FORWARD 2013.

Methods	Randomized trial to assess efficacy of PUFA for the maintenance of sinus rhythm in persistent atrial fibrillation (FORWARD) RCT, parallel, (n‐3 EPA + DHA vs MUFA), 12 months Summary risk of bias: low
Participants	Patients with paroxysmal atrial fibrillation N: 289 intervention, 297 control Level of risk for CVD: high Men: 57.8% intervention, 51.9% control Mean age in years (SD): 66.3 (12) intervention, 65.9 (10.5) control Age range: > 21 Smokers: 9% intervention, 6.2% control Hypertension: 92.2% intervention, 90.8% control Medications taken by at least 50% of those in the control group: aspirin, amiodarone, 'any antithrombotic treatment', beta‐blockers Medications taken by 20%‐49% of those in the control group: anticoagulants Medications taken by some, but less than 20% of the control group: none reported Location: Argentina Ethnicity: not reported
Interventions	Type: supplement (capsule) Comparison: EPA + DHA vs MUFA Intervention: one capsule/ day containing 1 g of n‐3 PUFA (Societá Prodotti Antibiotici and SigmaTau, Italy) (provided 850 mg to 882 mg EPA/DHA). Dose: 0.85 g/d EPA + DHA Control: identical placebo capsule containing olive oil Compliance: not reported. Length of intervention: 12 months
Outcomes	Main study outcome: survival free of atrial fibrillation Dropouts: 20 intervention, 25 control Available outcomes: mortality, MI, AF, heart failure, stroke, hospitalisation, side effects. Authors supplied further info on CVD events and methodology Response to contact: yes
Notes	Study funding: through unrestricted grants provided by companies that supplied study drugs, however "these companies did not have representatives on the Steering Committee" who terminated the trial after 1 year
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Participants were centrally assigned to receive either 1 g of n‐3 PUFA or placebo in a ratio of 1:1" – computer generated in blocks of 4 and 6 stratified by study location
Allocation concealment (selection bias)	Low risk	As above, centrally allocated. Communication from authors was ambiguous, stated that the person recruiting was aware of which arm the individual would be allocated to, but that the "study was double‐blind, placebo‐controlled."
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Each study site will be supplied with study drug and placebo in identically appearing packaging". "Both placebo and active treatment have the same odour and produce a comparable degree of fishy aftertaste"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Patients, investigator staff, persons performing the assessments, and data analysts will remain blind to the identity of the treatment from the time of randomisation until database lock" "The adjudication committee members are unaware of participant allocation and assess all available data and documentation with reference to pre‐established criteria".
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "the study was cut short by the trial steering committee due to 'a slower‐than‐expected recruitment rate and lower event rates'. This 'resulted in an underpowered clinical trial unable to verify its hypothesis'. Therefore the outcome data were not as complete as they were initially meant to be".
Selective reporting (reporting bias)	Low risk	Prospectively registered January 2008, study start January 2008, completion August 2011. All outcomes in trials registry appear to have been reported.
Attention	Low risk	Both intervention and control given the same exposure to research personnel. 2013 paper: "Clinical outcomes, adherence, and adverse events were assessed 2, 4, 8, and 12 months after randomization"
Compliance	Unclear risk	Not reported
Other bias	Low risk	None noted