RESUMO
Objective: To compare the effects of 2 types of robotic-assisted gait training (RAGT) devices that have been used in stroke rehabilitation. Design: Retrospective cohort. Setting: Rehabilitation hospital. Participants: 24 community dwelling people with stroke (N=24). Interventions: RAGT with either an exoskeleton (Lokomat) (mean age=53.8 years; 30% men; mean duration of stroke =17.8 months) or an end-effector (G-EO) (mean age=50.5 years; 77.8% men; mean duration of stroke =13.11) delivered 3 times per week (36 sessions total). Main Outcome Measures: The following tests/scales were employed before and after RAGT: Functional Ambulation Categories (FACs), timed Up and Go (TUG), 10-Meter Walk Test (10MWT), 6-Minute Walk Test (6MWT), Trunk Impairment Scale, Dynamic Gait Index (DGI), Berg Balance Scale (BBS), and ability to climb stairs (time to climb 6 steps of 15 cm each; ability to climb stairs). Results: There were 5 dropouts, all from the G-EO group. At the end, 10 participants in the Lokomat and 9 in the G-EO group completed the intervention. From pre- to post-RAGT, G-EO patients improved on all functional tests/scales, whereas Lokomat patients improved only on the TUG, DGI, and BBS. Most patients showed improvements above the relative smallest real difference in the TUG, 10MWT, and 6MWT. Conclusions: Both end-effectors and exoskeletons may improve clinically relevant aspects of walking function. However, this study had a small sample, was retrospective, non-randomized, and had a significant number of drop-outs, therefore its findings should be interpreted carefully. Future studies are needed for investigating potential differences in clinical results, side effects, contraindications, and cost effectiveness between these 2 different types of RAGT.