A new method for valuing health: directly eliciting personal utility functions

• Published in: The European journal of health economics Vol. 20; no. 2; pp. 257 - 270
• Main Authors: Devlin, Nancy J., Shah, Koonal K., Mulhern, Brendan J., Pantiri, Krystallia, van Hout, Ben
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer 01.03.2019; Springer Berlin Heidelberg; Springer Nature B.V
• Subjects: Adolescent; Adult; Choice Behavior; Computer Systems; Economic Policy; England; Female; Health Care Management; Health Economics; Health Status Indicators; Heterogeneity; Humans; Interviews as Topic - methods; Male; Medicine; Medicine & Public Health; Middle Aged; Original Paper; Pain Measurement; Pharmacoeconomics and Health Outcomes; Pilot Projects; Preferences; Public Finance; Public Health; Quality of Life; Questionnaires; Surveys and Questionnaires; Utility functions; Young Adult; England; EQ-5D; United Kingdom; Personal utility function; Health state valuation; Stated preferences
• ISSN: 1618-7598; 1618-7601
• DOI: 10.1007/s10198-018-0993-z

Background Standard methods for eliciting the preference data upon which 'value sets' are based generally have in common an aim to 'uncover' people's preferences by asking them to evaluate a subset of health states, then using their responses to infer their preferences over all dimensions and levels. An alternative approach is to ask people directly about the relative importance to them of the dimensions, levels and interactions between them. This paper describes a new stated preference approach for directly eliciting personal utility functions (PUFs), and reports a pilot study to test its feasibility for valuing the EQ-5D. Methods A questionnaire was developed, designed to directly elicit PUFs from general public respondents via computerassisted personal interviews, with a focus on helping respondents to reflect and deliberate on their preferences. The questionnaire was piloted in England. Results Seventy-six interviews were conducted in December 2015. Overall, pain/discomfort and mobility were found to be the most important of the EQ-5D dimensions. The ratings for intermediate improvements in each dimension show heterogeneity, both within and between respondents. Almost a quarter of respondents indicated that no EQ-5D health states are worse than dead. Discussion The PUF approach appears to be feasible, and has the potential to yield meaningful, well-informed preference data from respondents that can be aggregated to yield a value set for the EQ-5D. A deliberative approach to health state valuation also has the potential to complement and develop existing valuation methods. Further refinement of some elements of the approach is required.