Lessons learned from conducting a pragmatic, randomized, crossover trial on robot-assisted gait training in children with cerebral palsy (PeLoGAIT).

The trial stopped early with 16 children, although 30 patients were intended, so it had less power than planned. Children had to be able to walk at least 14 meters, so it says little about non-ambulatory children. Assessors were not blinded to the time point, which can affect subjective scoring. The primary GMFM dimension E score may not reflect individual family goals and may be insensitive for children at GMFCS levels III and IV. Half the children had previous robot training experience, but this could not be analyzed separately. Usual care varied by child and could include physiotherapy, occupational therapy, hippotherapy or circuit training, so the robot training was not compared with a fixed control. The second recruiting center enrolled no children.

The supplied text does not include a full conflict-of-interest or funding declaration. Publication types list non-U.S. government research support, and the sponsor, H.v.H., helped decide to close the trial early.

Do not read the null result as proof that robot-assisted gait training never helps children with cerebral palsy. The trial stopped early with 16 children, used 15 sessions over 5 weeks, and tested robot training as a single outpatient add-on rather than embedded in a broader multidisciplinary program.