Rehabilitation robotics has been proposed as an efficient strategy to improve upper extremity (UE) motor function after stroke especially at the subacute phase. Many trials with several robots have been performed but very few large multicenter RCTs at the subacute phase.

A multicenter parallel group, two arm, single blind, phase III, superiority, randomized, controlled trial has been conducted in France in 22 rehabilitation centers equipped with an Armeo Spring robot. Two hundred and twenty subjects must be included. The main inclusion criteria were an UE Fugl Meyer (UEFM) score between 10 and 40 and an occurrence of the stroke between 3 and 12 weeks before inclusion. The main outcome measure was the UE FM 30 days. Each patient performed usual treatment (2hours a day of motor rehabilitation) with 20 additional rehabilitation sessions (1h each, 5 days a week, 4 weeks). Patients were randomized either to rehabilitation robotic sessions with the Armeo Spring device or to self-rehabilitation (stretching and self paced movements) in equal timed sessions. Secondary outcome measures were, pain, hypertonia, functional improvement (FIM and Action research arm test ARAT) and quality of life (EQ5D and stroke impact scale). Evaluation were performed at day 30, 3, 6 and 12 months after inclusion.

Two hundred and eighteen subjects have been included and an attrition rate of 10% at 6 months and 20% at 12 months was observed. The gain in the main outcome criteria was not significantly different at day 30 nor any other secondary criteria except ARAT.

An additional time 1h a day during 4 weeks of rehabilitation robotics to usual intensive rehabilitation seems not significantly influence the motor deficiency prognosis of the upper limb after stroke. It might influence functional outcome but this has to be explored in another design randomized control study.