Robot-assisted gait training (RAGT) has been explored using various devices. Most studies have focused on bilateral-legged devices; however, few have evaluated single-legged devices. Therefore, we aimed to identify key factors associated with successful responders among individuals with severe hemiplegic subacute stroke who had ambulatory deficits and underwent gait training using a single-legged assistance robot. We retrospectively analyzed patients with stroke who received RAGT at Okayama Rehabilitation Hospital between February 2018 and September 2024. Eligible patients had severe hemiplegia after a first-time supratentorial stroke and required assistance for ambulation. Exclusion criteria included subarachnoid hemorrhage, concomitant neurological disease, or failure to complete 10 RAGT sessions. The primary outcome was achieving walking with supervision or better, rather than requiring physical assistance. Responders were defined as those who walked without physical assistance but required supervision. We explored factors affecting this outcome after 4 weeks of RAGT. Factor selection was performed using the Direct Linear Non-Gaussian Acyclic Model, followed by logistic regression to assess associations. A total of 126 participants were included. Stroke Impairment Assessment Set (SIAS) scores (knee extension, ankle dorsiflexion, joint position sense, trunk verticality, and abdominal strength), Functional Independence Measure scores (problem solving and memory), time from onset to RAGT initiation, and initial walking independence were associated with achieving walking with supervision. Among these factors, days to RAGT initiation (odds ratio OR = 0.19), SIAS trunk verticality (OR = 7.79), and SIAS joint position sense (OR = 3.37) were independently associated with the outcome. Achieving walking with supervision was associated with earlier initiation of RAGT, trunk control, and lower-limb proprioception in patients with severe hemiplegia undergoing gait training with a single-legged assistance robot. These findings highlight key factors for selecting candidates for single-legged gait-training robots and may guide more effective rehabilitation strategies. Trial Registration University Hospital Medical Information NetworkClinical Trials Registry UMIN000056551 January 4, 2025.
Asano et al. (Sat,) studied this question.