The risk for thromboembolic complications and. Original Research

Original Research Annals of Internal Medicine Risks for Stroke, Bleeding, and Death in Patients With Atrial Fibrillation Receiving Dabigatran or Warfarin in Relation to the CHADS 2 Score: A Subgroup Analysis of the RE-LY Trial Jonas Oldgren, MD, PhD; Marco Alings, MD, PhD; Harald Darius, MD, PhD; Hans-Christoph Diener, MD, PhD; John Eikelboom, MD; Michael D. Ezekowitz, MD, PhD; Gabriel Kamensky, MD, PhD; Paul A. Reilly, PhD; Sean Yang, MSc; Salim Yusuf, MBBS, DPhil; Lars Wallentin, MD, PhD; and Stuart J. Connolly, MD, on behalf of the RE-LY Investigators Background: CHADS 2 is a simple, validated risk score for predicting the risk for stroke in patients with atrial fibrillation not treated with anticoagulants. There are sparse data on the risk for thrombotic and bleeding complications according to the CHADS 2 score in patients receiving anticoagulant therapy. Objective: To evaluate the prognostic importance of CHADS 2 risk score in patients with atrial fibrillation receiving oral anticoagulants, including the vitamin K antagonist warfarin and the direct thrombin inhibitor dabigatran. Design: Subgroup analysis of a randomized, controlled trial. (ClinicalTrials.gov registration number: NCT00262600) Setting: Multinational study setting. Patients: 18 112 patients with atrial fibrillation who were receiving oral anticoagulants. Measurements: Baseline CHADS 2 score, which assigns 1 point each for congestive heart failure, hypertension, age 75 years or older, and diabetes mellitus and 2 points for stroke. score of 0 to 1, 1.22% in those with a score of 2, and 2.24% in those with a score of 3 to 6. Annual rates of other outcomes among all participants with CHADS 2 scores of 0 to 1, 2, and 3 to 6, respectively, were the following: major bleeding, 2.26%, 3.11%, and 4.42%; intracranial bleeding, 0.31%, 0.40%, and 0.61%; and vascular mortality, 1.35%, 2.39%, and 3.68% (P 0.001 for all comparisons). Rates of stroke or systemic embolism, major and intracranial bleeding, and vascular and total mortality each increased in the warfarin and dabigatran groups as CHADS 2 score increased. The rates of stroke or systemic embolism with dabigatran, 150 mg twice daily, and of intracranial bleeding with dabigatran, 150 mg or 110 mg twice daily, were lower than those with warfarin; there was no significant heterogeneity in subgroups defined by CHADS 2 scores. Limitation: These analyses were not prespecified and should be deemed exploratory. Conclusion: Higher CHADS 2 scores were associated with increased risks for stroke or systemic embolism, bleeding, and death in patients with atrial fibrillation receiving oral anticoagulants. Results: Distribution of CHADS 2 scores were as follows: 0 to 1 5775 patients; 2 6455 patients; and 3 to 6 5882 patients. Annual rates of the primary outcome of stroke or systemic embolism among all participants were 0.93% in patients with a CHADS 2 Primary Funding Source: Boehringer Ingelheim. Ann Intern Med. 2011;155:660-667. For author affiliations, see end of text. www.annals.org The risk for thromboembolic complications and bleeding in atrial fibrillation varies in different patient subsets. Whereas paroxysmal and permanent atrial fibrillation present similar stroke risk (1 3), other clinical features may help to stratify patients by risk. A history of stroke or transient ischemic attack is the most potent clinical risk factor (4 6). Hypertension, heart failure, diabetes, coronary artery disease, female sex, and advancing age are also associated with increased risk (5). Several stroke risk stratification schemes (7) have been proposed, mainly focusing on patients not receiving oral See also: Print Editors Notes... 661 Editorial comment....714 Related article....653 Web-Only Conversion of graphics into slides anticoagulants. The CHADS 2 score is a simple, validated measure of risk (8) that assigns 1 point for a history of congestive heart failure, hypertension, age 75 years or older, and diabetes mellitus and 2 points for a history of stroke or transient ischemic attack. There are sparse data on survival and bleeding rates in relation to the CHADS 2 risk score, especially during oral anticoagulant treatment (6, 9 11). Thus, the threshold at which benefit of oral anticoagulation for prevention of stroke in atrial fibrillation exceeds the risk for bleeding is unclear. International atrial fibrillation guidelines (12 14) uniformly recommend the use of warfarin for patients with a CHADS 2 score of 2 or higher. The U.S. guidelines (12, 13) suggest a choice between warfarin and aspirin in patients with lower stroke risk that is, a CHADS 2 score of 1 on the basis of concerns that the bleeding risk with warfarin may outweigh the benefit. Recent European guidelines (14) suggest further risk stratification by considering other risk factors, such as sex, vascular disease, and age 65 to 74 years, in patients with a CHADS 2 score less than 2 and recommend oral anticoagulation in patients with at least 1 of these additional risk factors. 660 2011 American College of Physicians

