Anticoagulation Therapeutic Ranges and Clinical Outcomes in Patients with a Mechanical Heart Valve Treated with Vitamin K Antagonists-a Nationwide Linked-data Dutch Study

Aims To examine the impact of different therapeutic international normalized ratio (INR) ranges on anticoagulation control and clinical outcomes in patients with mechanical heart valves (MHVs) treated with vitamin K antagonists (VKAs) in the Netherlands. Methods Data from 17 anticoagulation clinics (2013-2019) were linked to nation-wide data from Statistics Netherlands. Anticoagulation control metrics included significant dose adjustments, INR variance growth rate, and time in therapeutic range. Cause-specific Cox regression models were used to assess associations between therapeutic ranges and clinical outcomes, accounting for death as competing risk. Stratified analyses were performed for significant interactions by type of MHV recipient. R esults Among 3,473 MHV patients (median age: 67.0 [IQR: 58.0-76.0], 61.7% male, 68.2% acenocoumarol, 26.5% phenprocoumon), patients with lower therapeutic ranges (N = 1,866) (2.0-3.0 for isolated aortic valve without risk factors; 2.5-3.5 for all remaining MHV patients) had poorer anticoagulation control compared to those with higher ranges (N = 1,607) (2.5-3.5 and 3.0-4.0, respectively). No association was found between therapeutic ranges and major/clinically relevant bleeding (fully adjusted hazard ratio [aHR]: 0.80 [95%CI: 0.57-1.1]). However, in patients with a non-aortic valve and/or additional risk factors a lower therapeutic range was potentially associated with increased thromboembolic risk (aHR: 1.3 [95% CI: 0.94-1.9]), while no association was observed in patients with an isolated aortic valve (aHR: 0.71 [95%CI: 0.38-1.3]). Conclusion A lower therapeutic range does not apparently increase thromboembolic risk in most MHV patients but may be associated with a higher thromboembolic risk in higher risk patients. Lower therapeutic ranges were not associated with lower bleeding risk.