OBJECTIVES: To systematically review methodological variations for deriving health state values for EQ-5D-5L.

METHODS: A systematic literature search for articles published in English from January 1990 to September 2018 was conducted. Relevant databases, including PUBMED, PsycINFO, and CINAHL, EconLit, and Cochrane Library, were searched. In addition, targeted search on websites such as https://euroqol.org and Google Scholar was conducted to cover the grey areas and minimize bias in search outcomes. Inclusion criteria included health state valuation studies using EQ-5D-5L. Search key words included EQ-5D-5L, valuation, value set, and tariff. Based on the abstracts and the full texts of the articles, two researchers independently selected relevant studies. A consensus was needed to retain studies.

RESULTS: A total of 227 articles were identified through the initial search. After eligible criteria were applied, 16 articles were retained. The majority of the studies (n=11) used quota sampling technique, while other studies used stratified cluster (n=2), two stage sampling (n=1), and systematic approach (n=2). These studies used either face-to-face interview (n=15) or online survey (n=1) method to collect data from general populations. Sample sizes ranged from 1,000 to 8,222 participants. Final health state values in these studies were derived by using the hybrid (n=9), composite time trade off (cTTO) (n=6), and discrete choice experiment (DCE) (n=1) approaches. Tobit regression model was used to analyze the data in the majority of cTTO studies (n=5), while conditional logit and hybrid regression models were used in the DCE and all hybrid studies, respectively.

CONCLUSIONS: Although the EuroQol Group’s standardized valuation study protocol is recommended to derive health state values for EQ-5D-5L, methodological variations among studies exist. Further studies are needed to examine the impact of these variations on the health state values.