Validation of the Hong Kong version of Abe's behavioural and psychological symptoms of dementia score among care home residents with dementia.

Objectives: To translate and validate the Hong Kong version of Abe's behavioural and psychological symptoms of dementia (BPSD) score (HK-ABS) and examine its psychometric properties among older adults with dementia in Hong Kong care homes, and to examine correlations among BPSD items, patient and caregiver characteristics, and caregiver distress.

Methods: The content validity of the HK-ABS was assessed by a panel of three psychiatrists, one psychiatric nurse, one occupational therapist, and one medical social worker. Concurrent and discriminant validity of the HK-ABS were assessed using the Chinese or Hong Kong version of the Neuropsychiatric Inventory-Nursing Home (CNPI-NH), the Disability Assessment for Dementia, the Clinical Dementia Rating, and the Montreal Cognitive Assessment. Test-retest reliability and inter-rater reliability were determined. Exploratory factor analysis examined possible constructs by principal component analysis with varimax rotation. Factor extraction was based on Kaiser's eigenvalue criterion. Correlations between the HK-ABS and characteristics of patients and staff caregivers were examined.

Results: In total, 109 pairs of patients with major neurocognitive disorder and staff caregivers in care homes were included in the analysis. The median time to complete the HK-ABS was 70 seconds. The median HK-ABS score was 4. Internal consistency of the HK-ABS was good (Cronbach's alpha = 0.787); item-total correlations (excluding 'apathy and indifference') ranged from 0.337 to 0.653. The test-retest reliability (r_s = 0.981, p = 0.168) and inter-rater reliability (r_s = 0.987, p = 0.414) of the HK-ABS was excellent. For concurrent validity, the HK-ABS was highly correlated with the CNPI-NH (r_s = 0.857, p < 0.001). For discriminant validity, the HK-ABS was not correlated with the Chinese or Hong Kong version of the Montreal Cognitive Assessment (r_s = 0.103), the Disability Assessment for Dementia (r_s = -0.039), or the Clinical Dementia Rating (r_s = 0.067). In exploratory factor analysis, all items achieved factor loadings of >0.4. Three factors explained 70% of the variance. The HK-ABS was negatively correlated with work experience (r_s = -0.254, p = 0.008) and positively correlated with the length of hospital stay in the previous year (r_s = 0.193, p = 0.044) and the CNPI-NH occupational disruptive score (r_s = 0.759, p < 0.001).

Conclusion: The HK-ABS is a valid and reliable global assessment tool of BPSD for residential care settings and clinical communication. It may be used to identify caregiver distress to allow for early intervention.