Purpose: To examine inconsistencies between how people rate their own health and health scenarios using visual analog scales and how much they are willing to pay to improve their own health and the health scenarios.

Methods: We sampled equal numbers of Whites, Hispanics, and African Americans from two populations: adults living in Houston, Texas (N=193) and patients treated for knee osteoarthritis (OA) (N=198) in Houston. Face-to-face interviews were used to collect demographic and preference data. We elicited utilities for participants' own health and for two hypothetical scenarios depicting moderate and severe OA using visual analog scale (VAS) and willingness to pay (WTP) methodologies. We labeled a participant's response as “inconsistent” if they rated either own health better than or equal to a scenario, but were willing to pay more to improve their own health, or rated their own health worse, but were willing to pay less to improve it.

Results: Overall, 57% of participants had inconsistent responses with the moderate OA scenario and 58% with the severe scenario. The vast majority of these (84% and 95%) were due to the individual rating own health more highly but being willing to pay more to improve it. 130 individuals had one inconsistency, and 153 had two. African Americans and Hispanics had slightly higher percentage of inconsistencies than did Whites (58, 62, and 51% respectively). Individuals with a high school education or less had more inconsistencies than those more education (71 vs. 53%).

There was no correlation between inconsistencies and whether the respondent found the WTP or VAS questions easy or difficult. Inconsistencies in the severe scenario were correlated with lower education, being female, and being older. Using multivariate logit regression, we found that age [OR 1.03 (1.01, 1.04)] and education [OR 0.66 (0.45, 0.96)] were significantly related to inconsistencies in the severe scenario, and Hispanic [OR 1.92 (1.06, 3.46)], age [OR 1.03 (1.01, 1.05)] and whether the individual had OA [OR 0.46 (0.27, 0.80)] were related to inconsistencies in the moderate scenario.

Conclusions: Many people rate their own health as better than a hypothetical scenario, yet are WTP more to improve their own health. These findings illustrate the potential discrepancies in individuals' valuation under different methodologies between hypothetical scenarios and own health.