Perspective:

The authors performed an interesting analysis to help personalize anticoagulation therapy in AF. Through the use of easily available clinical and demographic data (e.g., age, gender, ethnicity, antiplatelet use, prior stroke, prior bleeding, renal function), they are able to predict the 5-year risk of stroke and major bleeding for various treatment options (no anticoagulation, warfarin, dabigatran 110 mg, dabigatran 150 mg). Using various weighting of stroke:bleeding, the authors predict that between just fewer than half (1:1 weighting) to nearly all patients (2:1 weighting) are likely to get benefit from treatment with dabigatran 150 mg. This approach is quite useful for patient-provider shared decision making discussion. Caution should be taken before extrapolating findings from randomized clinical trials to unselected populations and in populations of patients treated with other direct oral anticoagulants. This model needs further validation and calibration in those patient cohorts.