BackgroundAtrial fibrillation is the most common cardiac arrhythmia, posing a heavy burden on patients' wellbeing and healthcare budgets. Patients undergoing cardiac surgery are at risk of developing postoperative atrial fibrillation (POAF), new-onset atrial fibrillation and subsequent atrial fibrillation-related complications, including stroke. Sufficient clinical identification of patients at risk fails while the pathological substrate changes that precede atrial fibrillation remain unknown. Here, we describe the PREDICT AF study design, which will be the first study to associate tissue pathophysiology and blood biomarkers with clinical profiling and follow-up of cardiothoracic surgery patients for the prediction of future atrial fibrillation.MethodsPREDICT AF will include 150 patients without atrial fibrillation and a CHA2DS2-VASc score of at least 2 undergoing cardiac surgery. The left atrial appendage will be excised during surgery and blood samples will be collected before surgery and at 6 and 12 months' follow-up. Tissue and blood analysis will be used for the discovery of biomarkers including microRNAs and protein biomarkers. The primary study endpoint is atrial fibrillation, which will be objectified by 24 h Holters and ECGs after 30 days for POAF and after 6, 12 and 24 months for new-onset atrial fibrillation. Secondary endpoints include the dynamic changes of blood biomarkers over time and other atrial arrhythmias. PREDICT AF participants may benefit from extensive postoperative care with clinical phenotyping, rhythm monitoring and primary prevention of stroke.ConclusionWe here describe the PREDICT AF trial design, which will enable the discovery of biomarkers that truly predict POAF and new-onset atrial fibrillation by combining tissue and plasma-derived biomarkers with comprehensive clinical follow-up data.Trial registrationRetrospectively registered NCT03130985 27 April 2017.