Reliability of the Modified Ashworth Scale After Stroke for 13 Muscle Groups

To evaluate the reliability of the protocol for administration of the Modified Ashworth Scale (MAS) for all commonly affected muscle groups after stroke.

A repeated-measures design was used in administration of MAS for 13 muscle groups on 2 assessment days. Intrarater reliability and interrater reliability (between 3 raters) was assessed.

Inpatient rehabilitation.

30 patients, 1-19 months after stroke (age 55.1±13.5 years; N=30).

Not applicable.

Intra- and interrater reliability of the muscle tone assessment protocol with MAS for 7 upper and 6 lower limb muscle groups compiled from previous studies; 1 modified and 4 originally described. The weighted kappa was calculated.

The most and the least frequently assigned MAS grades were 0 and 4, respectively. Agreement was the highest for grade 0 (49% within raters, 32% between raters). Intrarater reliability was good to excellent for upper limb (κ=0.71-0.94) and moderate to excellent for lower limb (κ=0.55-0.97) muscles. Interrater reliability was poor to good for upper limb (κ=0.25-0.66) and moderate for lower limb (κ=0.41-0.54) muscles.

The intrarater reliability of MAS was moderate for the hip flexors. The reliability results for the other 4 muscles studied anew after stroke were similar to the predetermined ones. The better intrarater reliability results confirmed previous findings. Because of the low interrater reliability, caution is needed in interpreting the results when reassessment is not possible by the same examiner. A well-described protocol for administering the MAS may lead to its standardization.