MULTINATIONAL COMPARISON OF VALUES FOR THE 5-LEVEL EQ-5D BASED ON VISUAL ANALOGUE SCALE AND DISCRETE CHOICE MODELING

Wednesday, October 22, 2014
Poster Board # PS4-6

Elise Quik, PhD, University of Groningen; University Medical Center Groningen, Groningen, Netherlands, Nancy Devlin, PhD, Office of Health Economics, London, United Kingdom, Elly Stolk, PhD, Institute of Health Policy and Management/institute for Medical Technology Assessment Erasmus University, Rotterdam, Netherlands and Paul Krabbe, PhD, University Medical Center Groningen, Groningen, Netherlands

Purpose:

   To examine the consistency of the estimated parameters values derived after modeling the preference data and the relationship between the derived discrete choice (DC)  and visual analogue scale (VAS) values across countries in a multinational study.

Methods:

   Health states were quantified based on DC modeling and VAS scores using the 5 level version of the EQ-5D. Computer-based data were collected in Canada, England, the Netherlands, and the United States (US). To standardize the format of the choice tasks across countries, except for England where face-to-face interviewers were used, interactive software was developed. Respondents performed the paired comparison (DC) between two different EQ-5D health states. After each DC task, respondents were prompted to assess the two health states separately on a VAS. A main-effect alternative-specific multinomial probit regression model was used to estimate regression coefficients and to derive DC values. A linear regression model was used to calculate VAS values.

Results:

   At least 400 respondents participated from each country resulting in 1775 respondents in total. Cross-country comparison of DC values to the VAS values shows strong correlations and a quadratic relationship for the set of 400 EQ-5D-5L states. A high similarity of DC and moderate correspondence of VAS values granted to EQ-5D-5L states coming from four different countries were found. The DC as well as VAS values in the UK were found to be lower than the pooled values and values of the other 3 countries.

Conclusion:

   The 5 level version of the EQ-5D derived DC and VAS values show a strong positive relationship. However, the DC and VAS values are hard to compare directly, since this relation is nonlinear. Given some concerns with the VAS, choice methods based on ordinal responses may be a better alternative.

Fig2vasDC

Figure 1

VAS value of the 4 country's compared to the DC value of 400 EQ5D-states

See more of: POSTER SESSION IV & CONTINENTAL BREAKFAST

See more of: The 36th Annual Meeting of the Society for Medical Decision Making