Inter-rater Reliability of the Qol-measure QUALIDEM

Background

Improving access to health and nursing care aims to maintain and promote the quality of life (Qol) of care recipients. Especially in dementia research, Qol is a major outcome. QUALIDEM is an often-used dementia-specific Qol measure. It consists of two consecutive versions, one for mild to severe and one for very severe dementia. Previous results showed an insufficient inter-rater reliability (IRR).

Aims

Development of a user guide for the measurement and the evaluation of the IRR of QUALIDEM.

Methods

The study was conducted in two steps: (1) A QUALIDEM user guide was developed, based on 11 cognitive interviews with 16 nurses experienced in dementia care. (2) The item “Distribution” and the IRR of the QUALIDEM were evaluated in a field test including n=55 (mild to severe) and n=36 (very severe dementia) residents from nine nursing homes. The people with dementia were assessed four times by blinded proxy-raters. Proxy-raters were nurses and nursing assistants (n=40) who knew the people with dementia well. The proxy-ratings were led by a trained researcher and followed the user guide. The intra-class correlation coefficients (ICC) were calculated for each QUALIDEM subscale separately.

Results

The user guide includes definitions and examples for each item. The distribution of the responses of all 40 QUALIDEM items (n=40) was examined, indicating floor or ceiling effects for 13 items. All QUALIDEM subscales for people with mild to severe dementia showed a strong IRR (ICC ≥ 0.91). With the exception of one subscale (ICC: 0.64), the version for people with very severe dementia indicated a strong IRR (ICC ≥ 0.79) too. A second analysis excluding the items with floor and ceiling effects confirmed the strong IRR.

Conclusions

This study provides meaningful evidence for further development of the QUALIDEM and its future use as an outcome measure of dementia care intervention studies.