Robotic systems for gait re-education in cases of spinal cord injury: a systematic review

The evidence underlying robotic body weight supported treadmill training in patients with spinal cord injury remains poorly characterized. To perform a qualitative systematic review on the efficacy of this therapy. A search on PubMed, CINAHL, Cochrane Library and PEDro was performed from January 2005 to April 2016. The references in these articles were also reviewed to find papers not identified with the initial search strategy. The methodological level of the articles was evaluated with PEDro and Downs and Black scales. A total of 129 potentially interesting articles were found, of which 10 fulfilled the inclusion criteria. Those studies included 286 patients, who were predominantly young and male. Most of them had an incomplete spinal cord injury and were classified as C or D in ASIA scale. Robotic devices employed in these studies were Lokomat, Gait Trainer and LOPES. Improvement in walking parameters evaluated was more evident in young patients, those with subacute spinal cord injury, and those with high ASIA or LEMS scores. Conversely, factors such as etiology, level of injury or sex were less predictive of improvement. The methodological level of these studies was fair according to PEDro and Downs and Black scales. The evidence of gait training with robotic devices in patients with spinal cord injury is positive, although limited and with fair methodological quality.