POBBLE 2005.

Methods	Study type: double‐blind, randomised placebo‐controlled trial Study aim: to assess whether a policy of perioperative beta‐adrenergic blockade with metoprolol reduced 30‐day CV morbidity and mortality and reduced the length of hospital stay in people undergoing infrarenal vascular surgery Country: UK Setting: multicentre, vascular surgical units in 4 hospitals
Participants	Number randomised: total n = 103 (metoprolol n = 55; placebo n = 48) Age (median (IQR)): metoprolol = 73 (61‐79) years; placebo = 74 (66‐76) years Gender (M/F): total = 80/23; metoprolol = 40/13*; placebo = 35/9*, *6 participants did not undergo the procedure Inclusion criteria: people undergoing major elective infrarenal vascular surgery under general anaesthesia Exclusion criteria: already taking beta‐adrenergic blockers; beta‐adrenergic blockers could be dangerous; receiving current treatment for asthma; had aortic stenosis; had bradycardia or hypotension; perioperative beta‐adrenergic blockade had already been shown to be beneficial; had unstable angina, or angina with positive dobutamine stress test; previous MI CVD risk factors: current/ex smoker: metoprolol = 43%/47%, placebo = 10%/30%; diabetes: metoprolol = 19%, placebo = 18% Type of vascular surgery: infrarenal; aortic aneurysm repair, aortoiliac grafts for stenosis, femofemoral cross‐over grafts, femoropopliteal bypasses, femorodistal bypasses, amputation
Interventions	Treatment: metoprolol; 2‐4 mg in a slow intravenous injection over 5‐10 minutes before intubation, then 50 mg twice daily, oral route Control: placebo; equivalent 2‐4 mg in a slow intravenous injection over 5‐10 minutes before intubation, then 50 mg twice daily, oral route Duration: 7 days after surgery
Outcomes	Fatal and non‐fatal CV events within 30 days of surgery, myocardial ischaemia per 24 hours while wearing the Holter monitor, length of postoperative hospital stay, 2‐year survival
Notes	Study of average risk CV people undergoing vascular surgery, people with highest CV risk were excluded. All participants underwent a test dose of their allocated drug: metoprolol 50 mg or placebo equivalent for people weighing > 55 kg and 25 mg for people weighing ≤ 55 kg. People who did not tolerate the medication did not receive further beta‐adrenergic blockade
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Randomization as performed centrally at TheSealedEnvelope.com web site. Treatment was allocated in a 1:1 ratio by using random permuted blocks of size 2, 4, and 6 within four stratification factors..."
Allocation concealment (selection bias)	Low risk	Quote: "Randomization as performed centrally at TheSealedEnvelope.com web site"
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Trial drugs‐metoprolol and placebo‐of identical appearance were in gelatin‐coated capsules or ampoules..." Comment: study described as "double‐blind". For safety reasons the anaesthesiologists administering the study drug were unblinded, but precautions were taken to ensure remaining clinicians and trial co‐ordinators were blinded. There was no indication that blinding was broken for others involved in the trial, other than the anaesthesiologists
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Study did not specify whether the outcome assessors were blinded or not
Incomplete outcome data (attrition bias) All outcomes	Low risk	Drop‐out rate was even between treatment groups
Selective reporting (reporting bias)	Low risk	All outcomes reported on
Other bias	Unclear risk	Power calculations determined a study population of 300, but only 103 participants were recruited

BP: blood pressure; HR: heart rate; IQR: interquartile range; CHF: congestive heart failure; CV: cardiovascular; CVD: cardiovascular disease; 
 F: female; M: male; MI: myocardial infarction; 
 n: number; SD: standard deviation.