To examine the test-retest reliability and validity of the Norwegian Brief Agitation Rating Scale (BARS), a short form of the Cohen-Mansfield Agitation Inventory (CMAI) assessing the frequency of agitation in dementia.
We investigated the internal consistency, test-retest reliability and the validity of BARS. In the validity study, we compared the BARS scores with the Neuropsychiatric Inventory--Nursing Home Version subscale Agitation/Aggression (NPI-NH/AA) and the Cornell Scale for Depression in Dementia subscale Agitation (CSDD/A).
In the reliability study, Cronbach's alpha was 0.76; the test-retest reliability of the BARS showed a Spearman's rho of 0.64, but this increased to 0.86 when we deleted the item 'complaining'. In the validation study, the BARS score correlated with the NPI-NH/AA and the CSDD/A scores, Spearman's rho 0.55 and 0.52, respectively. These correlations changed when controlling for the Clinical Dementia Rating (CDR) Scale stages. The highest correlations between the BARS and the NPI-NH/AA and the BARS and the CSDD/A were found among patients with CDR score 2.
The study indicates that the Norwegian version of BARS is a reliable and valid instrument to test agitation in dementia, but a version without the item 'complaining' would be better.