed and future research described.