The hemiplegic arm: interrater reliability and concurrent validity of passive range of motion measurements

To assess whether our measurement protocol using two raters simultaneously yielded reliable passive range of motion measurements of the hemiplegic arm. Additionally, motion ranges were correlated to several factors to examine the concurrent validity of these measurements.

Two raters simultaneously assessed five arm motions at baseline, after five and ten weeks in respectively 18, 13 and 12 stroke patients. One tester made the passive movement and the other read the hydrogoniometer. Raters then switched roles.

Intraclass correlation coefficients revealed high agreement between the raters with intraclass correlation coefficients (ICCs) ranging between 0.84 and 0.99. Standard errors of measurement and smallest detectable differences were large for shoulder abduction. Significant correlations were found between shoulder external rotation and flexion. All arm motions correlated negatively to pain at the end range of these motions. Shoulder external rotation and flexion were significantly correlated to the time post stroke. Concurrent validity with Ashworth Scale, Fugl-Meyer Assessment and Barthel Index was limited.

The current measurement protocol yielded high reliability indices and seems useful for further use. However, standard error of measurement and smallest detectable difference for shoulder abduction were high, implying the necessity to include a large sample size in future studies. Correlations revealed that restricted range of arm motions relate to the time post-stroke and coincide with pain.