Effect of robotic-assisted gait training on objective biomechanical measures of gait in persons post-stroke: a systematic review and meta-analysis

Table 2 A summary of GRADE domains and overall certainty of evidence for each outcome of interest

Outcome	Certainty assessment							No of participants		Effect	Certainty
Outcome	№ of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	RAGT	non-RAGT	Absolute (95% CI)	Certainty
Gait speed	8	Randomised trials	Serious^a,b,c,d	Not serious	Serious^e	Serious^f	None	181	175	MD 0 m/s (0.05 lower to 0.05 higher)	⨁◯◯◯ Very low
Cadence	7	Randomised trials	Serious^a,b,c,d	Not serious	Serious^e	Serious^f	None	168	151	MD 1.44 steps/min higher (2.34 lower to 5.22 higher)	⨁◯◯◯ Very low
Other temporal outcomes	6	Randomised trials	Serious^a,b,c	Serious^g	Serious^e	Serious^f	None	104	112	see comment^	⨁◯◯◯ Very low
Step length	3	Randomised trials	Serious^a,b,c,d	Not serious	Serious^e	Serious^f	None	88	87	MD 1.22 cm higher (0.1 lower to 2.54 higher)	⨁◯◯◯ Very low
Stride length	5	Randomised trials	Serious^a,b,c	Not serious	Serious^e	Serious^f	None	82	79	MD 2.86 cm higher* (0.46 higher to 5.25 higher)	⨁◯◯◯ Very low
Temporal symmetry	4	Randomised trials	Serious^a,b,c	Not serious	Serious^e	Serious^f	None	76	92	MD 0.09 ratio higher* (0.04 higher to 0.15 higher)	⨁◯◯◯ Very low
Spatial symmetry	5	Randomised trials	Serious^a,b,c,d	Not serious	Serious^e	Serious^f	None	149	146	MD 0.01 ratio lower (0.06 lower to 0.04 higher)	⨁◯◯◯ Very low
Kinematics	3	Randomised trials	Serious^a,b,c,d	Not serious	Not serious	Serious^f	None	16	15	See comment^	⨁⨁◯◯ Low

Table generated with the GRADEpro Guideline Development Tool. McMaster University, 2015 [developed by Evidence Prime, Inc.]
RAGT robotic-assisted gait training, non-RAGT non-robotic gait training, CI confidence interval, MD mean difference
*Indicates significant differences (p < 0.05) between groups favouring RAGT
^Statistical pooling was not possible for these variables and the findings are therefore presented in a narrative form in the text
^aIncluding at least one study with unclear handling of missing data
^bIncluding at least one study with risk for selective outcome reporting
^cIncluding one or several studies with overall unclear risk of bias
^dIncluding study with high or unclear risk of bias arising from the randomization process
^eHeterogeneity in intervention settings and gait analysis
^fSmall total population size (< 400)
^gDowngraded by 1 due to inconsistency in findings across studies

ISSN: 1743-0003