Objective: In a pilot study of patients with a first stroke and hemiparesis, we sought to determine whether treatment of the upper limb with continuous passive motion (CPM) that was device delivered would alter impairment, disability or the associated adverse symptoms of shoulder joint instability, pain and tone.

Design: Patients were randomly assigned to receive daily CPM treatments or participate in self-range of motion groups under the supervision of an occupational therapist. All patients received standard daily poststroke therapy for 3.5 h per day. A blinded evaluator at admission and discharge assessed patients using standardized scales of impairment, disability and adverse symptoms.

Setting: Specialized stroke unit of an acute rehabilitation hospital.

Subjects: Two hundred and eighty consecutive patients were screened and 35 of these with a first unilateral stroke, 13 +/- 6 days following the acute event, provided informed consent and were randomly assigned to CPM treatment or supervised group self-range exercise.

Main measures: Thirty-two completed the study and were evaluated using standardized measures for motor impairment (Fugl-Meyer, Motor Status Scale and Medical Research Council Motor Power), adverse symptoms (gleno-humeral stability, pain and tone), and disability (Functional Independence Measure).

Results: CPM-treated patients demonstrated positive trends towards improved shoulder joint stability (p = 0.06, confidence interval -0.03, 2.3) when compared with patients performing therapist-supervised self-range of motion. There were no significant differences in motor impairment, disability, pain or tone.

Conclusions: Device-delivered continuous passive range of motion may offer an enhanced benefit for some adverse symptom reduction in the hemiplegic arm after stroke over traditional self-range of motion exercise.