Barrington 2005.

Methods	Parallel randomized controlled trial Ethics committee: approved by the ethics committee Informed consents: written informed consents obtained Site: St. Vincent’s Hospital, Melbourne, Australia Setting: university hospital Dates of data collection: from December 1999 to March 2002 Funding: grants from the Australian Society of Anaesthetists and the Australian and New Zealand College of Anaesthetists Registration: unspecified
Participants	120 participants scheduled for elective coronary artery bypass grafting surgery; mean age 62.5 years; sex distribution: 16 females and 104 males Inclusion criteria Patients scheduled for elective CABG surgery (using cardiopulmonary bypass (CPB)) were eligible Exclusion criteria Emergency or repeat CABG surgery Combined valve and CABG surgery Aspirin ingestion within 6 days of surgery Platelet count 150 × 10⁹/L International normalized ratio 1.1 Active neurological disease Cutaneous disorders at the epidural insertion site
Interventions	Intervention Epidural analgesia (N = 60) Comparator Systemic analgesia (N = 60) Premedication: temazepam, ranitidine, and morphine Induction: midazolam, fentanyl, propofol, and rocuronium Maintenance: propofol Surgery: CABG with CPB using a membrane oxygenator
Outcomes	Relevant to this review Risk of mortality Risk of myocardial infarction Risk of pulmonary complications Risk of atrial fibrillation or atrial flutter Risk of neurological complications (cerebrovascular accident) Tracheal extubation Pain scores Haemodynamic variables Others Arterial blood PaO₂ and PaCO₂
Notes	Correspondence: email sent 16 March 2018; no reply Conflict of interest: none reported DOI: 10.1213/01.ANE.0000146437.88485.47
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Patients were randomized the day before surgery to 2 groups. The random allocation sequence was computer‐generated in permuted blocks of 4 and was enclosed in sequentially numbered opaque sealed envelopes
Allocation concealment (selection bias)	Low risk	Opaque sealed envelopes
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not reported
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Twelve‐lead ECGs were recorded before surgery and on postoperative days 1 and 5 and were assessed by 2 observers blinded to group allocation and postoperative clinical course. No mention of blinding for any other outcome
Incomplete outcome data (attrition bias) All outcomes	Low risk	No loss to follow‐up. Two participants with failed epidural were kept in the analysis
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported
Other bias	Unclear risk	All participants were included in the intention‐to‐treat analysis. Prevalence of cerebrovascular and peripheral vascular disease was more frequent in the epidural group