Tests of motor function in patients suspected of having mild unilateral cerebral lesions.

Abstract

Though various textbooks describe clinical manoeuvres that help detect subtle motor deficits, their sensitivity, specificity and predictive values have not been determined. We investigated the sensitivity, specificity and predictive values of various manoeuvres in order to determine the most sensitive and reliable test or combination thereof. Straight arm raising (Barré), pronator drift, Mingazzini's manoeuvre, finger tap, forearm roll, segmental strength and deep tendon reflexes were tested in 170 patients with (86) and without (84) a proven lesion in the motor areas confirmed by computed tomography. Segmental motor strength bad good specificity (97.5%) but poor sensitivity (38.9%) and negative predictive value (NPV) (58.7%). The forearm roll had a similar profile. Finger tap had a sensitivity of 73.3% and a specificity of 87.5%. Barré and pronator testing had a sensitivity and specificity of 92.2% and 90.0% respectively. Hyperreflexia had a sensitivity of 68.9% and a specificity of 87.5%. An abnormality of pronator, reflexes or finger tap had a sensitivity of 97%, and when these three tests were positive, specificity was 97%. When all six tests were positive, the positive predictive value was 100%, when all six tests were negative the NPV was 100%. The detailed segmental examination has very good specificity for detecting motor deficits, but the sensitivity and NPV are unacceptably low. Pronator drift with finger tap and reflexes is the most reliable and time-effective combination of tests for the detection of subtle motor lesions, and could replace the segmental motor examination as a screening for motor lesions.