EXPERT SURVEY FOR INCREMENTAL COST-EFFECTIVENESS RATIO (ICER) THRESHOLD RANGE IN KOREA- DISCRETE CHOICE EXPERIMENTS
Author(s)
Lee J, Ko SPfizer Korea, Seoul, South Korea
OBJECTIVES: This study was performed to elicit ICER threshold range in Korea and to investigate the magnitude of relative influence of main attributes to reimbursement decision. METHODS: This study was based on experts survey using discrete choice experiments (DCE). DCE is stated preference method, in which respondents are requested to express preferences for sets of hypothetical choice alternatives constructed according to experimental design principles. Total 39 respondents (22 from the government or academia and 17 from industry) were responded to this survey among 95 selected experts who were working in pharmacoeconomics field. We identified 6 attributes (disease severity, availability of alternative therapy, burden of disease, ICER, uncertainty of cost-effectiveness ratio and budget impact) which are considered as main decision factors for reimbursement decision. Literatures related to reimbursement decision in Australia, UK, Canada and Korea were comprehensively investigated to elicit these attributes. 18 choice sets having two alternatives were selected using fractional factorial design (FFD). Each coefficient of 6 attributes and probabilities of a recommendation response by ICER range were estimated. RESULTS: Total 1346 observations were analyzed using probit model. 4 attributes (severity, burden, alternative, ICER) significantly influenced on respondents' choice on reimbursement decision (p<0.001). If the disease is severe and rare, the probability of recommending the alternative as cost-effective is higher. The lower ICER, the higher the probability of choosing the alternatives. ICER range threshold from survey results is 70,000,000 KRW approximately. If industry group were excluded in the analysis, the figure reduced to 40,000,000 KRW. CONCLUSIONS: This study showed that severity, burden, alternative, ICER is important factor for reimbursement decision in Korea as well expected. The ICER threshold estimated from this study between the estimators from previous studies. The study design and/or respondent characteristics could affect these figures.
Conference/Value in Health Info
2009-10, ISPOR Europe 2009, Paris, France
Value in Health, Vol. 12, No. 7 (October 2009)
Code
PMC28
Topic
Real World Data & Information Systems
Topic Subcategory
Health & Insurance Records Systems
Disease
Multiple Diseases