Rethinking the tools in the toolbox.

The commentary by Dr. Labruyere on the article by Kuo et al. (J Neuroeng Rehabil. 2021; 18:174) posits that randomized trials evaluating the comparative efficacy of robotic devices for patients with neurological injury may not be needed. The primary argument is that researchers and clinicians do not know how to optimize training parameters to maximize the benefits of this therapy, and studies vary in how they deliver robotic-assisted training. While I concur with the suggestion that additional trials using robotic devices as therapeutic tools are not warranted, an alternative hypothesis is that future studies will yield similar equivocal results regardless of the training parameters used. Attempts are made to detail arguments supporting this premise, including the notion that the original rationale for providing robotic-assisted walking training, particularly with exoskeletal devices, was flawed and that the design of some of the more commonly used devices places inherent limitations on the ability to maximize neuromuscular demands during training. While these devices arrived nearly 20 years ago amid substantial enthusiasm, we have since learned valuable lessons from robotic-assisted and other rehabilitation studies on some of the critical parameters that influence neuromuscular and cardiovascular activity during locomotor training, and different strategies are now needed to optimize rehabilitation outcomes.