OBJECTIVES: Respondents to health surveys often report that they have never reflected deeply about health states they have not experienced and sometimes struggle to initially value their own health. To assess whether self-rated health is impacted by the experience of thinking about and valuing health, we compared self-rated health before and after completing health valuation tasks.

METHODS: Data from the 2017 US EQ-5D-5L valuation was used (n=1,045). Respondents reported their EQ-5D health state and VAS twice: before (EQ-5D-5L + “pre-VAS”) and after (EQ-5D-3L + “post-VAS”) completing the EuroQol valuation task protocol. Mean pre- and post-VAS scores were compared using a paired t-test. Spearman correlation coefficient (Spearman’s rho) was used to estimate the rank order agreement between pre- and post-VAS scores. To examine how well VAS scores were explained by health classifier responses, OLS models were run on the pre-VAS/5L (Model 1) and post-VAS/3L (Model 2) responses. Models were compared based statistical significance of coefficients (p<0.05), logical ordering within dimensions, and explained variance.

RESULTS: Mean pre-VAS scores were significantly lower (80.0 [SD=16.0]) than post-VAS scores (84.7 [SD=14.5]) (p<0.0001), and there was little agreement between pre- and post-VAS scores (Spearman’s rho= -0.07; p=0.02). Most of the coefficients were statistically significant and logically ordered in both models (Model 1, R=0.42; Model 2, R=0.41).

CONCLUSIONS: Self-rating of health was significantly higher after participant exposure to health valuation exercises, suggesting that reflecting on a wide range of mild to severe health states changes how individuals value/rate their own health.