- Email: [email protected]
Significant Underuse of Warfarin in Patients with Nonvalvular Atrial Fibrillation: Results from the China National Stroke Registry
Significant Underuse of Warfarin in Patients with Nonvalvular Atrial Fibrillation: Results from the China National Stroke Registry Chunjuan Wang, MD,* Zhonghua Yang, MD,* Chunxue Wang, MD,* Yilong Wang, MD,* Xingquan Zhao, MD,* Liping Liu, MD,* David Z. Wang, DO,† Hao Li, PhD,* and Yongjun Wang, MD*
Background: Warfarin reduces the risk of stroke in patients with atrial fibrillation (AF) but is often underused in clinical practice. We aimed to examine the current state of warfarin use in nonvalvular atrial fibrillation (NVAF) patients with firstever ischemic stroke (IS) or transient ischemic attack (TIA) in China and to analysis factors causing such underuse. Methods: By accessing the China National Stroke Registry, data on consecutive patients with known NVAF who developed firstever IS or TIA were studied. Proportion of warfarin use was estimated in those patients with known NVAF who should be eligible for anticoagulation therapy. Factors associated with warfarin underuse were identified by multivariate logistic regression. Results: Of 11,080 patients with first-ever IS or TIA, 996 (9.7%) had NVAF and no contraindications to anticoagulation therapy and 592 of them had history of AF. Among these patients, only 96 (16.2%) were on warfarin and 496 (83.8%) were not. In those patients on warfarin, only 1 of his international normalized ratios on admission was in the therapeutic range of 2.0-3.0. Based on the CHADS2 scores, about 20.2% low-risk AF patients took warfarin; however, only 15.2% moderate and 16.4% high-risk patients were on warfarin. Older patients and patients with history of coronary heart disease (CHD) were less likely to be given warfarin, whereas patients with history of prestroke antiplatelet use were more likely to take warfarin. Conclusions: We found that warfarin was significantly underused in patients with known NVAF in China. Age and CHD and prestroke antiplatelet therapy were related factors. Key Words: Anticoagulation—atrial fibrillation—warfarin—stroke management—transient ischemic attack. Ó 2014 by National Stroke Association
Introduction Atrial fibrillation (AF) is a common arrhythmia not only in western countries but also in China. It causes about
From the *Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China; and †INI Stroke Center and Stroke Network, OSF Healthcare System, University of Illinois College of Medicine, Peoria, Illinois. Received August 30, 2013; revision received October 7, 2013; accepted October 9, 2013. Grant support: This study was funded by the Ministry of Science and Technology and the Ministry of Health of the People’s Republic of China. The grant numbers are National S & T Major Project of China (2008ZX09312-008), State Key Development Program of Basic Research of China (2009CB521905), the Research Special Fund For
15% of ischemic stroke (IS) each year.1 Since 1990s, studies have shown that anticoagulation with warfarin can effectively prevent IS in nonvalvular atrial fibrillation (NVAF) patients.2,3 However, despite its clear benefit for stroke Public Welfare Industry of Health (200902004), and the National 12th Five-year S & T Major Project (2011BAI08B02). The authors report no conflicts of interest. Address correspondence to Yongjun Wang, MD, Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, No. 6 Tiantanxili, Dongcheng District, Beijing 100050, China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2013.10.006
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 5 (May-June), 2014: pp 1157-1163
1157
C. WANG ET AL.
1158
prevention, warfarin has been underprescribed in clinical practice.4,5 Zhou et al6 have found that warfarin use was only about 2.7% in AF patients, and Qi7 have revealed that the anticoagulation rate was only 6.6% in hospitalized patients with AF in China. Most of these studies were based on AF population in general and few of them focused on NVAF patients who went on and developed first-ever IS or TIA. Furthermore, it has not been well studied about the factors that may contribute to warfarin underuse in this AF population. To evaluate this important issue of warfarin underuse and further determine the factors associated with the formation of such gap in China, we analyzed the data from the China National Stroke Registry (CNSR).
Materials and Methods Design and Setting CNSR is a national hospital-based and prospective registry sponsored by the Ministry of Health. The primary objective of the registry is to evaluate the quality of care during acute hospitalization of all stroke patients. The registry collects data on many performance measurements on the delivery of stroke care around the nation. A total of 132 hospitals representing 27 provinces and 4 municipalities (except for Taiwan and Macao regions) in Mainland China have been selected as registry hospitals. Among them, 100 are tertiary hospitals and 32 are secondary urban hospitals. Patients were eligible to be enrolled in the registry if they met the following criteria: 18 years of age or more, IS, TIA, intracranial hemorrhage (ICH), or subarachnoid hemorrhage confirmed by brain CT or MRI within 14 days after the onset of symptoms. Trained research co-ordinators at each institute reviewed medical records daily to identify, consent, and enroll qualified patients. Standardized data collection form was used by all sites. Trained research co-ordinators obtained these data directly through interviewing the patients and family: prehospital care, prestroke modified Rankin Scale, and baseline National Institutes of Health Stroke Scale and Glasgow Coma Scale scores. Other data extracted from the medical records included patient demographics, medical history, medication use, IS subtypes, ICH and causes, vascular malformation, coagulopathy, disease status (ie, National Institutes of Health Stroke Scale, neuroimaging studies), complications in hospital, disease management in different hospital settings, and discharge status. Patients were contacted by telephone at 3, 6, 12, 18, and 24 months after discharge by trained research personnel at Beijing Tiantan Hospital. Throughout the study period, all data elements were manually checked by a research specialist from an independent research organization. A professional data processing company was responsible for the computer data entry. Beijing Translational Medicine Research Center, an independent research organiza-
tion, served as the data analysis. Other details in design and rationale of CNSR have been described elsewhere.8,9
Eligibility and Criteria For the current study, we identified all patients who presented with TIA or first-ever IS. AF was defined as previously diagnosed AF documented in the medical record. Patients with newly detected AF on admission and those diagnosed with ICH or SAH or those with a history of IS were excluded. Patients with history of valvular heart disease or contraindications to warfarin use were also excluded. Other contraindications that have excluded the patients included metastatic cancer, dementia, cirrhosis, renal failure requiring dialysis, previous gastrointestinal bleed, and peptic ulcer diseases. Warfarin use was determined by the recorded prescriptions provided by the patient or his/her caregiver. The target population was divided into 2 groups: NVAF not-on warfarin and NVAF on warfarin. The baseline characteristics of these 2 groups included age, gender, marital status, body mass index, educational level (elementary or less, namely #6 years education; middle school, 6-9 years; and high school or above, namely .9 years), health insurance type (urban or governmental, rural, commercial or other, and self-payment), and geographic regions (eastern, central, or western according to the geographic location of the hospitals from which the patients were recruited). Data on risk factors and comorbidities of stroke included currently smoking (continuous or accumulated smoking for at least 6 months and $1 cigarette per day and had smoked at least once within 30 days before the stroke), heavy drinking ($5 standard alcoholic drinks per day), hypertension, hyperlipidemia, congestive heart failure (with definitive diagnosis on record), coronary heart disease (CHD, including silent myocardial ischemia, angina pectoris, myocardial infarction, or ischemic cardiomyopathy), peripheral vascular diseases (defined as peripheral vascular or lymph diseases including thromboangiitis obliterans, arteriosclerotic obliterans, arterial embolism, polyarteritis, Raynaud’s syndrome, varicose veins, or deep venous thrombosis), and diabetes mellitus (both types 1 and 2). The risk of stroke was estimated by using the CHADS2 score,10 which was derived from congestive heart failure (1 point), hypertension (1 point), age of 75 years or more (1 point), diabetes mellitus (1 point), and previous stroke or TIA (2 points). According to this scoring system, patients were categorized into 3 groups: low risk (0), moderate risk (1-2), and high risk ($3). Data from eligible patients for current analysis were obtained from the CNSR database after all information has been entered.
