DEFUSE 3.

Study characteristics
Methods	Multicentre, randomised, open‐label trial with blinded outcome assessment
Participants	Patients with acute ischaemic stroke presenting between 6 and 16 hours from last known well and with remaining brain tissue that was not yet infarcted. Patients with proximal MCA or ICA occlusion, an initial infarct size of less than 70 mL, and a ratio of the volume of the ischaemic tissue on perfusion imaging to infarct volume of 1.8 or more
Interventions	Endovascular thrombectomy plus standard medical treatment versus standard medical treatment alone
Outcomes	Primary outcome was the ordinal score on the mRS at day 90 follow‐up.  Secondary outcome was functional independence mRS 0 to 2 at day 90. Primary safety outcome was death within 90 days and the occurrence of symptomatic intracerebral haemorrhage within 36 hours.
Notes	Funding source: National Institute of Neurological Disorders and Stroke
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐based, dynamic randomisation system. Stratified according to age, core infarct volume, time from symptom onset to enrolment, baseline NIHSS, and trial site
Allocation concealment (selection bias)	Low risk	Computer‐based, dynamic randomisation system. Stratified according to age, core infarct volume, time from symptom onset to enrolment, baseline NIHSS, and trial site
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Outcome assessed by certified rater who was blinded to trial assignment. Open‐label trial
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinded endpoint assessment
Incomplete outcome data (attrition bias) All outcomes	Low risk	3 participants lost to follow‐up, and intention‐to‐treat analysis provided
Selective reporting (reporting bias)	Low risk	Intention‐to‐treat analysis provided.
Other bias	Low risk	After an early interim analysis after the DAWN trial results and after 182 randomised participants, the trial was halted due to efficacy. Maximal sample size calculated to 476 participants.