Agreement between dementia patient report and proxy reports using the Nottingham Health Profile

Objective The aim of the study was to examine the agreement between patient reports and their proxy reports (family and care provider proxies) on Health Status in a sample of patients with dementia. Method Ninety-nine patients with mild to moderate dementia and proxies completed the 38-item Nottingham Health Profile (NHP) questionnaire. Results Completion rates for the different NHP dimensions ranged from 78 to 90% for the dementia subjects. Inter-rater agreement between different proxies and subject was from moderate to good for physical assessment (ICCs from 0.54 to 0.78 for physical mobility scales). Patient/family proxy concordance was moderate to good for five out of six dimensions (physical mobility, social isolation, pain, energy, sleep) and poor for emotional reaction. Family proxies systematically reported lower functioning than did patients in the four subscales assessing: physical mobility (p < 0.0001), energy (p < 0.005), social isolation (p < 0.01) and sleep (p < 0.03). Care provider proxies only estimated physical mobility as lower (p < 0.0001). Conclusion Age and physical status of the patient significantly affected agreement in patient-care provider proxy ratings. Thus, caution is appropriate when resorting to proxies to estimate the Health Status of a dementia patient. Copyright (C) 2004 John Wiley Sons, Ltd.