Clinical determinants of oral anticoagulant selection in atrial fibrillation: insights from an Argentine registry

Background The current ESC guidelines recommend direct oral anticoagulants (DOACs) over vitamin K antagonists (VKAs) as the primary therapy for preventing stroke in patients with atrial fibrillation (AF). However, these guidelines do not prioritize one DOAC over another, leaving the final choice to clinical judgment or physician preference. Furthermore, there is a significant lack of evidence regarding utilization rates and the clinical determinants of molecule selection within South American populations, for which regional data remains scarce. Objective Identify independent clinical predictors that influence the selection of specific oral anticoagulants (OAC) in AF in a South American population Methods We analyzed 2,170 consecutive ambulatory patients receiving OAC due to AF from an Argentine institutional registry. Variables used were: age, gender, body weight and height, embolic risk, bleeding risk, and renal function. We used a multivariate logistic regression model to identify independent predictors of the prescription of VKA, apixaban, and rivaroxaban. We assessed renal function using the Cockcroft-Gault formula to estimate creatinine clearance (CrCl). Renal dysfunction was categorized as moderate (CrCl 30–49 mL/min) or severe (CrCl <30 mL/min). High embolic risk was defined as CHA2DS2-VASc ≥4, high bleeding risk as HAS-BLED ≥ 3 and advanced age as ≥80 years. Results The mean age of the study population was 72 (±9) years, and 48% were female. The median CHA2DS2-VASc and HAS-BLED scores were 4 (IQR 3–5) and 2 (IQR 1–3), respectively. Of the patients, 68% received DOACs: apixaban: 34%; rivaroxaban: 28%; and others: 6%. A further 32% received VKAs. Distinct clinical phenotypes were observed during the selection process: severe renal impairment was the strongest independent predictor of VKA use (OR: 5.2; 95% CI: 4.1–6.5; p <0.001). Apixaban was prescribed to high-risk individuals and was associated with a high ischemic risk (OR: 1.8; 95% CI: 1.3–2.4,p <0.01), advanced age (OR: 2.5; 95% CI: 1.9–3.2, p <0.001), moderate renal dysfunction (OR: 3.8; 95% CI: 3.1–4.6,p <0.001), and a high bleeding risk (OR: 1.5; 95% CI: 1.1–2.0, p <0.05). Rivaroxaban was preferred for individuals with a high body weight (>90 kg; OR 1.9; 95%CI: 1.4–2.5; p < 0.001), while advanced age was a negative predictor for its selection (OR 0.5; 95% CI 0.3–0.8;p < 0.01). Conclusions In this Argentine cohort, oral anticoagulant selection was risk-stratified based on patient phenotypes. VKAs were primarily reserved for patients with severe renal dysfunction. Almost 70% of the patients received DOACs: apixaban was preferred for elderly, renal-impaired, and high-risk patients, while rivaroxaban was favored for younger and overweight individuals. These findings suggest that, where guidelines provide no molecule-specific preference, patient characteristics appear to be stronger determinants of OAC selection than prescriber preference.