Do they just know more, or do they also have different preferences? An exploratory analysis of the effects of self-reporting serious health problems on health state valuation

Background There is a debate on whose health state preferences to use in economic evaluation. This may affect the values, but it is not clear whether this is caused by different understandings of the states (an “information mechanism”) and/or by different opinions of the states (a “preference mechanism”). The case for using better-informed preferences is straightforward, but if the opinions of health states depend on people's health experience, the case to justify privileging the opinions of some over others would be harder to make. This study proposes a new framework to examine the effect of having health problems on health state valuation by distinguishing “within-dimensional” effects (which can be due to better information or due to differences in opinion) and “cross-dimensional” effects (which must be due to differences in opinion). Methods Secondary data from the UK that valued EQ-5D-5L using Discrete Choice Experiments with duration (DCETTO) are remodelled controlling for whether a respondent self-reports serious (viz., severe or extreme) problems in ‘pain or discomfort’ (PD) or ‘anxiety or depression’ (AD). The main analysis uses respondents with serious PD or serious AD alongside matched respondents who do not, and assumes constant proportional time trade-off. Results Self-reporting serious PD or serious AD problems has no within-dimensional effect on health state preferences. However, self-reporting serious AD problems has negative cross-dimensional effects on the utility of having any problem in PD, which suggests that the preference mechanism is present. A similar pattern holds when all available (unmatched) data are used and when constant proportional time trade-off is relaxed. Conclusions Findings consistent with the preference mechanism indicate that those with serious health problems may have different opinions on the value of health states compared to the rest of the population, which has normative implications for the debate on whose and which values to use