Randomised controlled trials (RCTs) or quasi-RCTs that compared mechanically assisted walking with body weight support to assisted overground walking (over a period longer than 15 minutes) were eligible for inclusion in the review. Trials had to report on independent walking, Eligible participants were adults (over 18 years old) who had a stroke (>24 hours) or were subacute (stroke less than three months prior to the trial) and were undergoing inpatient stroke rehabilitation. Participants also had to be non-ambulatory as measured by the Functional Ambulatory Category (FAC<3), the Motor Arm Scale (MAS<3) or the Functional Independence Measure (FIM<5); if populations included mixed ambulatory and non-ambulatory patients, trials had to report separate data for non-ambulatory patients. Further information specifying types of eligible interventions and outcomes (primary and secondary) were reported in the review.
Included trials compared mechanically assisted walking (using treadmill with a harness, or robotic device and harness, or electro-mechanical gait trainer with harness). In most trials, the intervention group also received some degree of overground assisted walking. The control group in all of the trials received overground walking assisted by therapists. Training lasted from 20 to 80 minutes per day, from three to five days per week for four to six weeks, or until discharge from inpatient rehabilitation.
The mean age of included patients ranged from 57 to 73 years; the average time post-stroke was up to one month. All but one of the trials defined non-ambulatory as FAC<3; the remaining trial defined non-ambulatory as MAS<2. Details of the included outcome measures were also reported in the review.
Two reviewers independently assessed the studies for inclusion; discrepancies were resolved by consensus after discussion with a third reviewer.