| Summary: | Background <br/>Stroke-induced cognitive impairments are critical predictors of poor functional outcomes. They adversely affect recovery and reduce independent performance of basic activities of daily living (ADL) and instrumental ADL (Zinn et al., 2004). Choices of cognitive assessment tools specific to the Cantonese speaking stroke population in Hong Kong are limited. The Cantonese version of the Western Aphasia Battery (Cantonese-WAB) was specifically developed for examining language impairments. The Cantonese version of MMSE (Cantonese-MMSE) and Hong Kong Montreal Cognitive Assessment (HK-MoCA), designed to detect cognitive deficits associated with dementia, lacked important measures of writing, neglect, and praxis where impairments were commonly found in stroke. More critically, most tasks in these two screeners required relatively intact auditory comprehension and verbal responses from participants. Presence of aphasia can, therefore, lead to underestimation of cognitive abilities. <br/> <br/>Aims <br/>Extending Chan et al.’s (2013) development of a Cantonese version of the Birmingham Cognitive Screen (BCoS) to be used in Hong Kong, our first aim was to validate the Oxford Cognitive Screen (OCS), built on similar principles to the BCoS test but is shorter (15 minutes) and can be used in acute settings, for Cantonese-speaking stroke survivors. This tool, including assessment of aphasia, apraxia, attention, memory, and spatial neglect, was designed to be neglect- and aphasia-friendly by using multi-modal presentation, forced-choice testing procedures, and vertical layouts. The second aim was to determine which cognitive domain(s) in HK-OCS would best predict functional outcomes. <br/> <br/>Procedures <br/>Seventy normal individuals were recruited to establish the normative data of HK-OCS. Norm was developed for three age groups (< 50, 50-59, and >59 years). Direct percentile conversions for each sub-test scores were used and cut-off scores were set at the top 5th percentile. Forty six native Cantonese-speaking stroke participants (aged 21-83 years; 72% left hemispheric stroke) participated and received a short test of gestural production (Goldenberg, 1996), Albert’s test of visual neglect, Chinese version of the Modified Barthel Index, and Hong Kong Chinese version of Lawton Instrumental Activities of Daily Living scale, Cantonese-WAB, Cantonese-MMSE, HK-MoCA, and HK-OCS. <br/> <br/>Results <br/>Cut-off scores for most HK-OCS sections were comparable across the three age groups. Paired-sample t-tests revealed the stroke group achieved significantly lower scores across most sections than matched controls. The tool demonstrated strong concurrent validity with positive and significant correlations across related cognitive subtests in other batteries (Table 1). The HK-OCS also had excellent intra-rater and inter-rater, good test-retest reliability, and acceptable internal consistency. Regarding its predictive values on functional outcomes, while semantics and episodic memory best predicted basic ADL, number writing and orientation best predicted complex ADL. <br/>
|
|---|
