Reliability and validity study of Silver's test for Hong Kong-Chinese with dementia

There has been great concern regarding the functional and cognitive abilities of elderly people in performing activities of daily living. 'Silver's test' has been widely used to assess the abilities and mental functions of the aging population in Hong Kong. The purpose of this study was to establish the validity and reliability of a Chinese version of Silver's test (CVST) for Hong Kong-Chinese with dementia. A sample of 78 (21 male and 57 female) elderly subjects was recruited through two-stage cluster sampling. They were administered both the CVST and the Chinese Mini-Mental Status Examination (MMSE). Spearman rank-order correlation coefficient (rs) and intra-class correlation coefficient (ICC) showed that the CVST had a very good inter-rater reliability (rs=0.9573, P=0.000; ICC=0.9966) and test-retest reliability (rs=0.9693, P<0.001; ICC=0.9769). The internal consistency of the test was satisfactory (Cronbach's alpha=0.8396). The Mann-Whitney U-test showed that the CVST could discriminate the demented elderly subjects from the non-demented subjects, suggesting the best cut-off score to be 34, yielding a high sensitivity of 96.4% and equally good specificity of 96.7%. There was also a high positive correlation (rs=0.9218, P<0.001) between the total scores of the CVST and the MMSE. Suggestions for the modification of Silver's test were also made. In conclusion, the initial reliability and validity of the CVST has been established for Hong Kong-Chinese with dementia. It appears to be a sensitive and practical cognitive assessment acceptable to this population.