Objectives To describe the prescription patterns, medication adherence and quality of management of warfarin in Chinese patients with atrial fibrillation (AF).

Methods We conducted a prospective longitudinal cohort study of AF patients followed by either hospital visits or telephone every 6 months. From Aug 2011, patients were enrolled from 20 referral and 12 non-referral centres in Beijing. A CHA2DS2-VASc stroke-risk score was calculated for each patient at baseline. Among all the patients with CHA2DS2-VASc score> = 2, considering not using warfarin as an event, the fitted logistic regression analysis included gender, age, insurance, AF type, history of hypertension, diabetes, stroke or TIA, major bleeding, NYHA classification, blood pressure, creatinine level, and hospital level as independent variables. Proportion of non-adherent patients at 6 and 12 months was calculated in newly prescribed warfarin patients. The overall proportion of time in therapeutic range (TTR) was calculated for each patient on warfarin using Rosendaal’s method.

Results 6701 patients were enrolled (mean age 67.0, male 59.1%)and 2307 ablated patients were excluded from analysis. Among the remaining 4394 conservatively treated patients, 2843 and 887 patients have been followed up for 6 and 12 months, respectively. In 3687 patients with CHA2DS2-VASc Score≥2 who should be considered for anticoagulation, only 1130 (30.6%) patients received warfarin, while among 501 patients with CHA2DS2-VASc Score = 1 and 206 patients with CHA2DS2-VASc Score = 0, 124 (24.75%) and 41 (19.9%) were prescribed warfarin, respectively. Aspirin was concurrently used in 114/1295 (9.2%) patients using warfarin. Patients with the following characters were less likely to receive warfarin therapy: older than 75 years (HR 1.69, 95%CI 1.30-2.20), maximum out of pocket payment (HR 2.53, 1.72-3.72), new onset atrial fibrillation (HR 3.14, 2.22-4.43), paroxysmal AF (HR 1.87, 1.52-2.30), and systolic blood pressure higher than 140mmHg (HR1.75, 1.20-2.55). In newly prescribed warfarin patients followed for at least 6 or 12 months, 25.2% (56/222) and 44.1% (26/59) discontinued warfarin at 6 and 12 months, respectively. The median TTR is 0.50 (0.21-0.77) in patients on warfarin.

Conclusions Warfarin is significantly underutilized and poorly managed in AF patients in Beijing. Quality improvement initiatives are warranted in this area