We examine the relationship between Visual Analogue Scale (VAS) and Standard Gamble (SG) assumed in the development of the multiplicative multi-attribute utility functions (M-MAUFs) for the Health Utilities Index (HUI) Mark 2 and Mark 3, using data from a UK valuation study of the HUI2. A range of functional forms are considered, and are compared on the basis of their explanatory power and predictive ability.
A restricted cubic function fits the data better than a power curve with a mean absolute error (MAE) of 0.025 and root mean square error (RMSE) of 0.029 compared to a MAE of 0.135 and RMSE of 0.135 for the power curve. The use of a cubic mapping function instead of a power function leads to different predicted health state values.
We question the reliance on the assumption of a power curve relationship between VAS and SG data, in the Health Utilities Index valuation framework. Our results demonstrate that further work is required to examine the appropriateness of the published M-MAUFs for the Health Utilities Indices.