The unreliability of clinical measures of muscle tone: implications for stroke therapy

Background: the central tenet of the neurofacilitatory approach to stroke therapy is that muscle tone needs to be normal before normal movement can occur. A reliable clinical measure of the full spectrum of muscle tone is needed to test: (i) the purported relationship between muscle tone, other motor impairments and disability and (ii) the effectiveness of stroke therapy to restore movement. Aim: the purpose of the study was to test the inter-rater reliability of clinical categorization of muscle tone (spastic/normal/flaccid) and also a visual analogue scale with anchor points of 'lowest tone possible' (score 0) and 'highest tone possible' (score 100). Methods: four independent raters assessed tone of elbow flexors and knee extensors of 14 stroke rehabilitation inpatients using the categorical scale. Six independent raters assessed tone of elbow flexors and knee extensors of 25 chronic stroke patients and two healthy volunteers using the visual analogue scale. All assessment orders were randomized. Results: both scales were unreliable, with k coefficients for the categorical scale ranging from -0.046 to 0.56 for the categorical scale, and intraclass correlation coefficients for the visual analogue scale of 0.595 for elbow flexors and 0.451 for knee extensors. Assessment order effects for the visual analogue scale were non-significant for elbow flexors (P = 0.545) and knee extensors (P = 0.911). Conclusions: these results, and those of earlier studies, suggest that clinical measures of muscle tone are consistently unreliable. Systematic investigation of the therapy rationale for planning and evaluating treatment is required before relevant clinical measures can be developed.