Dabigatran vs. Warfarin and CHADS 2 in Atrial Fibrillation Original Research The RE-LY (Randomized Evaluation of Long-term Anticoagulation Therapy) trial (15), which compared the direct thrombin inhibitor dabigatran with the vitamin K antagonist warfarin for prevention of stroke and systemic embolism in patients with atrial fibrillation, provides the opportunity to assess thrombotic and bleeding risk according to the baseline CHADS 2 score in patients receiving anticoagulation. The aims of the present analysis were to evaluate the prognostic importance of CHADS 2 risk score in patients with atrial fibrillation receiving oral anticoagulants and to assess study treatment interactions by CHADS 2 risk groups. METHODS The RE-LY study (15) compared 2 blinded doses of dabigatran (150 mg and 110 mg twice daily) with openlabel warfarin in patients with atrial fibrillation at increased risk for stroke. The detailed design and primary results have been published (15 17). The study recruited 18 113 patients from 951 clinical centers in 44 countries. The inclusion criteria (15) were documented atrial fibrillation and at least 1 of the following risk factors for stroke: previous stroke or transient ischemic attack; congestive heart failure or reduced left ventricular ejection fraction ( 0.40); and aged at least 75 years or at least 65 years with diabetes mellitus, hypertension, or coronary artery disease. Thus, patients across all CHADS 2 scores were eligible because patients with the combination of the risk factors aged 65 to 74 years and coronary artery disease have a CHADS 2 score of 0. Patients with severe heart valve disorder, recent stroke, increased risk for hemorrhage, creatinine clearance less than 30 ml/min, or active liver disease were excluded (15). Median duration of follow-up was 2 years. The study was approved by all appropriate national regulatory authorities and ethics committees. All patients provided written informed consent before study entry. The primary RE-LY study outcome was stroke or systemic embolism. The primary safety outcome was major bleeding. Other outcomes were intracranial hemorrhage and vascular death (including death from bleeding) and total death. Stroke was defined as sudden onset of focal neurologic deficit consistent with the territory of a major cerebral artery and categorized as ischemic, hemorrhagic, or unspecified. Hemorrhagic transformation of ischemic stroke was not considered a hemorrhagic stroke. Intracranial hemorrhage included hemorrhagic stroke and subdural or subarachnoid hemorrhage. Systemic embolism was an acute vascular occlusion of an extremity or organ documented by imaging, surgery, or autopsy. Major bleeding was defined as a reduction in hemoglobin level of at least 20 g/l, or transfusion of at least 2 units of blood, or symptomatic bleeding in a critical area or organ. All other bleeding episodes were considered minor. All primary and secondary outcome events were blindly and doubly adjudicated. Context Baseline CHADS 2 scores predict risk for thromboembolic complications in patients with atrial fibrillation in the absence of anticoagulation. Whether these scores are useful for predicting thromboembolic and bleeding complications in patients treated with anticoagulants for atrial fibrillation, however, is not known. Contribution In this analysis of data from a randomized clinical trial in which patients received anticoagulant therapies for atrial fibrillation, the risks for stroke or thromboembolism, major and intracranial bleeding, and vascular mortality increased with higher CHADS 2 scores. Caution These were post hoc and exploratory analyses. Implication These data may be helpful in assessing the risks and benefits of anticoagulant therapy in patients with atrial fibrillation. The Editors Statistical Analysis Patient characteristics were evaluated with analysis of variance (continuous values) and a chi-square test (categorical values). One patient receiving 110 mg of dabigatran twice daily did not have a CHADS 2 score because of missing values and was excluded from the analyses. Quality of warfarin treatment was assessed by calculating the percentage of time in therapeutic range (international normalized ratio, 2.0 to 3.0) for individual patients by using Rosendaal and colleagues method (18). We excluded international normalized ratios during the first week of the study and those obtained during temporary or permanent discontinuation. The relationship between annual event rates and increasing CHADS 2 risk scores was assessed by using a Cochran Armitage test of linear trend. Because of low numbers of patients and events in the groups with CHADS 2 scores of 0 (n 452), 4 (n 1619), 5 (n 524), and 6 (n 85), patients were grouped into 3 almost equal-sized groups in the post hoc analyses of study treatment effects: CHADS 2 scores of 0 to 1 (n 5775), 2 (n 6455), and 3 to 6 (n 5882). Event rates per 100 patient-years were calculated as stratified by treatment and CHADS 2 risk group. Absolute risk reductions were calculated for outcome events by comparing the 3 treatment groups within the 3 CHADS 2 risk groups, with 95% CIs obtained by the Byar method (an approximation of Poisson distribution). Cox proportional hazards regression models were used to compare the cumulative event rates by treatment within the 3 CHADS 2 risk groups, and models contained terms for treatment and indicator variables for CHADS 2 group and their interactions. The proportional hazards assumption was checked by www.annals.org 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 661

Original Research Dabigatran vs. Warfarin and CHADS 2 in Atrial Fibrillation Table 1. Baseline Characteristics in Relation to CHADS 2 Scores Characteristic CHADS 2 Score P Value 0 1 2 3 6 Patients, n 5775 6455 5882 Mean age (SD), y 69.5 (7.4) 71.9 (9.1) 73.0 (9.0) 0.001 Mean weight (SD), kg 83.1 (19.0) 83.0 (20.2) 81.9 (19.6) 0.002 Mean systolic blood pressure (SD), mm Hg 130.4 (17.6) 131.6 (17.5) 130.9 (17.3) 0.001 Mean diastolic blood pressure (SD), mm Hg 77.5 (10.2) 77.4 (10.5) 76.3 (10.9) 0.001 Men, n (%) 3837 (66.4) 4032 (62.5) 3645 (62.0) 0.001 Type of atrial fibrillation, n (%) Persistent* 1785 (30.9) 2114 (32.7) 1890 (32.1) 0.09 Paroxysmal 2087 (36.1) 2083 (32.3) 1773 (30.1) 0.001 Permanent 1903 (33.0) 2254 (34.9) 2218 (37.7) 0.001 CHADS 2 components, n (%) Congestive heart failure 721 (12.5) 2229 (34.5) 2843 (48.3) 0.001 Hypertension 3396 (58.8) 5507 (85.3) 5380 (91.5) 0.001 Age 75 y 1044 (18.1) 2994 (46.4) 3220 (54.7) 0.001 Diabetes mellitus 162 (2.8) 1426 (22.1) 2633 (44.8) 0.001 Stroke or transient ischemic attack 0 (0.0) 377 (5.8) 3246 (55.2) 0.001 Previous myocardial infarction, n (%) 778 (13.5) 958 (14.8) 1269 (21.6) 0.001 Medications in use at baseline, n (%) Aspirin 2088 (36.2) 2611 (40.4) 2499 (42.5) 0.001 ACE inhibitor or ARB 3173 (54.9) 4466 (69.2) 4340 (73.8) 0.001 -Blocker 3561 (61.7) 4080 (63.2) 3734 (63.5) 0.09 Amiodarone 604 (10.5) 727 (11.3) 602 (10.2) 0.15 Statin 2284 (39.5) 2792 (43.3) 2962 (50.4) 0.001 Proton-pump inhibitor 696 (12.1) 881 (13.6) 914 (15.5) 0.001 Histamine-2 receptor antagonist 208 (3.6) 246 (3.8) 268 (4.6) 0.02 Median time in therapeutic range for warfarin recipients, % 68.9 67.7 65.3 0.001 ACE angiotensin-converting enzyme; ARB angiotensin-receptor blocker. * Atrial fibrillation episodes that do not terminate spontaneously but do convert with electrical or pharmacologic cardioversion. Atrial fibrillation episodes that terminate spontaneously. considering time-dependent variables for treatment and CHADS 2 group. The following outcome events were tested as the dependent factors in these analyses: stroke and systemic embolism, major bleeding, intracranial bleeding, vascular death, and total death. All analyses were performed separately for 110 mg of dabigatran twice daily versus warfarin and for 150 mg of dabigatran twice daily versus warfarin. All analyses were performed by using SAS statistical software, version 9.1 (SAS Institute, Cary, North Carolina). Role of the Funding Source The study was funded by Boehringer Ingelheim and coordinated by the Population Health Research Institute (Hamilton, Ontario, Canada), which independently managed the database and analyzed the data. An operations committee, with assistance from an international steering committee, was responsible for the design, conduct, and reporting of the study. The lead authors (J. Oldgren, S. Yusuf, L. Wallentin, S.J. Connolly) had full access to data and designed the statistical analysis. Dr. Oldgren had final responsibility for the decision to submit the manuscript for publication. RESULTS Baseline characteristics significantly differed among the patients with CHADS 2 scores of 0 to 1 (n 5775), 2 (n 6455), and 3 to 6 (n 5882) (Table 1). Hypertension was the most prevalent (58.8%) individual component of the CHADS 2 score in patients with a CHADS 2 score of 0 to 1, and less than 3% of patients in this group had diabetes mellitus. In the group with a CHADS 2 score of 3 to 6, more than 90% had hypertension, and the majority had previously had a stroke or transient ischemic attack or were aged 75 years or older. Mean time in the therapeutic range was slightly lower in patients with higher baseline CHADS 2 scores: 65.9% (SD, 19.0%), 64.9% (SD, 20.0%), and 62.4% (SD, 20.3%) in the groups with CHADS 2 scores of 0 to 1, 2, and 3 to 6, respectively. Primary Outcome In the overall cohort, the rate of stroke or systemic embolism increased for each 1-point increase in the CHADS 2 risk score, ranging from 0.53% per year with a CHADS 2 score of 0 to 5.40% per year with a CHADS 2 score of 6 (P 0.001) (Figure 1). Increasing CHADS 2 scores were associated with increased event rates in all 3 study treatment groups (Table 2). The reduction in stroke or systemic embolism with dabigatran, 150 mg twice daily, versus warfarin was consistent across the CHADS 2 risk groups (Table 2). The rates of stroke or systemic embolism were similar with dabigatran, 110 mg twice daily, and warfarin across CHADS 2 risk groups (Table 2). 662 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 www.annals.org

Dabigatran vs. Warfarin and CHADS 2 in Atrial Fibrillation Original Research Bleeding An almost linear increase in the annual rate of major bleeding for each 1-point increase in the CHADS 2 score occurred in the overall cohort: from 1.60% per year in the group with a CHADS 2 score of 0 to 5.40% per year in the group with a CHADS 2 score of 6 (P 0.001) (Figure 2). In addition, major bleeding increased according to the CHADS 2 score within each of the dabigatran and warfarin treatment groups (Table 2). Intracranial bleeding rates were also related to the CHADS 2 score, with rates ranging from 0.11% per year with a CHADS 2 score of 0 to 1.10% with a CHADS 2 score of 5 (P 0.001); no intracranial bleeding event occurred among patients with a CHADS 2 score of 6 (Figure 2). Increasing CHADS 2 scores were associated with increased event rates in all 3 study treatment groups (Table 2). Study treatment effects were consistent across CHADS 2 risk groups: Intracranial bleeding rates were lower in both dabigatran groups than in the warfarin group. Mortality In the overall cohort, vascular mortality rates were low in the group with a CHADS 2 score of 0 to 1 and mortality increased rates for each 1-point increase above 1 in the CHADS 2 score, ranging from 1.34% per year with a CHADS 2 score of 1 to 10.8% per year with a CHADS 2 score of 6 (P 0.001) (Figure 1). Total mortality rates were similarly associated with the CHADS 2 score, with rates from 2.28% (95% CI, 2.00% to 2.58%) per year with a CHADS 2 score of 1 to 13.5% (CI, 8.24% to 20.8%) per year with a CHADS 2 score of 6 (P 0.001) (Table 2). DISCUSSION Patients with atrial fibrillation across all CHADS 2 scores (ranging from 0 to 6) were included in the RE-LY trial (15), but the numbers of individuals with CHADS 2 scores at the extreme ends of the distribution were limited. Our results showed that higher CHADS 2 scores were associated with increased risks for not only stroke or systemic embolism but also major and intracranial bleeding and death in this large cohort of patients treated with oral anticoagulants. The increased rates of stroke and systemic embolism with increasing CHADS 2 scores in our current study group receiving anticoagulant treatment is similar to the 1.5-fold increase by each CHADS 2 point in the original observations in patients with atrial fibrillation not receiving oral anticoagulants (8). These results are furthermore consistent with observations in studies in which a proportion of patients received oral anticoagulants (6, 19), as well as the results from the pooled SPORTIF (Stroke Prophylaxis Using an Oral Thrombin Inhibitor in Atrial Fibrillation) III and V trials (n 7329) comparing the direct thrombin inhibitor ximelagatran with warfarin in patients with atrial fibrillation at risk for stroke (9). There are sparse data on survival rates in relation to the CHADS 2 risk score, especially during active anticoag- Figure 1. Annual rates of stroke or SE and vascular mortality in relation to CHADS 2 risk scores. 12 10 Stroke or SE Vascular mortality 17.5 Percentage per Year 8 6 4 2 0 0 1 2 3 4 5 6 CHADS 2 Score Patients, n 452 5323 6455 3654 1619 524 85 Stroke or SE/vascular mortality, n/n* 5/13 103/144 154/301 139/247 70/111 40/48 8/16 Cochran Armitage linear trend test P 0.001. Error bars represent 95% CIs. SE systemic embolism. * Total number of events (median follow-up, 2 y). www.annals.org 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 663

Original Research Dabigatran vs. Warfarin and CHADS 2 in Atrial Fibrillation Table 2. Stroke or Systemic Embolism, Death, Bleeding Events, Treatment Effects, and Interactions in CHADS 2 Groups* CHADS 2 Score per Outcome Rate in All Participants, %/y Dabigatran, 110 mg Twice Daily Total Events, n Rate, %/y Dabigatran, 150 mg Twice Daily Total Events, n Rate, %/y Total Events, n Warfarin Rate, %/y Absolute Risk Reduction (95% CI), %/y Dabigatran, 110 mg, vs. Warfarin Dabigatran, 150 mg, vs. Warfarin Stroke or systemic embolism 0 to 1 0.93 42 1.06 26 0.65 40 1.08 0.02 ( 0.44 to 0.49) 0.43 (0.01 to 0.85) 2 1.22 59 1.45 35 0.84 60 1.38 0.08 ( 0.59 to 0.43) 0.54 (0.09 to 0.98) 3 to 6 2.44 82 2.12 73 1.88 102 2.73 0.61 ( 0.09 to 0.31) 0.85 (0.16 to 1.53) P value 0.001 0.001 0.001 0.001 Major bleeding 0 to 1 2.26 74 1.86 84 2.11 105 2.84 0.98 (0.29 to 1.66) 0.73 (0.02 to 1.43) 2 3.11 121 2.98 127 3.04 144 3.30 0.32 ( 0.44 to 1.08) 0.27 ( 0.49 to 1.02) 3 to 6 4.42 147 3.80 188 4.85 172 4.60 0.80 ( 0.12 to 1.73) 0.25 ( 1.22 to 1.73) P value 0.001 0.001 0.001 0.001 Intracranial bleeding 0 to 1 0.31 8 0.20 8 0.20 20 0.54 0.34 (0.06 to 0.61) 0.34 (0.06 to 0.61) 2 0.40 9 0.22 11 0.26 30 0.69 0.47 (0.18 to 0.75) 0.43 (0.13 to 0.72) 3 to 6 0.61 10 0.26 20 0.52 40 1.07 0.81 (0.44 to 1.18) 0.55 (0.15 to 0.96) P value 0.001 0.64 0.01 0.02 Vascular death 0 to 1 1.35 55 1.38 43 1.08 59 1.59 0.21 ( 0.34 to 0.76) 0.51 ( 0.01 to 1.03) 2 2.39 104 2.56 84 2.01 113 2.59 0.03 ( 0.66 to 0.71) 0.58 ( 0.06 to 1.23) 3 to 6 3.68 130 3.36 147 3.79 145 3.88 0.52 ( 0.33 to 1.38) 0.09 ( 0.79 to 0.97) P value 0.001 0.001 0.001 0.001 Total death 0 to 1 2.30 93 2.34 77 1.94 98 2.65 0.31 ( 0.40 to 1.02) 0.71 (0.03 to 1.39) 2 3.88 165 4.07 142 3.40 182 4.18 0.11 ( 0.76 to 0.98) 0.78 ( 0.05 to 1.61) 3 to 6 5.35 188 4.86 219 5.65 207 5.54 0.68 ( 0.35 to 1.71) 0.11 ( 1.17 to 0.95) P value 0.001 0.001 0.001 0.001 * The numbers of patients assigned to the study groups per CHADS 2 score (0 to 1, 2, and 3 to 6, respectively) were as follows: dabigatran, 110 mg twice daily 1958, 2088, and 1968; dabigatran, 150 mg twice daily 1958, 2137, and 198; and warfarin 1859, 2230, and 1933. P value for linear trend. P value for interactions between groups. ulant treatment (10, 11). In all treatment groups, we demonstrated a strong relation between CHADS 2 risk groups and vascular and total mortality. There was also a progressive, almost linear increase in the risk for major and intracranial bleeding in relation to CHADS 2 scores of 0 to 5 during anticoagulant treatment, with a limited number of bleeding events in the small group of patients (n 85) with a CHADS 2 score of 6. These results are similar to previous reports of increased bleeding risk at higher CHADS 2 scores, at least in very elderly patients (6, 20, 21). Thus, beyond the risk prediction of thromboembolic events, higher CHADS 2 scores identify patients at increased risk for severe bleeding complications and death during both warfarin and dabigatran treatment. In the RE-LY study, dabigatran, 110 mg twice daily, was noninferior to warfarin for stroke prevention overall and dabigatran, 150 mg twice daily, was superior to warfarin overall. These results were generally consistent in all CHADS 2 risk groups, including lower-risk patients (those with a CHADS 2 score of 0 to 1). Accordingly, high-risk patients with CHADS 2 scores of 3 to 6 had the greatest absolute stroke risk reduction with dabigatran, 150 mg twice daily. The threshold of stroke risk at which the benefits of anticoagulation with warfarin exceeds the risk for bleeding remains unclear. International guidelines (12 14) recommend the use of oral anticoagulants for patients with a CHADS 2 score of 2 or higher, but less than two thirds of eligible patients receive this treatment in clinical practice (22 24). Beyond absolute contraindications for anticoagulants, reasons for undertreatment include patient-, physician-, and health care system related barriers and the inherent problems with vitamin K antagonists (25). Psychological barriers with perceived greater risk for bleeding or underestimation of the stroke risk may play a role (26, 27). The safety benefits of dabigatran compared with warfarin in patients with CHADS 2 scores of 2 or above with marked reduction in intracranial bleeding by both dabigatran doses might reduce the barriers to anticoagulation. In patients with a CHADS 2 score of 1, treatment guidelines (14 16) indicate a choice between oral anticoagulation and aspirin or further individual risk stratification because the bleeding risk with oral anticoagulation may outweigh the benefit. There is nevertheless an appreciable risk for stroke (23): for example, the 5-year risk for stroke in patients with a CHADS 2 score of 1 is about 9% 664 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 www.annals.org

Dabigatran vs. Warfarin and CHADS 2 in Atrial Fibrillation Original Research Table 2 Continued Dabigatran, 110 mg, vs. Warfarin Hazard Ratio (95% CI) Dabigatran, 150 mg, vs. Warfarin 0.98 (0.63 to 1.51) 0.61 (0.37 to 0.99) 1.06 (0.74 to 1.52) 0.61 (0.40 to 0.92) 0.78 (0.58 to 1.04) 0.69 (0.51 to 0.93) 0.37 0.84 0.65 (0.49 to 0.88) 0.74 (0.56 to 0.99) 0.90 (0.71 to 1.14) 0.92 (0.72 to 1.17) 0.83 (0.66 to 1.03) 1.07 (0.87 to 1.31) 0.26 0.14 0.37 (0.16 to 0.83) 0.37 (0.16 to 0.84) 0.32 (0.15 to 0.68) 0.38 (0.19 to 0.76) 0.24 (0.12 to 0.48) 0.48 (0.28 to 0.82) 0.70 0.82 0.87 (0.60 to 1.25) 0.68 (0.46 to 1.01) 0.99 (0.76 to 1.29) 0.78 (0.59 to 1.03) 0.86 (0.68 to 1.09) 0.97 (0.77 to 1.22) 0.73 0.22 0.88 (0.66 to 1.16) 0.73 (0.54 to 0.98) 0.98 (0.79 to 1.21) 0.81 (0.65 to 1.01) 0.87 (0.72 to 1.06) 1.02 (0.84 to 1.23) 0.72 0.12 and about 5% in those with a CHADS 2 score of 0. Two recent studies (28, 29) focusing on patients with a CHADS 2 score of 1 have shown a substantial reduction in stroke with vitamin K antagonists compared with antiplatelet or no antithrombotic treatment, although this is counterbalanced by a trend to increased major bleeding (29). In the present trial in patients with a CHADS 2 score of 0 to 1 (predominantly a score of 1), dabigatran, 150 mg twice daily, was superior for stroke prevention, and both dabigatran doses were associated with reductions in intracranial bleeding compared with warfarin. This study has several limitations. The subgroup analyses were not prespecified and should therefore be deemed exploratory. Given the large number of subgroup comparisons and the small numbers of events within each subgroup for some outcomes, the levels of statistical significance for interaction tests must be interpreted with considerable caution. There are several other stroke prediction schemes, including the CHA 2 DS2VASc score (30). However, all of them have, at best, only a modest discriminating ability for the individual patients and are in this respect not very different from the CHADS 2 risk score (23, 30, 31). The CHA 2 DS2VASc score extends the CHADS 2 risk score by considering additional moderate stroke risk factors, such as age between 65 and 74 years, female sex, and vascular disease, and also assigns 2 points for age 75 years or older. Thus, 99.8% of the RE-LY participants had acha 2 DS2VASc score of 1 or greater, in whom oral anticoagulation should be considered according to the recent European guidelines for treatment of atrial fibrillation Figure 2. Annual rates of major and intracranial bleeding in relation to CHADS 2 risk scores. 12 10 Intracranial bleeding Major bleeding Percentage per Year 8 6 4 2 0 0 1 2 3 4 5 6 CHADS 2 Score Patients, n 452 5323 6455 3654 1619 524 85 Major/intracranial bleeding, n/n* 15/1 248/35 392/50 277/11 159/24 63/11 8/0 Cochran Armitage linear trend test P 0.001 for both major and intracranial bleeding. Error bars represent 95% CIs. * Total number of major/intracranial bleeding episodes (median follow-up, 2 y). www.annals.org 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 665

Original Research Dabigatran vs. Warfarin and CHADS 2 in Atrial Fibrillation (14). However, it is not possible to draw conclusions on the benefits of oral anticoagulants compared with placebo or platelet inhibitors on the basis of the present study, in which all patients received oral anticoagulants. Furthermore, the group of very-low-risk patients with a CHADS 2 score of 0 was small. Increasing CHADS 2 scores were associated with increased risks for stroke or systemic embolism, major and intracranial bleeding, and death in patients with atrial fibrillation treated with the oral vitamin K antagonist warfarin or with 1 of 2 doses of the oral direct thrombin inhibitor dabigatran, 110 or 150 mg twice daily. Rates of stroke or systemic embolism were lower with dabigatran, 150 mg twice daily, and rates of intracranial bleeding were lower with both dabigatran doses than with warfarin treatment, without significant heterogeneity in subgroups defined by CHADS 2 scores. From Uppsala Clinical Research Centre and Department of Medical Sciences, Uppsala University, Uppsala, Sweden; Working Group on Cardiovascular Research, Utrecht, the Netherlands; Vivantes Klinikum Neukölln, Berlin, Germany; University Duisburg-Essen, Duisburg and Essen, Germany; Population Health Research Institute and McMaster University, Hamilton, Ontario, Canada; Lankenau Institute for Medical Research and the Heart Center, Wynnewood, Pennsylvania; University Hospital Bratislava, Bratislava, Slovakia; and Boehringer Ingelheim Pharmaceuticals, Ridgefield, Connecticut. Grant Support: By Boehringer Ingelheim. Potential Conflicts of Interest: Disclosures can be viewed at www.acponline.org/authors/icmje/conflictofinterestforms.do?msnum M11-0189. Reproducible Research Statement: Study protocol: Main RE-LY study protocol available from Dr. Oldgren (e-mail, Jonas.Oldgren@ucr.uu.se). Statistical code and data set: Not available. Corresponding Author: Jonas Oldgren, MD, PhD, Uppsala Clinical Research Center, Uppsala University, S-75185 Uppsala, Sweden; e-mail, Jonas.Oldgren@ucr.uu.se. Current author addresses and author contributions are available at www.annals.org. References 1. Hart RG, Pearce LA, Rothbart RM, McAnulty JH, Asinger RW, Halperin JL. Stroke with intermittent atrial fibrillation: incidence and predictors during aspirin therapy. Stroke Prevention in Atrial Fibrillation Investigators. J Am Coll Cardiol. 2000;35:183-7. [PMID: 10636278] 2. Hohnloser SH, Pajitnev D, Pogue J, Healey JS, Pfeffer MA, Yusuf S, et al; ACTIVE W Investigators. Incidence of stroke in paroxysmal versus sustained atrial fibrillation in patients taking oral anticoagulation or combined antiplatelet therapy: an ACTIVE W Substudy. J Am Coll Cardiol. 2007;50:2156-61. [PMID: 18036454] 3. Friberg L, Hammar N, Rosenqvist M. Stroke in paroxysmal atrial fibrillation: report from the Stockholm Cohort of Atrial Fibrillation. Eur Heart J. 2010;31: 967-75. [PMID: 19176537] 4. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med. 1994;154:1449-57. [PMID: 8018000] 5. Stroke Risk in Atrial Fibrillation Working Group. Independent predictors of stroke in patients with atrial fibrillation: a systematic review. Neurology. 2007; 69:546-54. [PMID: 17679673] 6. Healey JS, Hart RG, Pogue J, Pfeffer MA, Hohnloser SH, De Caterina R, et al. Risks and benefits of oral anticoagulation compared with clopidogrel plus aspirin in patients with atrial fibrillation according to stroke risk: the atrial fibrillation clopidogrel trial with irbesartan for prevention of vascular events (ACTIVE-W). Stroke. 2008;39:1482-6. [PMID: 18323500] 7. Stroke Risk in Atrial Fibrillation Working Group. Comparison of 12 risk stratification schemes to predict stroke in patients with nonvalvular atrial fibrillation. Stroke. 2008;39:1901-10. [PMID: 18420954] 8. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA. 2001;285:2864-70. [PMID: 11401607] 9. Baruch L, Gage BF, Horrow J, Juul-Möller S, Labovitz A, Persson M, et al. Can patients at elevated risk of stroke treated with anticoagulants be further risk stratified? Stroke. 2007;38:2459-63. [PMID: 17673721] 10. Crandall MA, Horne BD, Day JD, Anderson JL, Muhlestein JB, Crandall BG, et al. Atrial fibrillation significantly increases total mortality and stroke risk beyond that conveyed by the CHADS2 risk factors. Pacing Clin Electrophysiol. 2009;32:981-6. [PMID: 19659615] 11. Henriksson KM, Farahmand B, Johansson S, Asberg S, Terént A, Edvardsson N. Survival after stroke the impact of CHADS2 score and atrial fibrillation. Int J Cardiol. 2010;141:18-23. [PMID: 19144430] 12. Fuster V, Rydén LE, Cannom DS, Crijns HJ, Curtis AB, Ellenbogen KA, et al; Task Force on Practice Guidelines, American College of Cardiology/ American Heart Association. ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation-executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines (Writing Committee to Revise the 2001 Guidelines for the Management of Patients with Atrial Fibrillation). Eur Heart J. 2006;27:1979-2030. [PMID: 16885201] 13. Singer DE, Albers GW, Dalen JE, Fang MC, Go AS, Halperin JL, et al; American College of Chest Physicians. Antithrombotic therapy in atrial fibrillation: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest. 2008;133:546S-592S. [PMID: 18574273] 14. Camm AJ, Kirchhof P, Lip GY, Schotten U, Savelieva I, Ernst S, et al; European Heart Rhythm Association. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur Heart J. 2010;31:2369-429. [PMID: 20802247] 15. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, et al; RE-LY Steering Committee and Investigators. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361:1139-51. [PMID: 19717844] 16. Ezekowitz MD, Connolly S, Parekh A, Reilly PA, Varrone J, Wang S, et al. Rationale and design of RE-LY: randomized evaluation of long-term anticoagulant therapy, warfarin, compared with dabigatran. Am Heart J. 2009;157:805-10, 810.e1-2. [PMID: 19376304] 17. Connolly SJ, Ezekowitz MD, Yusuf S, Reilly PA, Wallentin L; Randomized Evaluation of Long-Term Anticoagulation Therapy Investigators. Newly identified events in the RE-LY trial [Letter]. N Engl J Med. 2010;363:1875-6. [PMID: 21047252] 18. Rosendaal FR, Cannegieter SC, van der Meer FJ, Briët E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost. 1993;69:236-9. [PMID: 8470047] 19. Ruiz Ortiz M, Romo E, Mesa D, Delgado M, Anguita M, Castillo JC, et al. Oral anticoagulation in nonvalvular atrial fibrillation in clinical practice: impact of CHADS(2) score on outcome. Cardiology. 2010;115:200-4. [PMID: 20160440] 20. Hylek EM, Evans-Molina C, Shea C, Henault LE, Regan S. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation. 2007;115:2689-96. [PMID: 17515465] 21. Poli D, Antonucci E, Marcucci R, Fatini C, Alterini B, Mannini L, et al. Risk of bleeding in very old patients with atrial fibrillationon warfarin: relationship with ageing and CHADS2 score. Thromb Res. 2007;121:347-52. [PMID: 17597186] patients with atrial fibrillation. 666 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 www.annals.org

Dabigatran vs. Warfarin and CHADS 2 in Atrial Fibrillation Original Research 22. Nieuwlaat R, Capucci A, Lip GY, Olsson SB, Prins MH, Nieman FH, et al; Euro Heart Survey Investigators. Antithrombotic treatment in real-life atrial fibrillation patients: a report from the Euro Heart Survey on Atrial Fibrillation. Eur Heart J. 2006;27:3018-26. [PMID: 16731536] 23. Rietbrock S, Heeley E, Plumb J, van Staa T. Chronic atrial fibrillation: Incidence, prevalence, and prediction of stroke using the Congestive heart failure, Hypertension, Age 75, Diabetes mellitus, and prior Stroke or transient ischemic attack (CHADS2) risk stratification scheme. Am Heart J. 2008;156:57-64. [PMID: 18585497] 24. Zimetbaum PJ, Thosani A, Yu HT, Xiong Y, Lin J, Kothawala P, et al. Are patients with atrial fibrillation receiving warfarin in accordance with stroke risk? Am J Med. 2010;123:446-53. [PMID: 20399322] 25. Bungard TJ, Ghali WA, Teo KK, McAlister FA, Tsuyuki RT. Why do patients with atrial fibrillation not receive warfarin? Arch Intern Med. 2000;160: 41-6. [PMID: 10632303] 26. Devereaux PJ, Anderson DR, Gardner MJ, Putnam W, Flowerdew GJ, Brownell BF, et al. Differences between perspectives of physicians and patients on anticoagulation in patients with atrial fibrillation: observational study. BMJ. 2001;323:1218-22. [PMID: 11719412] 27. Gattellari M, Worthington J, Zwar N, Middleton S. Barriers to the use of anticoagulation for nonvalvular atrial fibrillation: a representative survey of Australian family physicians [Letter]. Stroke. 2008;39:227-30. [PMID: 18048861] 28. Gorin L, Fauchier L, Nonin E, de Labriolle A, Haguenoer K, Cosnay P, et al. Antithrombotic treatment and the risk of death and stroke in patients with atrial fibrillation and a CHADS2 score 1. Thromb Haemost. 2010;103:833-40. [PMID: 20135077] 29. Lee BH, Park JS, Park JH, Park JS, Kwak JJ, Hwang ES, et al. The effect and safety of the antithrombotic therapies in patients with atrial fibrillation and CHADS score 1. J Cardiovasc Electrophysiol. 2010;21:501-7. [PMID: 20021521] 30. Lee BH, Park JS, Park JH, Park JS, Kwak JJ, Hwang ES, et al. The effect and safety of the antithrombotic therapies in patients with atrial fibrillation and CHADS score 1. J Cardiovasc Electrophysiol. 2010;21:501-7. [PMID: 20021521] 31. Fang MC, Go AS, Chang Y, Borowsky L, Pomernacki NK, Singer DE; ATRIA Study Group. Comparison of risk stratification schemes to predict thromboembolism in people with nonvalvular atrial fibrillation. J Am Coll Cardiol. 2008;51:810-5. [PMID: 18294564] READERS COMMENTS Readers wishing to comment on published articles should use the Comments: Submit a comment option at www.annals.org. While this service is free to Annals subscribers, readers without subscriptions who wish to comment on articles may purchase temporary access. www.annals.org 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 667

Annals of Internal Medicine Current Author Addresses: Drs. Oldgren and Wallentin: Uppsala Clinical Research Center, Uppsala University, S-75185 Uppsala, Sweden. Dr. Alings: Amphia Ziekenhuis, Molengracht 21, Breda, 4818 CK, the Netherlands. Drs. Darius: Vivantes Klinikum Neukölln, Rudower Str. 48, 12351 Berlin, Germany. Dr. Diener: Department of Neurology, University of Essen, Hufelandstrasse 55, Essen D45122, Germany. Drs. Eikelboom, Yusuf, and Connolly and Mr. Yang: Population Health Research Institute, McMaster University, 237 Barton Street East, Hamilton, Ontario L8L 2X2, Canada. Dr. Ezekowitz: Main Line Health Interventional Cardiology, Suite 380 Medical Science Building, 100 East Lancaster Avenue, Wynnewood, PA 19096. Dr. Kamensky: Department of Noninvasive Cardiovascular Diagnostics, University Hospital Bratislava, Ruzinovska 26, Bratislava 82606, Slovakia. Dr. Reilly: Boehringer Ingelheim Pharmaceuticals, 900 Ridgebury Road, PO Box 368, Ridgefield, CT 06877. Author Contributions: Conception and design: J. Oldgren, H.C. Diener, M.D. Ezekowitz, P.A. Reilly, S. Yusuf, L. Wallentin, S.J. Connolly. Analysis and interpretation of the data: J. Oldgren, H.C. Diener, J. Eikelboom, M.D. Ezekowitz, S. Yang, S. Yusuf, L. Wallentin, S.J. Connolly. Drafting of the article: J. Oldgren. Critical revision of the article for important intellectual content: J. Oldgren, M. Alings, H. Darius, J. Eikelboom, M.D. Ezekowitz, G. Kamensky, P.A. Reilly, S. Yusuf, L. Wallentin, S.J. Connolly. Final approval of the article: J. Oldgren, M. Alings, H. Darius, H.C. Diener, J. Eikelboom, M.D. Ezekowitz, P.A. Reilly, S. Yusuf, L. Wallentin, S.J. Connolly. Provision of study materials or patients: S.J. Connolly. Statistical expertise: S. Yang. Obtaining of funding: M.D. Ezekowitz, P.A. Reilly, L. Wallentin, S.J. Connolly. Administrative, technical, or logistic support: S. Yusuf, S.J. Connolly. Collection and assembly of data: J. Oldgren, M.D. Ezekowitz, P.A. Reilly, S. Yusuf, L. Wallentin, S.J. Connolly. W-204 15 November 2011 Annals of Internal Medicine Volume 155 Number 10 www.annals.org