EuroQol EQ-5D may not adequaltely describe the health of people with disabilities

PURPOSE: To assess whether life-circumstances and factors other than those described in the EuroQol EQ-5D instrument have a systematic influence on how the same EuroQol health-state is valued. METHOD: A simulation exercise whereby health professionals were asked to rate the health-state of 16 case-scenarios. Each case-scenario was designed to describe the same EuroQol health-state but was varied systematically using a 4 by 2 factorial design, in terms of ability to work, ability to perform usual leisure activity, age and type of disability. A convenience sample of 41 health professionals studying towards a postgraduate Diploma in Rehabilitation formed the study group. RESULTS: The average valuation was significantly higher than that derived from a social tariff model. The following factors contributed to a model which explained 10% of the variability in valuation scores: characteristics of the assessor--years of experience, experience with musculoskeletal disorders, and characteristics of the scenario--disability type, ability to work and ability to perform usual leisure activities (beta weights ranged from 0.093 to 0.253). Ability to perform usual leisure activities was associated with the greatest influence on the valuation rating. Additionally, professional type had a significant influence on rating with general practitioners giving significantly lower valuations. CONCLUSIONS: The same health-state, as defined by EuroQol EQ-5D cannot be regarded as describing similar life-circumstances, sufficient to be valued in the same way. Factors specific to the 'valuer' (experience and professional type) and specific to the 'valuee' (disability type, ability to work or perform usual leisure activities) that are not included in the EQ-5D description are significantly associated with different valuation scores. Caution should be used in applying valuations obtained from social tariff models to a disabled populatio