Statistical Analysis For descriptive analysis, proportions were used for categorical variables, and means with standard deviations
UNDERUSE OF WARFARIN IN CHINA
1159
Figure 1. Flow diagram defining patients with known NVAF but no contraindications of warfarin use from CNSR. Abbreviations: IS, ischemic stroke; NVAF, nonvalvular atrial fibrillation; TIA, transient ischemic attack; WFR, warfarin.
were used for continuous variables. The baseline characteristic differences between NVAF not-on warfarin group and NVAF on warfarin group were analyzed using nonparametric Kruskal–Wallis test for continuous variables, and the c2 or Fisher exact test were used for categorical variables. Univariate logistic regression analysis was used to identify the demographic and clinical factors associated with the underuse of warfarin such as age, gender, CHADS2 score, education, geographic region, etc. Variables (P , .20) were entered into a multivariate regression model to identify the independent factors related to warfarin underuse with significance level of P less than .05. The odds ratios (ORs) and 95% confidence interval (CI) of both univariate and multivariate analysis were reported and a 95% CI that did not contain 1.0 was considered to be statistically significant. All analyses were conducted with a commercially available software package (SAS version 9.1.3; SAS Institute, Inc.).
Results When the data bank was closed, 22,216 patients have been entered into the CNSR. Among them, 11,080 had first-ever IS or TIA without history of valvular heart disease. Of these patients, 793 were excluded with contraindications of using oral anticoagulant and 996 (9.7%) had
NVAF. After excluding 404 patients with newly diagnosed AF, a total of 496 (83.8%) patients in the AF noton warfarin group and 96 (16.2%) in the AF on warfarin group were entered into the analysis (Fig 1).
Patient Characteristics The average age in the AF on warfarin group (63.7 6 12.7 years) was younger than that of the AF without warfarin group (72.1 6 10.5 years, P , .0001). For subjects with known AF who were eligible for anticoagulation therapy, the rates of warfarin use decreased with older ages. There were more female patients in the not-on warfarin group (P 5 .0264). CHD showed a rather lower proportion in the AF on warfarin group (30.2%) compared with AF not-on warfarin group (42.7%, P 5 .0221). Based on the CHADS2 score, 483 of 592 (81.6%) AF patients had at least 1 risk factor of stroke, but 84.7% of them (409 of 483) were not on warfarin. As for combining warfarin and antiplatelet together before stroke, there were more patients taking antiplatelet agents in AF on warfarin group (59.4%) than in AF not-on warfarin group (19.6%, P , .0001). Of those on warfarin, about 42.1% had high school or higher education, compared with only 22.3% in the not-on warfarin group (P , .0001). As for different health insurance type, no significant difference
