How should EQ-5D data (value sets) from multinational clinical trials be used?

Data on EQ-5D health states gathered in the context of a multinational trial may be converted into a single summary index using one of the available EQ-5D-3L value sets. There are different options available to do this using appropriate value sets, however, the choice depends on the context in which researchers or decision makers will use the information.

In cases where data from a multinational trial are to be used to inform decision makers in a specific country, it seems reasonable to expect decision makers to be interested primarily in value sets that reflect the values for EQ-5D-3L health states in that specific country. For example, if applications for reimbursement of a drug are rolled out from country to country, then country-specific value sets should be applied and reported in each pharmaco-economic report. This is no different from the requirement to use country-specific costs. In the absence of a country-specific value set, the researcher should select another set of values for a population that most closely approximates that country.

Sometimes information about utilities is required to inform researchers or decision makers in an international context. In these instances, it is probably more appropriate to apply one value set to all the EQ-5D health state data. The decision about which value set to use will depend on whether the relevant decision-making body in each country specifies any requirements or preferences in regard to the methodology used in different contexts (e.g. TTO, standard gamble (SG), VAS or discrete choice modelling (DCM)). These guidelines are the topic of an international ongoing debate, but the EuroQol Group is planning to provide a summary of health-care decision-making bodies internationally, including their stated requirements regarding the valuation of health states.

Detailed information regarding the EQ-5D-3L valuation protocols, guidelines for choice of value set and tables of all available value sets have been published by Springer in: EuroQol Group Monograph series: Volume 2: EQ-5D value sets: inventory, comparative review and user guide’ (see section 8 for more information). Chapter 4 by Nancy Devlin and David Parkin will be of special interest to researchers pondering the issue of which value set to use.