Key factors for successful gait acquisition in individuals with severe hemiplegic subacute stroke using a single-legged gait-training assistance robot: a retrospective cohort study.

This was a single-center retrospective study, so the results may not apply to other hospitals or patient groups. There was no control group, so the findings cannot show that Welwalk caused the outcome or separate the robot from the rest of rehabilitation. All participants received a package that included Welwalk, lower-limb orthoses, physical therapy, occupational therapy, speech therapy, and other conventional treatments, so the contribution of any single component cannot be separated. The number and duration of sessions were controlled, but the content and intensity of training were individualized and not standardized. Outcomes were assessed only after 4 weeks, so longer training might produce different results. All participants started robot-assisted gait training within 3 months of stroke, so the study could not determine the best timing within that window. The main causal method was designed for continuous outcomes and was adapted to a yes/no outcome. The causal interpretation should therefore be cautious. The mixed-data sensitivity analysis did not show joint position sense as a direct predictor, so the proprioception finding should be read cautiously.

The paper states that the funder was Daiwa Securities Foundation. The supplied text does not report authors' conflict-of-interest declarations.

Do not read these results as proof that Welwalk alone caused walking achievement. Every participant received a combined rehabilitation program, and the study had no control group, so it can only show associations between baseline factors and the outcome.