“Because that’s what matters to me”. A pilot study to test the feasibility and reliability of ordinal valuation methods for health state valuation with children

K Stevens


This study sought to test whether it is feasible and reliable to obtain children's preferences for paediatric health states from the Child Health Utility 9D (CHU9D) using pair wise discrete choice experiment and best worst scaling techniques.

31 interviews with children ranging in age from 7-17 years were carried out and cognitive debriefing techniques were used to determine whether children could understand and respond to the tasks.

Pair wise comparison techniques were found to be feasible and reliable for children age 14-17 years and best worst scaling techniques were found to be feasible and reliable for children age 10-17 years. Children aged over 11 years could also do a second best and worst choice and those over 15 years could do a full ranking. There was a preference for best worst methods by children and it was rated as less difficult by them. Children commented that with a best worst question, they could keep reviewing to see if something else was better or worse whereas with pair wise comparison they found it difficult to retain all of the information in their head.

This research has demonstrated that it is feasible and reliable to obtain children’s preferences for health states from the CHU9D using either pair wise or best worst scaling techniques from age 14 years, or from 10 years for best worst scaling. There is potential for future valuation studies of child utility measures to obtain child preference weights using these methods.