Objectives: to establish the reliability of the modified Ashworth scale for measuring muscle tone in a range of muscle groups (elbow, wrist, knee and ankle; flexors and extensors) and of the Medical Research Council scale for measuring muscle power in the same muscle groups and their direct antagonists.
Design: a cross-sectional study involving repeated measures by two raters. We estimated reliability using the kappa statistic with quadratic weights (Kw).
Setting: an acute stroke ward, a stroke rehabilitation unit and a continuing care facility.
Subjects: people admitted to hospital with an acute stroke-35 patients, median age 73 (interquartile range 65-80), 20 men and 15 women.
Results: inter- and intra-rater agreement for the measurement of power was good to very good for all tested muscle groups (Kw = 0.84-0.96, Kw = 0.70-0.96). Inter- and intra-rater agreement for the measurement of tone in the elbow, wrist and knee flexors was good to very good (Kw = 0.73-0.96, Kw = 0.77-0.94). Inter- and intra-rater agreement for the measurement of tone in the ankle plantarflexors was moderate to good (Kw = 0.45-0.51, Kw = 0.59-0.64).
Conclusions: the Medical Research Council scale was reliable in the tested muscle groups. The modified Ashworth scale demonstrated reliability in all tested muscle groups except the ankle plantarflexors. If reliable measurement of tone at the ankle is required for a specific purpose (e.g. to measure the effect of therapeutic intervention), further work will be necessary.