Modified Ashworth scale and spasm frequency score in spinal cord injury: reliability and correlation

Intra- and inter-rater reliability study.

To assess intra- and inter-rater reliability of the Modified Ashworth Scale (MAS) and Spasm Frequency Score (SFS) in lower extremities in a population of spinal cord-injured persons, as well as correlations between the two scales.

Clinic for Spinal Cord Injuries, Rigshospitalet, Hornbaek, Denmark.

Thirty-one persons participated in the study and were tested four times in total with MAS and SFS by three experienced raters. Cohen's kappa (κ), simple and quadratic weighted (nominal and ordinal scale level of measurement), was used as a measure of reliability and Spearman's rank correlation coefficient for correlation between MAS and SFS.

Neurological level ranged from C2 to L2 and American Spinal Injury Association impairment scale A to D. Time since injury was (mean±s.d.) 3.4±6.5 years. Age was 48.3±20.2 years. Cause of injury was traumatic in 55% and non-traumatic for 45% of the participants. Antispastic medication was used by 61%. MAS showed intra-rater κsimple=-0.11 to 0.46 and κweighted=-0.11 to 0.83. Inter-rater κsimple=-0.06 to 0.32 and κweighted=0.08 to 0.74. SFS showed intra-rater κweighted=0.94 and inter-rater κweighted=0.93. Correlation between MAS and SFS showed non-significant correlation coefficients from-0.11 to 0.90.

Reliability of MAS is highly affected by the weighting scheme. With a weighted-κ it was overall reliable and simple-κ overall unreliability. Repeated tests should always be performed by the same rater and in a very standardized manner. SFS was found reliable. MAS and SFS are poorly correlated, and ratings were inversely distributed and suggest that it assesses different aspects of spasticity.