1000 Section - Exploring the integration of knee-ankle-foot-orthosis (KAFO) and robot-assisted gait training (RAGT) in real-world stroke rehabilitation: insights from a retrospective cohort

Both knee-ankle-foot orthoses (KAFO) and robot-assisted gait training (RAGT) improve gait after stroke. This study aimed to clarify real-world usage patterns of KAFOs and RAGT in post-stroke rehabilitation.

Design:

This is a single-center retrospective cohort study in a post-acute rehabilitation hospital. Stroke patients admitted between August 2018 and February 2025 who underwent more than five sessions of both KAFO training and RAGT were included. Demographic, clinical, and functional data and KAFO/RAGT session details were extracted from medical records. We focused (i) the interval between KAFO and RAGT initiation, (ii) the duration of their overlap, and (iii) sequential or prolonged use patterns. We analyzed the relationships of clinical and functional parameters with these patterns.

Results:

A total of 136 patients (88 male, 48 female) were included, with a mean age of 64.0 ± 13.1 years. Stroke types were hemorrhagic (n = 81), ischemic (n = 52), and subarachnoid hemorrhage (n = 3). KAFO training and RAGT were initiated a mean of 5.6 ± 5.9 and 16.0 ± 14.9 days after admission, respectively, with durations of 50.5 ± 31.2 and 33.9 ± 18.6 days. The interval between them was 10.4 ± 16.1 days and was negatively correlated with motor FIM score at admission (ρ = –0.42, p < 0.01), but not with lower-limb or trunk motor function scales. Concurrent use was significantly prolonged in patients with more severe leg paresis (β = –1.41, p = 0.031) and poorer balance function (β = –0.55, p = 0.023), while shorter overlap was associated with earlier KAFO termination (ρ = –0.421, p < 0.001).

Conclusion: KAFOs are typically introduced before RAGT. The overlap period tended to be longer in patients with more severe impairments and shorter in those who achieved KAFO goals early.