Mapping EuroQol-5 Dimensions (EQ-5D) utilities from cancer-specific non-preference measures have used ordinary least squares regression and, more recently, a variety of more complex statistical regression methods.

We have shown that these should be rejected in favour of three-part models that are more able to take into account the tri-modal distribution of the 3-level (EQ-5D-3L) measures.

Further research should be undertaken to validate our results in other cancer data and with the more recent 5-level (EQ-5D-5L) questionnaire.