Most atrial fibrillation (AF) patients have coexisting cardiovascular diseases and risk factors that worsen prognosis and complicate management. Due to this complexity, few patients achieve all guideline recommended treatment targets. Improving the application of recommended comprehensive cardiovascular care potentially improves patient outcomes. Our aim was to test the effect of a computer-generated individualized audit and feedback intervention targeting physicians and patients to improve clinical outcomes in anticoagulated clinical trial AF patients. At the end of RE-LY, a randomized trial testing two dabigatran doses for stroke prevention in AF, patients were offered to continue dabigatran treatment in the extended RELY-ABLE study. Clinical centres participating in RELY-ABLE were randomized to receive intervention, or not. All centres received general educational material. The intervention encompassed computer-assisted decision support using audit of clinical data to generate individual patient-specific feedback letters regarding achievement of guideline-recommended management targets. Letters in 38 different local languages were sent to investigators two weeks in advance of the patient's scheduled study visit with the request to send a copy to the patient and discuss the content at the upcoming visit. Feedback addressed management targets distilled from international guidelines, including control of blood pressure, cholesterol and HbA1c, rhythm and rate control therapy, aspirin use, heart failure therapy and smoking. Clinical centres also received two report cards with feedback on their attainment of guideline targets for all their study patients. The primary outcome was a composite of vascular death, stroke, myocardial infarction, systemic embolism, major bleeding and cardiovascular hospitalization, assessed during 28 months of follow-up. Secondary outcome was the change in proportion of achieved guideline recommended management targets at the last study visit compared with baseline. The primary outcome was analyzed using a marginalized Cox model and the secondary outcome with a generalized estimating equation model, both to account for clustering of patients within clinical centers. In 34 participating countries, 269 practices were randomized to Intervention and 276 to Control, enrolling 3,010 and 2,853 AF patients respectively. Overall, the mean age (standard deviation) was 73 (8) years, 35% were female, 82% had hypertension, 32% coronary artery disease, 23% heart failure and 26% diabetes. At baseline, 78.3% (10.8) of all relevant guideline-recommended targets were met in the Intervention group and 77.6% (10.2) in the Control group. No difference in the occurrence of the primary outcome was observed between Intervention (12.7/100 person years) and Control (12.1/100 person years; p=0.47). The secondary outcome of proportional change in achievement of guideline-recommended treatment targets since baseline also showed no difference (Intervention 0.3% (10.5) versus Control 0.2% (10.7); p=0.98). Also no differences were observed regarding the individual components of the primary and secondary outcome. A computer-generated individualized audit and feedback intervention targeting patients and physicians did not improve clinical outcomes or comprehensive cardiovascular care for anticoagulated AF patients. A process evaluation will shed light on why this intervention did not establish an effect, to inform the design of future research.