The predicted rates of thromboembolism under dabigatran or warfarin treatment in randomized controls among patients with atrial fibrillation (AF) were close to the observed rates seen in routine care; however, rates of major bleeding were underestimated in clinical trials when compared to routine care. This according to a cohort study of 21,934 adults with AF initiating dabigatran (150 mg dose only) or warfarin treatment as part of routine care. Researchers found:

• 6,516 (30%) and 15,418 (70%) of the patient population initiated dabigatran and warfarin, respectively.

• Annual event rates per 100 patients were 1.7 from thromboembolism and 4.6 for major bleeding.

• For thromboembolism, calibration of estimates from randomized controlled trials was similar to calibration for model based predications.

• However, trial estimates for major bleeding consistently underestimated the rate of bleeding among patients in routine care.

• Underestimation of bleeding rates was particularly pronounced in warfarin initiators with high HAS-BLED scores.

Citation: Wang SV, Franklin JM, Glynn RJ, Schneeweiss S, Eddings W, Gagne JJ. Prediction of rates of thromboembolic and major bleeding outcomes with dabigatran or warfarin among patients with atrial fibrillation: new initiator cohort study. [Published online ahead of print May 24, 2016]. BMJ. doi:http://dx.doi.org/10.1136/bmj.i2607.

Commentary: Anticoagulation decreases the risk of stroke by 60 to 80%, but of course increases the risk of major bleeding. Prediction scores such as the CHADS₂-VAS_C score to predict the probability of thromboembolism and the HAS-BLED score to predict the probability of bleeding, can be useful in helping to determine the risk vs benefit of using anticoagulation in patients with atrial fibrillation. These risk scores though, were derived from randomized trial data. This trial shows that randomized trials underestimate the risk of bleeding in the real world. Upon reflection, this is not surprising; as to enter most randomized trials patients at high risk of bleeding would be excluded from the trial, making the trial have less bleeding than might occur in an unselected population. This trial lends support to what every good primary care physician already knows, that risk assessment has to be made on an individual basis and that evidence from randomized trials can inform us, but cannot be the final word in telling us what to do. To give the best patient care requires knowing the evidence and then applying the evidence with wisdom and attention to our individual patient. —Neil Skolnik, MD