Abstract P6: Vitamin K Antagonist Dosing Methods and Individual Patient Quality of INR Control in the International Active W Anticoagulation Study

Abstract

Background The net benefit of vitamin K antagonists (VKAs) depends on the time spent in the therapeutic range (TTR) for the International Normalized Ratio in individual patients. Evidence-based methods are recommended by guidelines. We assessed VKA dosing methods among ACTIVE W study sites and the association with TTR in individual atrial fibrillation patients. Methods ACTIVE W sites received a survey questionnaire after the study to assess VKA dosing methods. Univariable and multivariable linear mixed models, to account for the random effect of clinic-level survey data, were used to assess the association of dosing methods with patient TTR. Patient-level covariates in multivariable analysis were: age, sex, CHADS 2 stroke risk score, mini-mental state examination score, history of VKA use, VKA type, and use of aspirin, amiodarone and insulin. Results The questionnaire was returned by 333 of 493 ACTIVE W sites (68%) who had at least one patient randomized to VKA. Responding sites had a higher mean study TTR than non-responding sites (64 vs. 60%; p=0.0101) and were mainly specialized in cardiology (87%). Only 28% of sites managed VKA dosing with an evidence-based method: an anticoagulation clinic, computer dosing system or patient self-management. Also taking in account (non-validated) manual algorithms, 64% of sites managed VKA dosing primarily based on clinical experience. In univariable analysis, patients achieved a higher TTR when managed by an anticoagulation clinic vs. by the study physician (67.3 vs. 62.1%; p=0.0027), when managed using a computer dosing system vs. using clinical experience (72.9 vs. 63.6%; p=0.0026), and when managed using at least one evidence-based method vs. not using evidence-based methods (67.3 vs. 62.8%; p=0.0045). However, when adding patient data in multivariable analysis, these three associations became non-significant (p-values 0.4659, 0.6555 and 0.6058, respectively). Conclusion The use of evidence-based VKA dosing methods was reported by only 28% of ACTIVE W sites, but was not significantly associated with an improved TTR when accounting for patient characteristics.