Measurement properties of the Motor Evaluation Scale for Upper Extremity in Stroke patients (MESUPES).

PURPOSE To investigate inter-rater reliability of the Motor Evaluation Scale for Upper Extremity in Stroke patients (MESUPES), to provide estimates of the minimal detectable change (MDC) of the MESUPES and to investigate concurrent validity in relation to the arm scores of the Modified Motor Assessment Scale (M MAS). METHODS Forty-two stroke patients (mean age 56 ± 12 years) were independently assessed within a 48-hours window by two raters in different pairs (total available raters = 4). RESULTS Weighted κ analysis indicated good to very good agreement at item level (range 0.63-0.96). The relative and absolute reliability of the total score of MESUPES (maximum 58) was high according to the intraclass correlation coefficients (ICC = 0.98) and the standard error of measurement (SEM = 2.68). The MDC for three levels of confidence was calculated: A score change of 8, 7 and 5 is necessary for a MDC to have confidence of 95%, 90% and 80%, respectively, of a genuine change. Correlation between the MESUPES and M MAS was high (r(s) = 0.87). CONCLUSIONS The MESUPES shows high inter-rater reliability, and our study provides useful estimates of MDC for different levels of certainty. Additional research to confirm concurrent validity and to examine other psychometric properties of the MESUPES such as sensitivity is needed.