C. WANG ET AL.
1160
Table 1. Baseline characteristics for patients with known NVAF and no contraindications of warfarin therapy categorized by warfarin use Patients with known NVAF no contraindications
Age (mean, y) 18-45 46-65 66-75 .75 Female, n (%) Marital status Single Married Divorced/widowed/remarried Lives Alone With others In nursing home Body mass index ,25 25-30 $30 Current smoker Heavy drinking Hypertension Hyperlipidemia Diabetes mellitus Coronary heart diseases Congestive heart failure Peripheral vascular diseases CHADS2 (0-6 points) Low risk: 0 Moderate risk: 1-2 High risk: $3 Drug administration prestroke Antiplatelet agents Antihypertensive agents Antihyperlipidemia agents Insulin or oral hypoglycemic agents Education Elementary or below Middle school High school or above Health insurance type Urban or governmental Rural Commercial or other Self-payment Geographic region Eastern Central Western
Overall (n 5 592)
No warfarin (n 5 496)
With warfarin (n 5 96)
P value
70.7 6 11.3 15 (2.5%) 159 (26.9%) 194 (32.8%) 224 (37.8%) 308 (52.0%)
72.1 6 10.5 8 (1.6%) 111 (22.4%) 173 (34.9%) 204 (41.1%) 268 (54.0%)
63.7 6 12.7 7 (7.3%) 48 (50.0%) 21 (21.9%) 20 (20.8%) 40 (41.7%)
,.0001 ,.0001
4 (.7%) 501 (85.1%) 84 (14.3%)
3 (.6%) 415 (84.2%) 75 (15.2%)
1 (1.0%) 86 (89.6%) 9 (9.4%)
.2985
24 (4.1%) 561 (95.7%) 1 (.2%)
21 (4.3%) 469 (95.5%) 1 (.2%)
3 (3.2%) 92 (96.8%) 0 (.0%)
.7979
304 (61.3%) 164 (33.1%) 28 (5.6%) 161 (27.3%) 37 (6.3%) 373 (63.0%) 63 (10.6%) 111 (18.8%) 241 (40.7%) 60 (10.1%) 7 (1.2%)
255 (62.3%) 129 (31.5%) 25 (6.1%) 125 (25.3%) 30 (6.0%) 308 (62.1%) 47 (9.5%) 89 (17.9%) 212 (42.7%) 53 (10.7%) 4 (.8%)
49 (56.3%) 35 (40.2%) 3 (3.4%) 36 (37.5%) 7 (7.3%) 65 (67.7%) 16 (16.7%) 22 (22.9%) 29 (30.2%) 7 (7.3%) 3 (3.1%)
.2270 . . .0141 .6451 .2972 .0365 .2532 .0221 .3132 .0544
109 (18.4%) 428 (72.3%) 55 (9.3%)
87 (17.5%) 363 (73.2%) 46 (9.3%)
22 (22.9%) 65 (67.7%) 9 (9.4%)
.4499
154 (26.0%) 285 (48.1%) 25 (4.2%) 87 (14.7%)
97 (19.6%) 247 (49.8%) 15 (3.0%) 68 (13.7%)
57 (59.4%) 38 (39.6%) 10 (10.4%) 19 (19.8%)
,.0001 .0667 .0010 .1234
304 (52.1%) 130 (22.3%) 149 (25.6%)
274 (56.1%) 105 (21.5%) 109 (22.3%)
30 (31.6%) 25 (26.3%) 40 (42.1%)
,.0001
374 (63.2%) 89 (15.0%) 18 (3.0%) 111 (18.8%)
309 (62.3%) 78 (15.7%) 16 (3.2%) 93 (18.8%)
65 (67.7%) 11 (11.5%) 2 (2.1%) 18 (18.8%)
.6387
413 (69.8%) 86 (14.5%) 93 (15.7%)
356 (71.8%) 66 (13.3%) 74 (14.9%)
57 (59.4%) 20 (20.8%) 19 (19.8%)
.0468
Abbreviation: NVAF, nonvalvular atrial fibrillation.
.0264
UNDERUSE OF WARFARIN IN CHINA
1161
Figure 2. Comparison of warfarin use in different AF populations. ‘‘*,’’ An American retrospective cohort study with total number of 13,709 recruited AF outpatients and 10.2% with a history of stroke. ‘‘†,’’ A subpopulation of Euro Heart Survey with total number of 2706 recruited patients and 336 (12.4%) had a prior stroke or TIA. ‘‘‡,’’ A study based on data of the REACH Registry with total number of AF patients 6814 and 43.67% had a prior stroke or TIA. ‘‘x,’’ A subpopulation from phase 3 of the RCSN. Here we used only the subgroup (n 5 597) data of first-ever ischemic stroke patients for comparison. Abbreviations: AP, antiplatelet; OAC, oral anticoagulation; RCSN, Registry of the Canadian Stroke Network; REACH, Reduction of Atherothrombosis for Continued Health. (Color version of figure is available online.)
was found between AF on warfarin and not-on warfarin group (P 5 .6387), but as for geographic region, a significant difference among eastern, central, and western was detected between both groups (P 5 .0468; Table 1).
Current State of Warfarin Use For patients with known NVAF and without any contraindication, only 16.2% received warfarin (9.6% on both warfarin and an antiplatelet agent), 16.4% were on antiplatelet alone, and 67.4% were on neither warfarin nor antiplatelet agents (Fig 2). The international normalized ratio (INR) was in the desired range of 2.0-3.0 in only 1 of 96 NVAF patients on warfarin while the INR in majority (55.2%) was below this therapeutic range (INR , 2.0) and the INRs of the other 40 patients (41.7%) were missing. About 20.2% of 109 patients with low CHADS2 scores were on warfarin while 15.2% with moderate and 16.4% with high CHADS2 scores were on warfarin (Fig 3).
Factors Related to Warfarin Use In patients with known NVAF, the ORs of warfarin use gradually reduced with older age. The lowest use was
Figure 3. Proportions of warfarin use with increased risk of stroke based on CHADS2 score in different AF populations. *, †, and ‡, please refer to the footnote of Figure 2. Note: CHADS2 score in the US study by Boulanger et al12 was classified as low risk (0-1), moderate risk (2-3), and high risk (.4), which was a small difference from the other 3 studies. Abbreviation: AF, atrial fibrillation. (Color version of figure is available online.)
among those older than 75 years (OR 5 .05; 95% CI, .01.19, age of 18-45 years as a reference). AF patients with CHD were much less likely to be on warfarin (OR 5 .47; 95% CI, .27-.83). However, patients with antiplatelet agents before stroke were more likely to receive warfarin (OR 5 6.99; 95% CI, 4.08-11.99). Gender, marital status, CHADS2 score, educational level, health insurance type, geographic region, and other clinical characteristics did not have significant impact on the use of warfarin (Fig 4).
Discussion From this study, we found several concerns in AFrelated stroke prevention in China. First of all, warfarin was significantly underused in patients with known NVAF and developed first-ever IS or TIA. Only about one sixth were on warfarin for primary prevention. Furthermore, many patients (55.2%) who were on warfarin did not reach the therapeutic range. Overall, approximately 90% of first-ever IS patients with known NVAF were not appropriately anticoagulated before their catastrophic event. This undertreatment with warfarin as seen in CNSR was similar to the findings in the Registry of Canadian Stroke Network.11 Comparing with other 3
C. WANG ET AL.
1162
Figure 4. Uni- and multivariate logistic regression analysis for factors related to warfarin use. Abbreviation: OR, odds ratio.
