IJCS | Volume 33, Nº1, January / February 2019

74 Chart 3 - Frequency of use of anticoagulants regarding the HASBLED score in adult patients admitted with atrial fibrillation at a tertiary hospital in Salvador, Bahia, Brazil. Geraldes et al. Oral anticoagulation in AF Int J Cardiovasc Sci. 2020;33(1):68-78 Original Article the number of AF patients receiving oral anticoagulants, since they no longer consider antiplatelet agents as acceptable alternatives to anticoagulants in patients with lower thromboembolic risk. 16,20 Another important finding of our study was the lack of linearity between thromboembolic risk and the percentage of anticoagulant use. The frequency of anticoagulant prescription increased with the thromboembolic risk score up to CHA 2 DS 2 -VASc 3. However, in scores > 3, there was a progressive decrease in the use of anticoagulants. Similar findings were observed in other cohort studies, where anticoagulation rates peaked at CHA 2 DS 2 -VASc 3-4 scores, followed, unlike our study, with a plateau without an observable decline. 18,19 Thus, despite the known increase in the annual risk of stroke at each point of the CHA 2 DS 2 -VASc