Author(s)
Yang Z1, Purba FD2, Shafie AA3, Igarashi A4, Wong EL5, Ahn J6, Pattanaphesaj J7, Jo M8, Lin H9, Lam HY10, Minh HV11, Mai VQ11, Luo N12, van Busschbach J2
1Erasmus University Rotterdam MC, Rotterdam, The Netherlands, 2Erasmus University Medical Center, Rotterdam, The Netherlands, 3Universiti Sains Malaysia, Penang, Malaysia, 4University of Tokyo, Tokyo, Japan, 5The Chinese University of Hong Kong, Hong Kong, Hong Kong, 6Ewha Womans University, Seoul, Korea, Republic of (South), 7Ministry of Public Health, Bangkok, Thailand, 8University of Ulsan College of Medicine, Seoul, Korea, Republic of (South), 9Pharmerit International, Newton, MA, USA, 10Institute of Health Policy and Development Studies, National Institutes of Health, Manila, Philippines, 11Hanoi University of Public Health, Hanoi, Viet Nam, 12National University of Singapore, Singapore, Singapore
OBJECTIVES: METHODS: In each study, 196 pairs of health states were valued using DCE, with each respondent valued 7 pairs. A main effect conditional logit model was used to estimate the regression coefficients for each country. For comparison, we created interaction terms between 20 main effect dummy variables and country dummies and make comparison between each two countries. Then, we summarize i) the relative importance of the five dimensions; ii) the relative utility decrement, or distances between each of the five levels. RESULTS: Between country comparison showed significant difference in regression coefficients, especially for the severe/extreme levels. Comparing the relative weights of five dimensions, there appeared to be three groups.The relative importance of severity levels (how much accounted for level 5 for each level) showed less variations across countries (20% for level 2, 30% for level 3, 65% for level 4). CONCLUSIONS: Our results showed that health preferences differed over countries, when in absence of the extraneous factors affecting the valuation task. It is therefore justifiable to announce that the difference in EQ-5D-5L value set reflects the genuine difference in cultural notions and that therefore different national values sets have to be used in different countries.
Conference/Value in Health Info
2018-11, ISPOR Europe 2018, Barcelona, Spain
Value in Health, Vol. 21, S3 (October 2018)
Code
PRM207
Topic
Methodological & Statistical Research
Topic Subcategory
PRO & Related Methods
Disease
Multiple Diseases