well-known AF research in the United States,12 Europe,13 and the Reduction of Atherothrombosis for Continued Health Registry,14 the rate of warfarin use in China was much lower (Fig 2). Significant numbers of strokes could be prevented in AF patients if these patients were appropriately anticoagulated. Second, AF patients with higher CHADS2 scores and increased stroke risks had less chance to be on warfarin in CNSR. As we know, warfarin can be given to patients with moderate risk and should be given to those with high risk. Some recent studies even suggest that patients with CHADS2 score of 1 should receive warfarin rather than aspirin.15,16 However, the proportions of warfarin use in CNSR were unexpectedly decreased with increased stroke risks. This finding was very different from the other studies not only in the percentage but also in the trend (Fig 3). There was a big gap between understanding the evidence-based guidelines and practicing accordingly among Chinese physicians. It shows that much work needs to be done to improve the compliance with the guidelines. Third, both age and CHD were independent risk factors of lower use of warfarin, whereas prestroke antiplatelet agent use was associated with higher warfarin treatment. Although patients, treating physicians, and regional health care system may have an impact on the use of anticoagulation,11,17 our discovery of the impact of patient’s older age associated with the unlikelihood of being on warfarin was not unique to China.18 Unfortunately, many physicians were reluctant to prescribe warfarin especially for older patients because they overestimated the risk of bleeding and underestimated the risk of stroke and the benefit of warfarin,19 even though age itself should not be regarded as a contraindication to anticoagulation.20 It seemed that Chinese AF patients with CHD were less likely to receive warfarin. This might be because of the use of antiplatelet agents for CHD prevention in these patients who fear to be on warfarin. Lastly, the INR was inappropriately monitored in AF patients on warfarin in China. More than two fifths (41.7%) of them had no INRs. One of the reasons was that INR was not checked on admission, an indication
that the treating physicians did not pay enough attention to INR monitoring. In China, checking INR is rather expensive and mainly done in the tertiary hospitals. This could be another reason why warfarin may be underprescribed. Although portable INR monitor is now available in China, it is too expensive to be used widely. We also found surprisingly that prestroke antiplatelet treatment was associated with better warfarin utilization. Combining antiplatelet agents with warfarin are recommended in those AF patients with acute myocardial infarction for at least 3 months and those with mechanical or bioprosthetic heart valves.21 An interesting phenomenon is that Chinese physicians often prescribe lower dose of warfarin and yet add an antiplatelet agent such as aspirin to the regimen. The common belief is that lower doses of warfarin may avoid hemorrhage and yet provide an anticoagulation effect. This unrealistic fear of bleeding is popular among Chinese physicians even though they fully understand what the target INR should be. There are several limitations of our study. Because only patients admitted to the neurology departments were captured in CNSR and no patients from cardiology departments were included, the prevalence of AF and rate of anticoagulation could be underestimated. This study may have also underestimated the magnitude of warfarin underuse because the data in CNSR were from tertiary or secondary hospitals, and rural hospitals were excluded. Furthermore, our study only examined the warfarin use in NVAF patients with the first-ever IS or TIA and not with other medical conditions.
Conclusions China faces significant challenges in stroke and TIA prevention in patients with NVAF. Many factors impede the use of warfarin in AF patients who have no obvious contraindications. Appropriate use of warfarin may significantly decrease first-ever IS or TIA in patients with AF. The main obstacle seems to be the belief held by the Chinese physicians that warfarin may cause excessive unforeseen hemorrhage. With the aging population1 and likelihood of increased prevalence of AF in China, it
UNDERUSE OF WARFARIN IN CHINA
has become urgent that effective strategies are needed to address these deficiencies. It is rather clear that improving stroke prevention is now a pressing issue not only for Chinese physicians but also for the Chinese government. Acknowledgment: We thank all participating hospitals, colleagues, nurses, and imaging and laboratory technicians.
1163
11.
12.
13.
References 1. Wolf P, Abbott R, Kannel W. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke 1991;22:983-988. 2. The Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators. The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. N Engl J Med 1990;323:1505-1511. 3. Ezekowitz MD, Bridgers SL, James KE, et al. Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. N Engl J Med 1992;327:1406-1412. 4. Ogilvie IM, Newton N, Welner SA, et al. Underuse of oral anticoagulants in atrial fibrillation: a systematic review. Am J Med 2010;123:638-645.e634. 5. Mercaldi CJ, Ciarametaro M, Hahn B, et al. Cost efficiency of anticoagulation with warfarin to prevent stroke in medicare beneficiaries with nonvalvular atrial fibrillation. Stroke 2011;42:112-118. 6. Zhou ZQ, Hu DY, Chen J, et al. An epidemiological survey of atrial fibrillation in China. Chin J Intern Med 2004;43:491-494. 7. Qi WH. Retrospective investigation of hospitalized patients with atrial fibrillation in major parts of China. Chin J Cardiol 2003;31:913-915. 8. Wang Y, Cui L, Ji X, et al. The China National Stroke Registry for patients with acute cerebrovascular events: design, rationale, and baseline patient characteristics. Int J Stroke 2011;6:355-361. 9. Li S, Zhao X, Wang C, et al. Risk factors for poor outcome and mortality at 3 months after the ischemic stroke in patients with atrial fibrillation. J Stroke Cerebrovasc Dis 2013;13:146-148. 10. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification schemes for predicting stroke: re-
14.
15.
16.
17.
18.
19.
20.
21.
