Validation of the UCSD Performance-Based Skills Assessment (UPSA) for functional evaluation of people with severe mental illness in Hong Kong.

BACKGROUND/HYPOTHESIS: This study aims to translate and adapt the University of California San Diego Performance-Based Skills Assessment Extended Range (UPSA-2-ER) to Chinese for assessing independent living skills of people with severe mental illness in Hong Kong.

STUDY

DESIGN

An expert panel was conducted to evaluate the translation, content validity, and cultural relevance. The UPSA-2ER (HK) was administered to 30 Hong Kong Chinese with severe mental illness (SMI) to examine the inter-rater reliability and 210 more to examine the internal consistency, structural validity and predictive validity. Sixty-six participants also completed the Chinese Functional Needs Assessment (CFNA) for concurrent validity.

STUDY

RESULTS

Good content validity was established with modifications of items to address cultural relevance. Both inter-rater reliability (ICC ranged from 0.98 to 0.99) and internal consistency (Cronbach's α = 0.97) of total score were excellent, while internal consistency of sub-domains ranged from fair to excellent (Cronbach's α ranged from 0.76 to 0.96). In Rasch analysis, raw variance explained by measures ≥40% was interpreted as good in terms of unidimensionality (Sumintono & Widhiarso, 2015), study results provided modest evidence that the UPSA-2ER(HK) has a unidimensional structure, as the variance explained by the items was 43.5% and the unexplained variance in the 1st contrast was 4.4%. Classification & regression tree analysis showed that the UPSA-2ER(HK) correctly classified 78.5% of the 210 participants into their residential status of independent living, halfway house or long stay care home. For concurrent validity, the UPSA-2ER(HK) demonstrated high correlation with the local golden standard, CFNA (r = 0.82).

CONCLUSIONS

The results showed that the instrument has satisfactory psychometric properties for assessing capacity to perform community living tasks.