Vedel 2018.

Study characteristics
Methods	Study design: randomised, open‐label controlled clinical trial Study duration: 19 months between July 2014 and January 2016 'Run‐in' period: none Number of study centre and location: 1 in Denmark
Participants	Total number of study participants: 197 Number of randomised participants: 197 Number lost to follow‐up: 0 Number withdrawn: 3 Number analysed: 197 Mean age: 67.2 years Gender: 177 men Inclusion criteria: aged ≥ 18 years and in need of elective or subacute CABG or left heart valve surgery (or both) using CPB Exclusion criteria: history of stroke or intracranial bleeding, history of reversible ischaemic deficits (duration of symptoms 24–72 hours), history of transient ischaemic attacks (duration of symptoms < 24 hours), diagnosis of neurodegenerative disorders such as Alzheimer's disease and multiple sclerosis, or contraindications to magnetic resonance imaging
Interventions	Experimental: maintaining MAP at 70–80 mmHg during CPB Comparison: maintaining MAP at 40–50 mmHg during CPB Concomitant medications: an intended fixed, equal, and non‐pulsatile blood flow of 2.4 L/minute/m² body surface area plus 10–20% was applied in both groups, and assigned MAP levels were targeted with intermittent intravenous doses of phenylephrine to a total maximum dose of 2.0 mg followed by continuous intravenous infusion of noradrenaline up to 0.4 μg/kg/minute. Concomitant treatment interventions were at the treating clinicians' discretion. According to departmental guidelines, CPB was performed with arterial oxygen saturation > 96% (and partial pressure of oxygen in the arterial blood > 13.0 kPa), normocapnia (partial pressure of carbon dioxide in the arterial blood 4.5–6.0 kPa), normothermia (body temperature > 36.5 °C), α‐stat pH management, and transfusion of packed red blood cells if haematocrit was < 24% (or at higher haematocrit levels in case of lactic acidosis or low mixed venous saturation). Excluded medications: no vasodilatory drugs were accepted in the low‐target group, and MAP levels above the low‐target window were accepted.
Outcomes	Primary outcome: total volume of new ischaemic lesions (sum in millimetres cubed), expressed as the difference between DWI conducted preoperatively and again between days 3 and 6. Secondary outcomes: total number of new ischaemic cerebral lesions, POCD, new focal neurological deficits, both evaluated as a change from baseline neuropsychological and neurological test performance to the result at 1 week, at discharge from the hospital, or at healthcare relocation from the cardiac surgery ward to a local hospital, whichever came first. Cognitive function was re‐evaluated at 2–4 months postoperatively. Used the International Study of Postoperative Cognitive Dysfunction test battery, 14–16 which was developed for people with an intact cognitive capability. Also conducted a Mini‐Mental State Examination to screen for potential signs of dementia. If a participant had a baseline Mini‐Mental State Examination score ≤ 24, there was no further cognitive testing.
Notes	Dr Vedel received grants from the Danish Heart Foundation (14‐R97‐ A5179‐22868 and 15‐R99‐A6034‐22905) and the Research Foundations at Rigshospitalet (E‐22329‐01), University of Copenhagen, Denmark.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Centralised web‐based randomisation was performed with a computer‐generated allocation sequence with varying block size of 4–8.
Allocation concealment (selection bias)	Low risk	Centralised web‐based randomisation was performed with a computer‐generated allocation sequence with varying block size of 4–8.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Patients were blinded from the group allocation. Healthcare providers in the intensive care unit and ward were unaware of the assigned MAP strategy unless they specifically consulted the handwritten anaesthesia report and guessed to which group the participant was allocated based on the continuous registration of MAP data. Throughout the trial, the study authors stressed that the staff involved in the experimental setup in the operating room at no point disclosed group allocation.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	According to the protocol, assessors of the primary and selected secondary endpoints (DWI scans and POCD) were blinded to treatment allocation. Furthermore, blood samples were labelled with a unique participant number, and personnel carrying out blood sample analysis did not know the group allocation.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Lost to follow‐up 3.0% (6/197).
Selective reporting (reporting bias)	Low risk	All outcome data but MRS described in the protocol (Vedel 2016 – see under Vedel 2018) were reported.
Other bias	Low risk	Review authors believed the study free of other sources of bias.

AKI: acute kidney injury; CABG: coronary artery bypass graft; CPB: cardiopulmonary bypass; DWI: diffusion‐weighted imaging; MAP: mean arterial pressure; MRS: magnetic resonance spectroscopy; POCD: postoperative cognitive dysfunction; SF‐36: 36‐item Short Form.