sults from the National Registry of Atrial Fibrillation. JAMA 2001;285:2864-2870. Gladstone DJ, Bui E, Fang J, et al. Potentially preventable strokes in high-risk patients with atrial fibrillation who are not adequately anticoagulated. Stroke 2009;40:235-240. Boulanger L, Kim J, Friedman M, et al. Patterns of use of antithrombotic therapy and quality of anticoagulation among patients with non-valvular atrial fibrillation in clinical practice. Int J Clin Pract 2006;60:258-264. Nieuwlaat R, Capucci A, Lip GY, et al. Antithrombotic treatment in real-life atrial fibrillation patients: a report from the Euro Heart Survey on Atrial Fibrillation. Eur Heart J 2006;27:3018-3026. Goto S, Bhatt DL, Rother J, et al. Prevalence, clinical profile, and cardiovascular outcomes of atrial fibrillation patients with atherothrombosis. Am Heart J 2008; 156:855-863. 863.e2. Hopps S, Marcy TR. Warfarin versus aspirin: using CHADS2 to guide therapy for stroke prevention in nonvalvular atrial fibrillation. Consult Pharm 2009;24:841-844. Lee BH, Park JS, Park JH, et al. The effect and safety of the antithrombotic therapies in patients with atrial fibrillation and CHADS score 1. J Cardiovasc Electrophysiol 2010;21:501-507. Bungard TJ, Ghali WA, Teo KK, et al. Why do patients with atrial fibrillation not receive warfarin? Arch Intern Med 2000;160:41-46. Partington SL, Abid S, Teo K, et al. Pre-admission warfarin use in patients with acute ischemic stroke and atrial fibrillation: the appropriate use and barriers to oral anticoagulant therapy. Thromb Res 2007;120:663-669. Man-Son-Hing M, Laupacis A. Anticoagulant-related bleeding in older persons with atrial fibrillation: physicians’ fears often unfounded. Arch Intern Med 2003; 163:1580-1586. Mant J, Hobbs FD, Fletcher K, et al. Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised controlled trial. Lancet 2007;370:493-503. Becker RC, Meade TW, Berger PB, et al. The primary and secondary prevention of coronary artery disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008;133: 776S-814S.
Background: Warfarin reduces the risk of stroke in patients with atrial fibrillation (AF) but is often underused in clinical practice. We aimed to examine the current state of warfarin use in nonvalvular atrial fibrillation (NVAF) patients with firstever ischemic stroke (IS) or transient ischemic attack (TIA) in China and to analysis factors causing such underuse. Methods: By accessing the China National Stroke Registry, data on consecutive patients with known NVAF who developed firstever IS or TIA were studied. Proportion of warfarin use was estimated in those patients with known NVAF who should be eligible for anticoagulation therapy. Factors associated with warfarin underuse were identified by multivariate logistic regression. Results: Of 11,080 patients with first-ever IS or TIA, 996 (9.7%) had NVAF and no contraindications to anticoagulation therapy and 592 of them had history of AF. Among these patients, only 96 (16.2%) were on warfarin and 496 (83.8%) were not. In those patients on warfarin, only 1 of his international normalized ratios on admission was in the therapeutic range of 2.0-3.0. Based on the CHADS2 scores, about 20.2% low-risk AF patients took warfarin; however, only 15.2% moderate and 16.4% high-risk patients were on warfarin. Older patients and patients with history of coronary heart disease (CHD) were less likely to be given warfarin, whereas patients with history of prestroke antiplatelet use were more likely to take warfarin. Conclusions: We found that warfarin was significantly underused in patients with known NVAF in China. Age and CHD and prestroke antiplatelet therapy were related factors. Key Words: Anticoagulation—atrial fibrillation—warfarin—stroke management—transient ischemic attack. Ó 2014 by National Stroke Association
Introduction Atrial fibrillation (AF) is a common arrhythmia not only in western countries but also in China. It causes about
From the *Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China; and †INI Stroke Center and Stroke Network, OSF Healthcare System, University of Illinois College of Medicine, Peoria, Illinois. Received August 30, 2013; revision received October 7, 2013; accepted October 9, 2013. Grant support: This study was funded by the Ministry of Science and Technology and the Ministry of Health of the People’s Republic of China. The grant numbers are National S & T Major Project of China (2008ZX09312-008), State Key Development Program of Basic Research of China (2009CB521905), the Research Special Fund For
15% of ischemic stroke (IS) each year.1 Since 1990s, studies have shown that anticoagulation with warfarin can effectively prevent IS in nonvalvular atrial fibrillation (NVAF) patients.2,3 However, despite its clear benefit for stroke Public Welfare Industry of Health (200902004), and the National 12th Five-year S & T Major Project (2011BAI08B02). The authors report no conflicts of interest. Address correspondence to Yongjun Wang, MD, Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, No. 6 Tiantanxili, Dongcheng District, Beijing 100050, China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2013.10.006
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 5 (May-June), 2014: pp 1157-1163
1157
C. WANG ET AL.
1158
prevention, warfarin has been underprescribed in clinical practice.4,5 Zhou et al6 have found that warfarin use was only about 2.7% in AF patients, and Qi7 have revealed that the anticoagulation rate was only 6.6% in hospitalized patients with AF in China. Most of these studies were based on AF population in general and few of them focused on NVAF patients who went on and developed first-ever IS or TIA. Furthermore, it has not been well studied about the factors that may contribute to warfarin underuse in this AF population. To evaluate this important issue of warfarin underuse and further determine the factors associated with the formation of such gap in China, we analyzed the data from the China National Stroke Registry (CNSR).
