How you ask is what you get: Framing effects in willingness-to-pay for a QALY

In decisions on financing new and innovative health care technologies a central question is how to determine the value citizens place on the gains in health and life expectancy that result from respective medical treatments. We report results of surveys of four representative samples of the German population. In 2010 and 2012, in total about 5000 respondents were asked for their willingness-to-pay (WTP) for either an extension of their life or an improvement in their health corresponding to a gain of one quality-adjusted life year (QALY). Specific changes of the study design allow for ceteris paribus comparisons of different survey versions. While the initial version exactly copied a questionnaire used in the EuroVaQ (European Value of a QALY) project, which was conducted in nine European countries and Palestine, but not in Germany, in other versions the wording and the survey technique were modified. The findings show that the technique of posing the questions plays an important role when respondents are asked to imagine being in hypothetical situations. This clearly refers to the wording of the questions and the survey setting (personal or online interview). But even simple design elements such as putting a yes/no filter in front greatly affect the answers in terms of both the frequency of zero WTP and the distribution of positive amounts. From the different results, we conclude that it is inevitable to conduct studies comprising a broad variety of versions when trying to elicit WTP for a specific type of QALY in order to achieve an array of values combined by insights into the principles of their sensitivity.