Li 2014.

Methods	Randomisation via minimisation software Single‐blind ‐ assessors blinded No significant differences in baseline comparability tests in all groups of participants
Participants	Recruitment through newspaper advertisements and flyers in China 118 participants with dysphagia and hemispheric stroke
Interventions	Rx 1: neuromuscular electrical stimulation (VitalStim) Rx 2: combined NMES and traditional swallowing therapy C: traditional swallowing therapy (Data from Rx 2 vs control used in this review)
Outcomes	Swallow score, oral transit time, pharyngeal transit time, laryngeal closure duration, PAS
Notes	Exclusion: progressive stroke, other neurological disease, neoplastic disease, previous surgery to swallowing apparatus, nasogastric tube
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer randomisation via minimisation software
Allocation concealment (selection bias)	Low risk	Allocation concealed by sealed envelope
Blinding (performance bias and detection bias) All outcomes	Low risk	Outcome assessor blinded
Blinding of participants and personnel (performance bias) All outcomes	High risk	Participants and technicians not blinded
Blinding of outcome assessment (detection bias) All outcomes	Low risk	As above
Incomplete outcome data (attrition bias) All outcomes	High risk	17 participant dropouts
Selective reporting (reporting bias)	Low risk	All outcomes reported
Other bias	Low risk	None identified