Materials and Methods Design and Setting CNSR is a national hospital-based and prospective registry sponsored by the Ministry of Health. The primary objective of the registry is to evaluate the quality of care during acute hospitalization of all stroke patients. The registry collects data on many performance measurements on the delivery of stroke care around the nation. A total of 132 hospitals representing 27 provinces and 4 municipalities (except for Taiwan and Macao regions) in Mainland China have been selected as registry hospitals. Among them, 100 are tertiary hospitals and 32 are secondary urban hospitals. Patients were eligible to be enrolled in the registry if they met the following criteria: 18 years of age or more, IS, TIA, intracranial hemorrhage (ICH), or subarachnoid hemorrhage confirmed by brain CT or MRI within 14 days after the onset of symptoms. Trained research co-ordinators at each institute reviewed medical records daily to identify, consent, and enroll qualified patients. Standardized data collection form was used by all sites. Trained research co-ordinators obtained these data directly through interviewing the patients and family: prehospital care, prestroke modified Rankin Scale, and baseline National Institutes of Health Stroke Scale and Glasgow Coma Scale scores. Other data extracted from the medical records included patient demographics, medical history, medication use, IS subtypes, ICH and causes, vascular malformation, coagulopathy, disease status (ie, National Institutes of Health Stroke Scale, neuroimaging studies), complications in hospital, disease management in different hospital settings, and discharge status. Patients were contacted by telephone at 3, 6, 12, 18, and 24 months after discharge by trained research personnel at Beijing Tiantan Hospital. Throughout the study period, all data elements were manually checked by a research specialist from an independent research organization. A professional data processing company was responsible for the computer data entry. Beijing Translational Medicine Research Center, an independent research organiza-
tion, served as the data analysis. Other details in design and rationale of CNSR have been described elsewhere.8,9
Eligibility and Criteria For the current study, we identified all patients who presented with TIA or first-ever IS. AF was defined as previously diagnosed AF documented in the medical record. Patients with newly detected AF on admission and those diagnosed with ICH or SAH or those with a history of IS were excluded. Patients with history of valvular heart disease or contraindications to warfarin use were also excluded. Other contraindications that have excluded the patients included metastatic cancer, dementia, cirrhosis, renal failure requiring dialysis, previous gastrointestinal bleed, and peptic ulcer diseases. Warfarin use was determined by the recorded prescriptions provided by the patient or his/her caregiver. The target population was divided into 2 groups: NVAF not-on warfarin and NVAF on warfarin. The baseline characteristics of these 2 groups included age, gender, marital status, body mass index, educational level (elementary or less, namely #6 years education; middle school, 6-9 years; and high school or above, namely .9 years), health insurance type (urban or governmental, rural, commercial or other, and self-payment), and geographic regions (eastern, central, or western according to the geographic location of the hospitals from which the patients were recruited). Data on risk factors and comorbidities of stroke included currently smoking (continuous or accumulated smoking for at least 6 months and $1 cigarette per day and had smoked at least once within 30 days before the stroke), heavy drinking ($5 standard alcoholic drinks per day), hypertension, hyperlipidemia, congestive heart failure (with definitive diagnosis on record), coronary heart disease (CHD, including silent myocardial ischemia, angina pectoris, myocardial infarction, or ischemic cardiomyopathy), peripheral vascular diseases (defined as peripheral vascular or lymph diseases including thromboangiitis obliterans, arteriosclerotic obliterans, arterial embolism, polyarteritis, Raynaud’s syndrome, varicose veins, or deep venous thrombosis), and diabetes mellitus (both types 1 and 2). The risk of stroke was estimated by using the CHADS2 score,10 which was derived from congestive heart failure (1 point), hypertension (1 point), age of 75 years or more (1 point), diabetes mellitus (1 point), and previous stroke or TIA (2 points). According to this scoring system, patients were categorized into 3 groups: low risk (0), moderate risk (1-2), and high risk ($3). Data from eligible patients for current analysis were obtained from the CNSR database after all information has been entered.
Statistical Analysis For descriptive analysis, proportions were used for categorical variables, and means with standard deviations
UNDERUSE OF WARFARIN IN CHINA
1159
Figure 1. Flow diagram defining patients with known NVAF but no contraindications of warfarin use from CNSR. Abbreviations: IS, ischemic stroke; NVAF, nonvalvular atrial fibrillation; TIA, transient ischemic attack; WFR, warfarin.
were used for continuous variables. The baseline characteristic differences between NVAF not-on warfarin group and NVAF on warfarin group were analyzed using nonparametric Kruskal–Wallis test for continuous variables, and the c2 or Fisher exact test were used for categorical variables. Univariate logistic regression analysis was used to identify the demographic and clinical factors associated with the underuse of warfarin such as age, gender, CHADS2 score, education, geographic region, etc. Variables (P , .20) were entered into a multivariate regression model to identify the independent factors related to warfarin underuse with significance level of P less than .05. The odds ratios (ORs) and 95% confidence interval (CI) of both univariate and multivariate analysis were reported and a 95% CI that did not contain 1.0 was considered to be statistically significant. All analyses were conducted with a commercially available software package (SAS version 9.1.3; SAS Institute, Inc.).
Results When the data bank was closed, 22,216 patients have been entered into the CNSR. Among them, 11,080 had first-ever IS or TIA without history of valvular heart disease. Of these patients, 793 were excluded with contraindications of using oral anticoagulant and 996 (9.7%) had
NVAF. After excluding 404 patients with newly diagnosed AF, a total of 496 (83.8%) patients in the AF noton warfarin group and 96 (16.2%) in the AF on warfarin group were entered into the analysis (Fig 1).
Patient Characteristics The average age in the AF on warfarin group (63.7 6 12.7 years) was younger than that of the AF without warfarin group (72.1 6 10.5 years, P , .0001). For subjects with known AF who were eligible for anticoagulation therapy, the rates of warfarin use decreased with older ages. There were more female patients in the not-on warfarin group (P 5 .0264). CHD showed a rather lower proportion in the AF on warfarin group (30.2%) compared with AF not-on warfarin group (42.7%, P 5 .0221). Based on the CHADS2 score, 483 of 592 (81.6%) AF patients had at least 1 risk factor of stroke, but 84.7% of them (409 of 483) were not on warfarin. As for combining warfarin and antiplatelet together before stroke, there were more patients taking antiplatelet agents in AF on warfarin group (59.4%) than in AF not-on warfarin group (19.6%, P , .0001). Of those on warfarin, about 42.1% had high school or higher education, compared with only 22.3% in the not-on warfarin group (P , .0001). As for different health insurance type, no significant difference
