Do Individuals Consider Expected Income When Valuing Health States?

Background: When individuals value health states there is uncertainty about how expected income related to the health states affects the valuations. This may furthermore have implications for which costs that should be included in the health economic analysis, such as productivity costs. Purpose: The purpose was to empirically explore whether individuals consider expected income or not when valuing hypothetical health states, in order to receive guidance for how to handle productivity costs caused by morbidity in a cost-effectiveness analysis. Method: Two hundred students at Linköping University in Sweden each valued four hypothetical health states using a questionnaire. Fifty per cent of them (group I) were explicitly asked to consider their expected income related to the health states in the valuations. For the others (group II), income was not mentioned, illustrating the most common valuation scenario. After valuing the health states, both groups answered a few following questions. The health states were described from the EQ-5D instrument. The valuations were made by both the time trade-off (TTO) technique and the visual analogue scale (VAS). A given income at best imaginable health was presented for group I and they were asked to state how they thought the health states would have affected this income. The valuations of the health states were compared between the two groups by using independent-sample t-tests. Furthermore, a regression analysis was undertaken to test certain variables' ability to explain the variation in the results. Results: Among the students in group I, 40% felt that their expected income had affected their valuations of the health states. Among the students in group II, 96% stated that they had not thought about their expected income when they valued the health states. The valuations made by group I were lower compared to the valuations made by group II for health states that were assumed to affect the expected income. None of the variables - age, sex, or course - could explain these differences. Discussion: The results indicate that individuals are not considering their expected income when they value different health states, as long as income is not mentioned. These findings give support for adding productivity costs to the health economic analysis as a decrease in income does not seem to be captured in the valuations of health states