Validation of the Chinese Revised Algase Wandering Scale-Community Version for persons with dementia in northern...

Wandering in persons with dementia is perceived as significant due to its prevalence and negative outcomes. However, lack of a validated wandering measure in Taiwan has limited scientific investigation and clinical practice. Therefore, the purpose of this study was to validate the Chinese Revised Algase Wandering Scale-Community Version (CRAWS-CV) in northern Taiwan.

For this cross-sectional study, the sample comprised 180 elders with dementia and their family caregivers (FCs). FCs responded to the CRAWS-CV in interviews with trained research assistants.

The structure of CRAWS-CV was examined by exploratory principal component analysis with varimax rotation. This analysis derived nine factors, explaining 71.48% of variance: eloping behavior (EB), mealtime impulsivity/temporal aspects, getting lost inside the house (GLI), pacing, impulsivity, negative outcomes, random pattern (RANDOM), and getting lost outside. The total scale and subscales showed excellent internal consistency. Good construct validity was shown by significant inter-scale correlation coefficients, and significant correlations between scores on the total CRAWS-CV and its subscales with scores on the Mini-Mental State Examination and Chinese Neuropsychiatric Inventory. The receiver operating characteristic curve showed a cutoff score of 67, with sensitivity and specificity of 83.6% and 76.9%, respectively. CRAWS-CV scores were significantly different for wanderers and non-wanderers. The one-week test-retest reliability using intra-class correlation coefficients (ICCs) showed significant correlations except for the EB and RANDOM subscales. Inter-rater reliability using an ICC was significant and acceptable except for GLI.

This study supports the CRAWS-CV as a valid measure of wandering in community-dwelling elders with dementia in northern Taiwan.