Estimating the minimal clinically important difference of the stroke rehabilitation assessment of movement measure

Background. The minimal clinically important difference (MCID) of the Stroke Rehabilitation Assessment of Movement (STREAM) measure is unknown, which limits the application and interpretation of change scores. Objective. To estimate the MCID of the 3 subscales (ie, lower extremity, upper extremity, and mobility subscales) of the STREAM using the method of patients' global ratings of change. Methods. Eighty-one stroke patients participated in this study. The patients' global ratings of change were used to rate their changes of limb movements and basic mobility skills on the 15-point Likert scale. The mean change scores on the 3 subscales of STREAM of the MCID group (ie, scored on +2 to +3 or -2 to -3) served as the estimates of the MCID. Results. There were 42, 38, and 43 patients in the MCID group, and the estimates of the MCID were 2.2, 1.9, and 4.8 points for the upper-extremity subscale, lower-extremity subscale, and mobility subscale, respectively. Conclusions. These findings suggest that if the mean change scores on the 3 subscales of the STREAM within a stroke group have reached 2.2, 1.9, and 4.8 points, the change scores on the 3 subscales of the STREAM can be perceived by patients as clinically important. Copyright ? 2008 The American Society of Neurorehabilitation.