Chua 2016.

Methods	RCT Method of randomisation: computer‐generated sequence Blinding of outcome assessors: stated as 'yes' by the investigator Adverse events: no Deaths: yes Dropouts: yes (7 in treatment group, 13 in control group) ITT: yes
Participants	Country: Singapore 108 participants (53 in treatment group, 53 in control group) Not ambulatory at start of study Mean age: 62 years experimental and 61 years control group Inclusion criteria: unilateral haemorrhagic/ischaemic stroke, age between 18 and 80 years, independent ambulation pre‐stroke Exclusion criteria: > 8 weeks poststroke, FAC ≥ 4, cardiovascular instability, MMSE < 16, communication deficits, lower limb joint contractures
Interventions	2 arms: Experimental group: electromechanical gait training (20 minutes) and conventional physiotherapy including stance/gait (25 minutes) 6 days a week for 8 weeks (45 minutes per day) Control group: conventional physiotherapy (including therapy to improve stance/gait), same time and frequency as experimental group
Outcomes	Outcomes were recorded at baseline and 4, 8, 12, 24, and 48 weeks after baseline Outcome measures: walking ability (FAC), Barthel Index, gait velocity (10‐metre walk test), gait endurance (6‐minute walk test), health status (Stroke Impact Scale)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Authors state: "Randomization was performed using a computer‐generated sequence of random numbers."
Allocation concealment (selection bias)	Low risk	Authors state: "An independent department generated the random group allocation sequence and transferred the sequence to a series of serially numbered opaque envelopes, which were not opened and revealed until after acceptance into the study and the baseline tests, therefore ensuring allocation concealment."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Authors state: "The data assessors were blinded to group allocation, but it was not possible to blind participants or the health care professionals providing interventions."
Incomplete outcome data (attrition bias) All outcomes	Low risk	All data for all participants provided and analysed. Authors state: "An intention‐to‐treat approach was used. Data from subjects were analysed according to the group to which they were randomised, regardless of whether they completed the intervention. Participants failing to complete either intervention were asked to return for follow‐up."