Tonic stretch reflex threshold as a measure of spasticity after stroke: Reliability, minimal detectable change and responsiveness

Objective: To determine inter-rater reliability, minimal detectable change and responsiveness of Tonic Stretch Reflex Threshold (TSRT) as a quantitative measure of elbow flexor spasticity. Methods: Elbow flexor spasticity was assessed in 55 patients with sub-acute stroke by determining TSRT, the angle of spasticity onset at rest (velocity = 0°/s). Elbow flexor muscles were stretched 20 times at different velocities. Dynamic stretch-reflex thresholds, the elbow angles corresponding to the onset of elbow flexor EMG at each velocity, were used for TSRT calculation. Spasticity was also measured with the Modified Ashworth Scale (MAS). In a sub-group of 44 subjects, TSRT and MAS were measured before and after two weeks of an upper-limb intervention. Results: The intraclass correlation coefficient was 0.65 and the 95% minimal detectable change was 32.4°. In the treated sub-group, TSRT, but not MAS significantly changed. TSRT effect size and standardized response mean were 0.40 and 0.35, respectively. Detection of clinically meaningful improvements in upper-limb motor impairment by TSRT change scores ranged from poor to excellent. Conclusions: Evaluation of stroke-related elbow flexor spasticity by TSRT has good inter-rater reliability. Test responsiveness is low, but better than that of the MAS. Significance: TSRT may be used to complement current scales of spasticity quantification.