Purpose: To assess the ability of community-perspective utility scores to incorporate the effects of adaptation to disease by explicitly stating the time of onset of disease in hypothetical health state scenarios.

Methods: An internet-based utility survey evaluating 3 hypothetical health scenarios describing an unspecified disease of varying severity (mild, moderate and severe, in succession) was completed by a convenience sample of 355 adult members of the community, randomly assigned to 2 times of onset of the health condition: within the last week and 6 months prior, or a control group that was given no information about the time of onset. Standard gambles were used to assess scores. Respondents who provided illogical, inconsistent or invariant responses were omitted from the analysis, as were respondents who indicated that they were not confident of their responses. ANOVA was used to compare scores across onset times.

Results: 61% of the total sample (n=215) successfully completed the standard gamble tasks with confidence (31% (n=111) had inconsistent, illogical or invariant scores, and 8% (n=29) stated they did not have confidence in their responses). Among these 215 respondents, hypothetical health states of varying severity described as initiating within the last week were scored similarly to those initiating 6 months previously and to those for which there was no indication of when they commenced. Mean standard gamble utilities for the different onset times for the 3 hypothetical health states were: mild severity: current onset=0.89, prior onset=0.88, unspecified onset=0.90 (p=0.49), moderate severity: current onset=0.70, prior=0.72, unspecified=0.71 (p=0.97), and severe: current onset=0.39, prior=0.39, unspecified=0.39 (p=1.0). Hypothetical health states that were described without mention of the time of onset of the condition were assessed similarly to those that were described as commencing very recently or months prior.

Conclusions: Among the subgroup of community members who accurately and confidently provide utility scores for hypothetical health states, it may not be possible to induce adaptation to disease by stating the time of onset of a health condition. It may be that individuals do not focus on the time of onset of the condition, or that they do not anticipate adaptation to disease. In order to account for the effects of adaptation to disease in economic evaluation, patient utility scores should be considered as an alternative to community-perspective utilities.