Engineering and Technology
SANAZ POURNAJAF, PhD
Senior Rehabilitation Scientist - Research Area Coordinator
IRCCS San Raffaele Roma
Rome, Lazio, Italy
Giovanni Morone, MD, PhD
Associate Professor
University of L'Aquila
L'Aquila, Abruzzi, Italy
Marco Iosa, PhD
Associate Professor
Sapienza University
Rome, Lazio, Italy
Sofia Straudi, Prof., PhD
Associate Professor
Ferrara University
Ferrara, Emilia-Romagna, Italy
andrea Santamato, Prof., MD
Full Professor
Foggia University
Foggia, Puglia, Italy
Rocco Salvatore Calabrò, MD
Senior Rehabilitation Scientist and Research Area Lead
IRCCS San Raffaele Roma
Messina, Sicilia, Italy
Marco Franceschini, Prof., MD
Research Area Head
IRCCS San Raffaele Roma
Rome, Lazio, Italy
Robot-assisted Arm Therapy (RAT) is a well-established approach to support post-stroke motor recovery. However, evidence on three-dimensional (3-D) exoskeleton-based RAT remains limited, as most available studies involve planar 2-D end-effector systems. This multicentre RCT evaluated the efficacy of 3-D exoskeleton-based RAT as an add-on intervention to usual rehabilitation care in subacute stroke inpatients. In this single-blind RCT, individuals with moderate-to-severe upper extremity (UE) motor impairment due to subacute stroke (< 3 months from onset) were randomly assigned (1:1) to receive 25 sessions (5/week for 5 weeks) of 3-D exoskeleton-RAT or conventional upper limb therapy. The primary outcome was the change in the Fugl–Meyer Assessment for Upper Extremity from baseline to end of treatment. A 3-month telephonic follow-up was performed. The odds ratio for achieving the minimal clinically important difference (MCID) was calculated. ClinicalTrials.gov NCT04697368. This study demonstrates the significant clinical benefit of 3-D exoskeleton-based RAT during the early post-stroke rehabilitation window, with markedly higher MCID rates than conventional therapy. Further research is needed to assess long-term outcomes and optimize protocols according to individual patient characteristics.
Design:
Results: A total of 109 individuals (35% women; mean age 62.5 ± 13.5 years; mean time from onset 34 ± 28 days) were randomized, with 82 completing the intervention. The Robotic Group (RG) showed significantly greater improvement in the primary outcome compared to the Control Group (CG), with a median between-group difference of 22 points (p < 0.001). Clinically meaningful gains (MCID ≥10 points) were achieved in 68.4% of the RG versus 31.8% of the CG, yielding an odds ratio of 4.64 (95% CI: 1.83–11.8; p < 0.001). Secondary outcomes favored the RG, although spasticity and dexterity measures did not show significant differences.
Conclusion:
