OBJECTIVES: To examine the performance of a generic health preference-based instrument, Child Health Utility – Nine Dimensions (CHU-9D), compared to a disease-specific preference-based instrument, Atopic Dermatitis Quality-of-Life (ADQoL) in valuing the quality-of-life for children with atopic eczema.

METHODS: Participants in the CLOTHES Trial, (www.nets.nihr.ac.uk/projects/hta/116501) evaluating the provision of silk garments for children with moderate to severe eczema, were asked to complete the CHU-9D, ADQoL, and Patient Oriented Eczema Measure (POEM) pre- and post-intervention (6 months). For both utility measures, using data on participants aged 5 and over at baseline, we assessed their practicality (response rate), agreement (using Lin’s Concordance correlation coefficient), construct validity (ability to discriminate between POEM severity levels using ANOVA), and responsiveness (ability to discriminate between those whose total POEM score showed i) No improvement, ii) Improvement of <3 points, and iii) Improvement of ≥

RESULTS: The baseline response rate was 100% (of 137 participants) for ADQoL compared to 97.8% for CHU-9D, and 91.2% versus 90.5% post-intervention respectively. There was fair agreement between the two instruments at baseline (0.313 (95% CI 0.210, 0.415); and moderate agreement at follow-up (0.413 (95% CI 0.318, 0.508). Both instruments were able to discriminate between participants with different levels of POEM severity at baseline (p<0.01) and follow-up (p<0.01). Further, the mean ADQoL change was 0.001 for group i), 0.011 for group ii), and 0.082 for group iii), compared to -0.001, 0.035 and 0.059 on the CHU-9D. These change scores were significantly different according to the ADQoL (3.27, p<0.01) and CHU-9D (4.44, p<0.01).

CONCLUSIONS: This is the first study to examine comparative performance of these two instruments in eczema. It suggests that the ADQoL and CHU-9D were largely comparable for the criteria assessed. Further research is needed to extend the criteria assessed and to corroborate the findings in other samples.