• Gary Naglie, George Tomlinson, Catherine Tansey, Jane Irvine, Paul Ritvo, Sandra E Black, Morris Freedman, Michel Silberfeld, and Murray Krahn.
• Division of General Internal Medicine and Clinical Epidemiology, University Health Network, Toronto, Ontario, Canada. naglie.gary@tornotorehab.on.ca
• Qual Life Res. 2006 May 1;15(4):631-43.

ObjectivesTo explore whether Alzheimer's disease patients could rate their quality of life (QOL) using utility-based health indexes, and to provide new knowledge about the measurement properties of these instruments for patient and caregiver proxy ratings.MethodsA convenience sample of 60 mild-moderate AD patients and their caregivers were randomized to complete the Quality of Well-Being Scale (QWB), Health Utilities Index (HUI3) or EQ-5D and visual analogue scale (VAS) on two occasions. Test-retest reliability (intraclass correlation coefficients) and convergent validity (Spearman correlations) of utility scores with global health status, activities of daily living, comorbidity, mood, cognition and other utility measures were assessed.ResultsCompletion time was shortest for the combined EQ-5D and VAS. For patients with mild dementia and for proxies, reliability was >or= 0.70 for the EQ-5D, QWB and HUI3. The EQ-5D had a ceiling effect for patient ratings. Convergent validity was demonstrated for patient and proxy ratings, with the strongest validity for EQ-5D ratings and the weakest validity for HUI3 patient ratings. Mean patient utility scores were significantly higher than mean proxy scores for all measures (p<0.001).ConclusionsFor patient and proxy ratings, the EQ-5D had the best combination of measurement properties, although it had a substantial ceiling effect for patient ratings. Proxy QOL ratings did not accurately reflect patients' ratings.

23 keywords...

hide…