Nodari 2011 AF

Methods	RCT, parallel, (DHA + EPA vs MUFA), 12 months Summary risk of bias: moderate or high
Participants	Patients with persistent atrial fibrillation with at least 1 relapse after cardioversion N: 102 intervention, 103 control. (analysed, intervention: 94 control: 94) Level of risk for CVD: high Men: 70% intervention, 63% control Mean age in years (SD): 70 (6) intervention, 69 (9) control Age range: not reported (18‐80 inclusion criteria) Smokers: 10% intervention, 9.1% control Hypertension: 47% intervention, 40% control Medications taken by at least 50% of those in the control group: beta‐blockers, ACE inhibitors, anticoagulant therapy, amiodarone Medications taken by 20%‐49% of those in the control group: diuretics, antiplatelet, statins Medications taken by some, but less than 20% of the control group: calcium channel blockers Location: Italy Ethnicity: not reported
Interventions	Type: supplement (omega‐3‐acid ethyl esters 90: Omacor) Comparison: EPA + DHA vs MUFA Intervention: 2 × 1 g/d Omacor (total 1.7 g/d EPA + DHA at a ratio of 0.9 to 1.5). Dose: 1.7 g/d EPA + DHA Control: 2 × 1 g/d olive oil (gelatin capsules identical in appearance to Omacor) Compliance: no details Duration of intervention: 12 months
Outcomes	Main study outcome: probability of maintenance of sinus rhythm Dropouts: 6 intervention, 5 control Available outcomes: adverse events, AF recurrence (nil death) Response to contact: no
Notes	Study funding: 'Centro per lo Studio ed il Trattamento dello Scompenso Cardiaco' of the University of Brescia, Brescia, Italy. The work of Dr Campia was supported by National Institutes of Health grant K12 HL083790‐01a1
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random assignment followed a computer‐generated randomisation list obtained using blocks of size 4
Allocation concealment (selection bias)	Low risk	The randomisation schedule was kept in the research pharmacy area and was available only to unblinded pharmacy personnel until after the database was locked. At that time, the unblinded patient treatment information was made available to the investigators.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Placebo gelatin capsules identical in appearance to Omacor. However no information provided as to their smell and taste.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No details
Incomplete outcome data (attrition bias) All outcomes	Low risk	All randomised were accounted for. ITT analysis for main outcomes
Selective reporting (reporting bias)	Unclear risk	NCT01198275. Registered retrospectively in September 2010, study started January 2006, completed May 2008, main publication 2011
Attention	Low risk	No difference between groups
Compliance	Unclear risk	No details
Other bias	Low risk	None noted