C. WANG ET AL.
1160
Table 1. Baseline characteristics for patients with known NVAF and no contraindications of warfarin therapy categorized by warfarin use Patients with known NVAF no contraindications
Age (mean, y) 18-45 46-65 66-75 .75 Female, n (%) Marital status Single Married Divorced/widowed/remarried Lives Alone With others In nursing home Body mass index ,25 25-30 $30 Current smoker Heavy drinking Hypertension Hyperlipidemia Diabetes mellitus Coronary heart diseases Congestive heart failure Peripheral vascular diseases CHADS2 (0-6 points) Low risk: 0 Moderate risk: 1-2 High risk: $3 Drug administration prestroke Antiplatelet agents Antihypertensive agents Antihyperlipidemia agents Insulin or oral hypoglycemic agents Education Elementary or below Middle school High school or above Health insurance type Urban or governmental Rural Commercial or other Self-payment Geographic region Eastern Central Western
Overall (n 5 592)
No warfarin (n 5 496)
With warfarin (n 5 96)
P value
70.7 6 11.3 15 (2.5%) 159 (26.9%) 194 (32.8%) 224 (37.8%) 308 (52.0%)
72.1 6 10.5 8 (1.6%) 111 (22.4%) 173 (34.9%) 204 (41.1%) 268 (54.0%)
63.7 6 12.7 7 (7.3%) 48 (50.0%) 21 (21.9%) 20 (20.8%) 40 (41.7%)
,.0001 ,.0001
4 (.7%) 501 (85.1%) 84 (14.3%)
3 (.6%) 415 (84.2%) 75 (15.2%)
1 (1.0%) 86 (89.6%) 9 (9.4%)
.2985
24 (4.1%) 561 (95.7%) 1 (.2%)
21 (4.3%) 469 (95.5%) 1 (.2%)
3 (3.2%) 92 (96.8%) 0 (.0%)
.7979
304 (61.3%) 164 (33.1%) 28 (5.6%) 161 (27.3%) 37 (6.3%) 373 (63.0%) 63 (10.6%) 111 (18.8%) 241 (40.7%) 60 (10.1%) 7 (1.2%)
255 (62.3%) 129 (31.5%) 25 (6.1%) 125 (25.3%) 30 (6.0%) 308 (62.1%) 47 (9.5%) 89 (17.9%) 212 (42.7%) 53 (10.7%) 4 (.8%)
49 (56.3%) 35 (40.2%) 3 (3.4%) 36 (37.5%) 7 (7.3%) 65 (67.7%) 16 (16.7%) 22 (22.9%) 29 (30.2%) 7 (7.3%) 3 (3.1%)
.2270 . . .0141 .6451 .2972 .0365 .2532 .0221 .3132 .0544
109 (18.4%) 428 (72.3%) 55 (9.3%)
87 (17.5%) 363 (73.2%) 46 (9.3%)
22 (22.9%) 65 (67.7%) 9 (9.4%)
.4499
154 (26.0%) 285 (48.1%) 25 (4.2%) 87 (14.7%)
97 (19.6%) 247 (49.8%) 15 (3.0%) 68 (13.7%)
57 (59.4%) 38 (39.6%) 10 (10.4%) 19 (19.8%)
,.0001 .0667 .0010 .1234
304 (52.1%) 130 (22.3%) 149 (25.6%)
274 (56.1%) 105 (21.5%) 109 (22.3%)
30 (31.6%) 25 (26.3%) 40 (42.1%)
,.0001
374 (63.2%) 89 (15.0%) 18 (3.0%) 111 (18.8%)
309 (62.3%) 78 (15.7%) 16 (3.2%) 93 (18.8%)
65 (67.7%) 11 (11.5%) 2 (2.1%) 18 (18.8%)
.6387
413 (69.8%) 86 (14.5%) 93 (15.7%)
356 (71.8%) 66 (13.3%) 74 (14.9%)
57 (59.4%) 20 (20.8%) 19 (19.8%)
.0468
Abbreviation: NVAF, nonvalvular atrial fibrillation.
.0264
UNDERUSE OF WARFARIN IN CHINA
1161
Figure 2. Comparison of warfarin use in different AF populations. ‘‘*,’’ An American retrospective cohort study with total number of 13,709 recruited AF outpatients and 10.2% with a history of stroke. ‘‘†,’’ A subpopulation of Euro Heart Survey with total number of 2706 recruited patients and 336 (12.4%) had a prior stroke or TIA. ‘‘‡,’’ A study based on data of the REACH Registry with total number of AF patients 6814 and 43.67% had a prior stroke or TIA. ‘‘x,’’ A subpopulation from phase 3 of the RCSN. Here we used only the subgroup (n 5 597) data of first-ever ischemic stroke patients for comparison. Abbreviations: AP, antiplatelet; OAC, oral anticoagulation; RCSN, Registry of the Canadian Stroke Network; REACH, Reduction of Atherothrombosis for Continued Health. (Color version of figure is available online.)
was found between AF on warfarin and not-on warfarin group (P 5 .6387), but as for geographic region, a significant difference among eastern, central, and western was detected between both groups (P 5 .0468; Table 1).
Current State of Warfarin Use For patients with known NVAF and without any contraindication, only 16.2% received warfarin (9.6% on both warfarin and an antiplatelet agent), 16.4% were on antiplatelet alone, and 67.4% were on neither warfarin nor antiplatelet agents (Fig 2). The international normalized ratio (INR) was in the desired range of 2.0-3.0 in only 1 of 96 NVAF patients on warfarin while the INR in majority (55.2%) was below this therapeutic range (INR , 2.0) and the INRs of the other 40 patients (41.7%) were missing. About 20.2% of 109 patients with low CHADS2 scores were on warfarin while 15.2% with moderate and 16.4% with high CHADS2 scores were on warfarin (Fig 3).
Factors Related to Warfarin Use In patients with known NVAF, the ORs of warfarin use gradually reduced with older age. The lowest use was
Figure 3. Proportions of warfarin use with increased risk of stroke based on CHADS2 score in different AF populations. *, †, and ‡, please refer to the footnote of Figure 2. Note: CHADS2 score in the US study by Boulanger et al12 was classified as low risk (0-1), moderate risk (2-3), and high risk (.4), which was a small difference from the other 3 studies. Abbreviation: AF, atrial fibrillation. (Color version of figure is available online.)
among those older than 75 years (OR 5 .05; 95% CI, .01.19, age of 18-45 years as a reference). AF patients with CHD were much less likely to be on warfarin (OR 5 .47; 95% CI, .27-.83). However, patients with antiplatelet agents before stroke were more likely to receive warfarin (OR 5 6.99; 95% CI, 4.08-11.99). Gender, marital status, CHADS2 score, educational level, health insurance type, geographic region, and other clinical characteristics did not have significant impact on the use of warfarin (Fig 4).
