Is robotic gait training effective for individuals with cerebral palsy? A systematic review and meta-analysis of randomized controlled trials.

Only 188 participants in total, all with spastic diplegia or triplegia. The authors note the result may not extend to other CP subtypes (e.g., dyskinetic, hemiplegic) and that a small benefit is still possible. The meta-analysis could only pool outcomes measured in at least four studies. If assistive robotic training helps with an outcome that was not measured by enough trials, that benefit would be invisible here. Most underlying trials were rated Level 2b evidence, with small sample sizes and high dropout rates in the control groups. All devices were tethered, which the authors note limits how well the results translate to real-world, independent walking. Several studies did not report exactly how biofeedback or virtual-reality features of the devices were used during training, making it hard to judge whether those features contributed anything. Not all studies used the same device or protocol, though no significant heterogeneity was detected.

One author (ZFL) is a named inventor on a pending utility patent for a robotic device similar to those reviewed and is a co-founder of a company seeking to commercialize it. No funding source is named in the text provided.

Do not read this as 'robotic gait training is useless for cerebral palsy.' Every device tested was assistive, meaning the robot guided the legs and unloaded body weight while the patient went along. One RCT found that resistive training, which forces the child to generate the movement, was more effective than assistive training. The 188 participants were all spastic diplegia or triplegia, so the null result may not apply to other CP subtypes. The absence of a benefit for assistive devices does not rule out a benefit for a different robotic paradigm.