Abstract
BACKGROUND: The Quality of Life in Alzheimer's disease scale (QoL-AD) is a widely used Health Related Quality of Life (HRQoL) instrument. However, studies investigating the instrument's inter-rater reliability (IRR) are missing. This study aimed to determine the item distribution and IRR of the German proxy version of the QoL-AD (13 Items) and a nursing home-specific instrument version (QoL-AD NH, 15 Items). METHODS: The instruments were applied to 73 people with dementia living in eight nursing homes in Germany. Individuals with dementia were assessed two times by blinded proxy raters. The IRR analyses were based on methodological criteria of the quality appraisal tool for studies of diagnostic reliability (QAREL), the COSMIN group and the single-measure Intra-Class Correlation Coefficient (ICC) for absolute agreement ≥0.70. RESULTS: All items for both instrument versions demonstrated acceptable item difficulty, with the exception of one item (QoL-AD proxy). The IRR was moderate for the QoL-AD (ICC: 0.65) and insufficient for the QoL-AD NH (ICC: 0.18). The additional computation of the average measure ICC for two proxy-raters demonstrated a strong IRR (ICC: 0.79) for the QoL-AD and a weak IRR for the QoL-AD NH (ICC: 0.31). The detailed analysis of the IRR for each item underpinned the need for the further development of both instruments. CONCLUSIONS: The unsatisfactory IRRs for both instruments highlight the need for the development of a user guide including general instructions for instrument application as well as definitions and examples reflecting item meaning. Priority should be given to the development of reliable proxy-person versions of both instruments. TRIAL REGISTRATION: ClinicalTrials.gov: NCT02295462 , Date of registration: 11-20-2014.