Discussion From this study, we found several concerns in AFrelated stroke prevention in China. First of all, warfarin was significantly underused in patients with known NVAF and developed first-ever IS or TIA. Only about one sixth were on warfarin for primary prevention. Furthermore, many patients (55.2%) who were on warfarin did not reach the therapeutic range. Overall, approximately 90% of first-ever IS patients with known NVAF were not appropriately anticoagulated before their catastrophic event. This undertreatment with warfarin as seen in CNSR was similar to the findings in the Registry of Canadian Stroke Network.11 Comparing with other 3
C. WANG ET AL.
1162
Figure 4. Uni- and multivariate logistic regression analysis for factors related to warfarin use. Abbreviation: OR, odds ratio.
well-known AF research in the United States,12 Europe,13 and the Reduction of Atherothrombosis for Continued Health Registry,14 the rate of warfarin use in China was much lower (Fig 2). Significant numbers of strokes could be prevented in AF patients if these patients were appropriately anticoagulated. Second, AF patients with higher CHADS2 scores and increased stroke risks had less chance to be on warfarin in CNSR. As we know, warfarin can be given to patients with moderate risk and should be given to those with high risk. Some recent studies even suggest that patients with CHADS2 score of 1 should receive warfarin rather than aspirin.15,16 However, the proportions of warfarin use in CNSR were unexpectedly decreased with increased stroke risks. This finding was very different from the other studies not only in the percentage but also in the trend (Fig 3). There was a big gap between understanding the evidence-based guidelines and practicing accordingly among Chinese physicians. It shows that much work needs to be done to improve the compliance with the guidelines. Third, both age and CHD were independent risk factors of lower use of warfarin, whereas prestroke antiplatelet agent use was associated with higher warfarin treatment. Although patients, treating physicians, and regional health care system may have an impact on the use of anticoagulation,11,17 our discovery of the impact of patient’s older age associated with the unlikelihood of being on warfarin was not unique to China.18 Unfortunately, many physicians were reluctant to prescribe warfarin especially for older patients because they overestimated the risk of bleeding and underestimated the risk of stroke and the benefit of warfarin,19 even though age itself should not be regarded as a contraindication to anticoagulation.20 It seemed that Chinese AF patients with CHD were less likely to receive warfarin. This might be because of the use of antiplatelet agents for CHD prevention in these patients who fear to be on warfarin. Lastly, the INR was inappropriately monitored in AF patients on warfarin in China. More than two fifths (41.7%) of them had no INRs. One of the reasons was that INR was not checked on admission, an indication
that the treating physicians did not pay enough attention to INR monitoring. In China, checking INR is rather expensive and mainly done in the tertiary hospitals. This could be another reason why warfarin may be underprescribed. Although portable INR monitor is now available in China, it is too expensive to be used widely. We also found surprisingly that prestroke antiplatelet treatment was associated with better warfarin utilization. Combining antiplatelet agents with warfarin are recommended in those AF patients with acute myocardial infarction for at least 3 months and those with mechanical or bioprosthetic heart valves.21 An interesting phenomenon is that Chinese physicians often prescribe lower dose of warfarin and yet add an antiplatelet agent such as aspirin to the regimen. The common belief is that lower doses of warfarin may avoid hemorrhage and yet provide an anticoagulation effect. This unrealistic fear of bleeding is popular among Chinese physicians even though they fully understand what the target INR should be. There are several limitations of our study. Because only patients admitted to the neurology departments were captured in CNSR and no patients from cardiology departments were included, the prevalence of AF and rate of anticoagulation could be underestimated. This study may have also underestimated the magnitude of warfarin underuse because the data in CNSR were from tertiary or secondary hospitals, and rural hospitals were excluded. Furthermore, our study only examined the warfarin use in NVAF patients with the first-ever IS or TIA and not with other medical conditions.
Conclusions China faces significant challenges in stroke and TIA prevention in patients with NVAF. Many factors impede the use of warfarin in AF patients who have no obvious contraindications. Appropriate use of warfarin may significantly decrease first-ever IS or TIA in patients with AF. The main obstacle seems to be the belief held by the Chinese physicians that warfarin may cause excessive unforeseen hemorrhage. With the aging population1 and likelihood of increased prevalence of AF in China, it
UNDERUSE OF WARFARIN IN CHINA
has become urgent that effective strategies are needed to address these deficiencies. It is rather clear that improving stroke prevention is now a pressing issue not only for Chinese physicians but also for the Chinese government. Acknowledgment: We thank all participating hospitals, colleagues, nurses, and imaging and laboratory technicians.
1163
11.
12.
13.
References 1. Wolf P, Abbott R, Kannel W. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke 1991;22:983-988. 2. The Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators. The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. N Engl J Med 1990;323:1505-1511. 3. Ezekowitz MD, Bridgers SL, James KE, et al. Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. N Engl J Med 1992;327:1406-1412. 4. Ogilvie IM, Newton N, Welner SA, et al. Underuse of oral anticoagulants in atrial fibrillation: a systematic review. Am J Med 2010;123:638-645.e634. 5. Mercaldi CJ, Ciarametaro M, Hahn B, et al. Cost efficiency of anticoagulation with warfarin to prevent stroke in medicare beneficiaries with nonvalvular atrial fibrillation. Stroke 2011;42:112-118. 6. Zhou ZQ, Hu DY, Chen J, et al. An epidemiological survey of atrial fibrillation in China. Chin J Intern Med 2004;43:491-494. 7. Qi WH. Retrospective investigation of hospitalized patients with atrial fibrillation in major parts of China. Chin J Cardiol 2003;31:913-915. 8. Wang Y, Cui L, Ji X, et al. The China National Stroke Registry for patients with acute cerebrovascular events: design, rationale, and baseline patient characteristics. Int J Stroke 2011;6:355-361. 9. Li S, Zhao X, Wang C, et al. Risk factors for poor outcome and mortality at 3 months after the ischemic stroke in patients with atrial fibrillation. J Stroke Cerebrovasc Dis 2013;13:146-148. 10. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification schemes for predicting stroke: re-
14.
15.
16.
17.
18.
19.
20.
21.
sults from the National Registry of Atrial Fibrillation. JAMA 2001;285:2864-2870. Gladstone DJ, Bui E, Fang J, et al. Potentially preventable strokes in high-risk patients with atrial fibrillation who are not adequately anticoagulated. Stroke 2009;40:235-240. Boulanger L, