Among patients with atrial fibrillation (AF), the risk of thromboembolism is a significant concern. However, the reported use of warfarin among patients with AF at elevated risk of stroke remains low. In the present study, we have provided information on anticoagulation use reported during the recent Atrial Fibrillation: Focus on Effective Clinical Treatment Strategies (AFFECTS) Registry. Among patients identified by their physician at baseline to be at an increased risk of stroke, as determined from an assessment of the medical history, 74% received warfarin and 29% received aspirin. Post hoc analysis of warfarin use stratified by Congestive heart failure, Hypertension, Age, Diabetes, Stroke, (CHADS 2 ) doubled score revealed that at the end of the study, warfarin use was 73% (155 of 213) and 66% (185 of 280) in the rate- and rhythm-control patients with a score of ≥2, respectively, compared to 60% (183 of 306) and 49% (322 of 662) in the rate- and rhythm-control patients with a score of <2, respectively. The practicing cardiologists who participated in this registry initiated anticoagulation therapy in most of their patients with AF. However, warfarin use is not yet in line with the guidelines and evidence-based recommendations. Patients considered at no risk of stroke appear to have been overprescribed anticoagulant agents, and a considerable portion of high-risk patients did not receive warfarin. In conclusion, these results suggest that continued physician education of appropriate anticoagulation use in patients with AF is needed.
The American College of Cardiology/American Heart Association/European Society of Cardiology (ACC/AHA/ESC) atrial fibrillation (AF) management guidelines have deemed long-term anticoagulation appropriate for most patients with nonvalvular AF who have more than one risk factor for thromboembolism, regardless of the treatment strategy. Previously published data have indicated that anticoagulation has been substantially underused or inadequate in patients appropriate for this therapy. The results from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) and Rate Control versus Electrical Cardioversion (RACE) trials revealed that most strokes occurred after discontinuation of anticoagulation or when patients had received inadequate anticoagulant therapy (international normalized ratio <2.0). The Atrial Fibrillation: Focus on Effective Clinical Treatment Strategies (AFFECTS) Registry was designed to assess the treatment patterns of AF among practicing cardiologists in the United States who had received training with the 2001 or 2006 ACC/AHA/ESC AF guidelines and to assess the patient treatment outcomes. We report the use of anticoagulant agents among patients with AF in the AFFECTS Registry. The present analysis was conducted to determine anticoagulant use in the post-AFFIRM and -RACE era among subjects with varying degrees of stroke risk, as determined by physician assessment and post hoc Congestive heart failure, Hypertension, Age, Diabetes, Stroke, doubled (CHADS 2 ) stratification and to compare these strategies in patients for whom rate or rhythm control was chosen as the initial strategy.
Methods
Registry design, including site recruitment, inclusion and exclusion criteria for patient enrollment, and patient assessments have been fully described in the accompanying report, Practice Patterns Among United States Cardiologists for Managing Adults With Atrial Fibrillation (from the AFFECTS Registry) . The registry was open from March 2005 through December 2007. Each physician was trained using the ACC/AHA/ESC guidelines (2001 or 2006, depending on the date of enrollment). Only patients with AF without clinically significant structural heart disease (SHD) or AF with hypertension without left ventricular hypertrophy (as defined by the ACC/AHA/ESC guidelines ) were candidates for enrollment. All enrolled patients were expected to be seen by their respective physicians on a regular basis, according to that physician’s standard of care.
The data were collected on each patient at the enrollment (baseline) visit, and quarterly for 1 year after enrollment, for a total of 5 visits. All patients completed informed consent forms after being provided with an overview of the registry’s objective and procedures. The medical therapy for all patients was prescribed at the discretion of the investigator. No medications were provided by the study sponsor.
The patients were assigned to a primary therapy goal at baseline—either maintenance of sinus rhythm (rhythm-control group) or control of the heart rate (rate-control group)—by their physician. The baseline therapy goal for each patient did not bind the physician in any way to specific treatment modalities during the registry period. Serious adverse events (SAEs) were recorded at each quarterly visit. SAEs were reported for the all-enrolled population and for patients receiving warfarin and/or aspirin.
A post hoc review of patients’ medical history, either reported at baseline or noted in the diagnostic test results, was performed to assign a score for the risk of stroke according to the CHADS 2 stratification scheme. Patients with any literal terms indicative of congestive heart failure in their medical history were assigned 1 point according to the CHADS 2 stratification scheme. An additional 1 point each was assigned for a history of hypertension and diabetes and 2 points for previous stroke or transient ischemic attack. Patients who were ≥75 years old were assigned 1 point, in addition to any other identified risk factors. Once a patient’s medical history had been assessed for these risk factors, a CHADS 2 score (range 0 to 6 points) was computed. The CHADS 2 classification for the all-enrolled cohort was then stratified by the initial treatment strategy (rate or rhythm control), and the proportions of patients receiving warfarin at the beginning and end of the registry in each CHADS 2 classification were displayed. For the present analysis, patients with a CHADS 2 score of ≥2 were considered at “high risk” of stroke.
Results
A full accounting of physician recruitment and patient enrollment, including the flow of patients through the study and the patient baseline demographics and AF characteristics, has been previously published. Of the 14,804 sites invited to participate in the registry, 633 expressed interest in participating in the registry, 455 received institutional review board approval, 322 were activated, and ≥1 patient was enrolled at 248 of the 322 sites. Of the 322 sites, 50 enrolled >50% of the patients. Of the 1,535 patients enrolled, 1,461 had analyzable data and they comprise the all-enrolled population. The per-protocol cohort consisted of 1,110 patients (76%) from the all-enrolled population who met all the inclusion/exclusion criteria. The data reported in the present study were from the all-enrolled population, because this cohort included the patients at a greater risk of stroke (40 patients with hypertension and substantial left ventricular hypertrophy, 58 with coronary artery disease, and 70 with heart failure).
In the all-enrolled population, 942 patients (64%) were allocated to a rhythm-control therapy goal and 519 (36%) to a rate-control therapy goal ( Table 1 ) at entry. Overall, 83% of patients received antithrombotic or antiplatelet therapy—warfarin (64%) or aspirin (32%)—during the registry. Physicians primarily chose warfarin instead of aspirin, irrespective of whether a rate- or rhythm-control goal had been chosen. Warfarin use before any cardioversion in the rhythm-control group was not considered separately. Of the patients identified by their physician at baseline to be at increased risk of stroke (as determined by the assessment of the medical history and physician judgment), 74% received warfarin and 29% aspirin. Overall, 89% of patients at increased risk of stroke received warfarin or aspirin during the registry. In patients with no increased risk of stroke, 73% received anticoagulation with warfarin and/or aspirin (46% received warfarin alone).
| Variable | Rhythm Control | Rate Control | Total |
|---|---|---|---|
| Overall | 942 | 519 | 1,461 |
| Warfarin | 563 (60%) | 367 (71%) | 930 (64%) |
| Aspirin | 310 (33%) | 154 (30%) | 464 (32%) |
| Warfarin or aspirin | 761 (81%) | 455 (88%) | 1,216 (83%) |
| Increased stroke risk | 551 | 343 | 894 |
| Warfarin | 390 (71%) | 270 (79%) | 660 (74%) |
| Aspirin | 160 (29%) | 95 (28%) | 255 (29%) |
| Warfarin or aspirin | 478 (87%) | 317 (92%) | 795 (89%) |
| No increased stroke risk † | 360 | 168 | 528 |
| Warfarin | 154 (43%) | 91 (54%) | 245 (46%) |
| Aspirin | 137 (38%) | 57 (34%) | 194 (37%) |
| Warfarin or aspirin | 256 (71%) | 131 (78%) | 387 (73%) |
| Unknown increased stroke risk | 31 | 8 | 39 |
| Warfarin | 19 (61%) | 6 (75%) | 25 (64%) |
| Aspirin | 13 (42%) | 2 (25%) | 15 (39%) |
| Warfarin or aspirin | 27 (87%) | 7 (88%) | 34 (87%) |
⁎ Patients could have been administered warfarin and/or aspirin therapy.
† Determined by the investigator at baseline, with the prompt that increased risk factors for stroke included ≥65 years old, recent congestive heart failure, diabetes mellitus, history of stroke/transient ischemic attack, hypertension, or coronary heart disease.
A post hoc analysis of warfarin use stratified by CHADS 2 score is provided in Figure 1 . At the end of the registry period, the rate of warfarin use was 73% (155 of 213) and 66% (185 of 280) in the rate- and rhythm-control patients with a CHADS 2 score of ≥2, respectively. It was 60% (183 of 306) and 49% (322 of 662) in the rate- and rhythm-control patients with a CHADS 2 score of <2, respectively.
Central nervous system bleeding and embolic rate was low (13 of 1,461 [0.9%]) during the registry period; 4 patients experienced a cerebrovascular accident (0.3%), 3 had transient ischemic attacks (0.2%), 2 had an intracerebral hemorrhage (0.1%), 2 experienced a defined embolic stroke (0.1%), and 2 experienced an intracranial hemorrhage (0.1%). Of all the patients taking warfarin, 2 SAEs (0.1%) were thought to be associated with anticoagulation therapy—1 transient ischemic attack and 1 cerebral hemorrhage. No patients were reported to have experienced more than one of these events. No reported SAEs were associated with aspirin use.
Results
A full accounting of physician recruitment and patient enrollment, including the flow of patients through the study and the patient baseline demographics and AF characteristics, has been previously published. Of the 14,804 sites invited to participate in the registry, 633 expressed interest in participating in the registry, 455 received institutional review board approval, 322 were activated, and ≥1 patient was enrolled at 248 of the 322 sites. Of the 322 sites, 50 enrolled >50% of the patients. Of the 1,535 patients enrolled, 1,461 had analyzable data and they comprise the all-enrolled population. The per-protocol cohort consisted of 1,110 patients (76%) from the all-enrolled population who met all the inclusion/exclusion criteria. The data reported in the present study were from the all-enrolled population, because this cohort included the patients at a greater risk of stroke (40 patients with hypertension and substantial left ventricular hypertrophy, 58 with coronary artery disease, and 70 with heart failure).
In the all-enrolled population, 942 patients (64%) were allocated to a rhythm-control therapy goal and 519 (36%) to a rate-control therapy goal ( Table 1 ) at entry. Overall, 83% of patients received antithrombotic or antiplatelet therapy—warfarin (64%) or aspirin (32%)—during the registry. Physicians primarily chose warfarin instead of aspirin, irrespective of whether a rate- or rhythm-control goal had been chosen. Warfarin use before any cardioversion in the rhythm-control group was not considered separately. Of the patients identified by their physician at baseline to be at increased risk of stroke (as determined by the assessment of the medical history and physician judgment), 74% received warfarin and 29% aspirin. Overall, 89% of patients at increased risk of stroke received warfarin or aspirin during the registry. In patients with no increased risk of stroke, 73% received anticoagulation with warfarin and/or aspirin (46% received warfarin alone).
