85:oral The EQ-5D-5L OPUF survey: quantifying health priorities on the individual person level using compositional preference elicitation techniques. (28th April 2022)

• Record Type: Journal Article
• Authors: Schneider, Paul; Brazier, John; Devlin, Nancy; Hout, Ben van
• Is Part Of: BMJ global health. Volume 7(2022)Supplement 2
• Page Start: A32
• Page End: A32
• Publication Date: 2022-04-28
• Subjects: World health -- Periodicals; 362.105
• DOI: 10.1136/bmjgh-2022-ISPH.88
• Languages: English
• ISSNs: 2059-7908

Abstract

Abstract : Introduction: Social value sets (=QALY/DALY-weights) are essential inputs for HTA. To derive a social value set, various methods have been used, including time trade-off, or discrete choice experiments. However, all of these methods suffer from a major limitation: they are inefficient. Little information is obtained from each participant. As a result, data from hundreds if not thousands of participants is required. This limits the ability to derive value sets in resource-constrained settings or for small (patient) groups. Here, we report on the development of the 'OPUF tool'; a new online survey method for estimating value sets for the EQ-5D-5L (or any other health descriptive system). The approach is more efficient than conventional methods, and even allows estimating value sets on the individual level. Methods: The OPUF approach combines different compositional preference elicitation techniques into a new type of online survey. It broadly consists of three steps: dimension weighting, level rating, and anchoring. We tested the feasibility of using the OPUF survey to derive group-, subgroup-, and individual-level value sets for the EQ-5D-5L in the UK. An interactive demo version of the survey is available at: https://eq5d5l.me. Results: A representative sample (N = 1, 000) of the UK population was recruited in August 2021. On average, it took participants about nine minutes to complete the survey. Data from 874 participants were included in the analysis. We successfully constructed a personal EQ-5D-5L value set for each participant, and aggregate value sets for various subgroups. The validity of the models were assessed against the results from discrete choice experiments: the constructed personal value sets predicted participants' choices with an accuracy of 78.5%. Conclusion: Although the development of the OPUF tool is in an early stage, we think there are multiple potential applications and avenues for further research (e.g. patient decision-aid).