Revista Panamericana de Salud Pública
Print version ISSN 1020-4989
SALIBA, Viviane Amaral et al. Cross-cultural adaptation and analysis of the psychometric properties of the Brazilian version of the Motor Activity Log. Rev Panam Salud Publica [online]. 2011, vol.30, n.3, pp. 262-271. ISSN 1020-4989. http://dx.doi.org/10.1590/S1020-49892011000900011.
OBJECTIVE: To describe the translation into Brazilian Portuguese and cross-cultural adaptation of the Motor Activity Log (MAL), an instrument specifically designed to assess function of the more severely affected upper limb in hemiplegics. METHODS: The MAL was translated and adapted according to standardized procedures and submitted to test-retest reliability assessment (intraclass correlation coefficient, ICC). Other psychometric properties were investigated using Rasch analysis in 77 chronic hemiplegics (41 men, mean age = 57.5 ± 12.4 years). RESULTS: An excellent ICC (0.98) was obtained for the total scores of both the quantity and quality MAL scales. When assessing degree of difficulty, "using a key to unlock the door" was the most difficult item on the quantity scale, whereas "washing hands" was the easiest one. On the quality scale, the most difficult item was "using the TV remote control," and the easiest one was "drying hands." The analyses showed that the set of items as a whole fit into the model; however, the individual analyses indicated that four items did not meet the expectations of the model in both the quantity (items 21, 16, 14, and 13) and quality (items 9, 21, 23, and 22) scales. Irregular response patterns were observed, and five subjects obtained the minimum score. There was disagreement between item difficulty and sample ability, suggesting that the abilities of individuals were below the degree of difficulty of the assessed items. A significant correlation was observed between grip strength of the more severely affected upper limb and motor skill measurement among individuals on the quantity (r = 0.51, P < 0.0001) and quality scales (r = 0.57, P < 0.0001), and also between the two scales when measuring the individuals' motor skills (r = 0.97, P < 0.0001). CONCLUSIONS: MAL-Brazil is potentially useful to evaluate the more severely affected upper limb in Brazilian patients with chronic hemiplegia. However, the instrument has limitations for use with individuals with severe limb impairments. Also, construct validity was affected by the presence of irregular score patterns. MAL-Brazil should be applied to additional samples to further investigate its validity.
Keywords : Stroke; psychometrics; upper extremity; validity of tests; reproducibility of results.