Purpose: There are many validated quality-of-life (QoL) measures designed for people living with dementia. However, the majority of these are completed via proxy-report, despite indications from community-based studies that consistency between proxy-reporting and self-reporting is limited. The aim of this study was to understand the relationship between self- and proxy-reporting of one generic and three disease-specific quality-of-life measures in people living with dementia in care home settings.
Methods: As part of a randomised controlled trial, four quality-of-life measures (DEMQOL, EQ-5D-5L, QOL-AD and QUALID) were completed by people living with dementia, their friends or relatives or care staff proxies. Data were collected from 726 people living with dementia living in 50 care homes within England. Analyses were conducted to establish the internal consistency of each measure, and inter-rater reliability and correlation between the measures.
Results: Residents rated their quality of life higher than both relatives and staff on the EQ-5D-5L. The magnitude of correlations varied greatly, with the strongest correlations between EQ-5D-5L relative proxy and staff proxy. Internal consistency varied greatly between measures, although they seemed to be stable across types of participants. There was poor-to-fair inter-rater reliability on all measures between the different raters.
Discussion: There are large differences in how QoL is rated by people living with dementia, their relatives and care staff. These inconsistencies need to be considered when selecting measures and reporters within dementia research.
Keywords: Care homes; Dementia; Inter-rater reliability; Proxy-report.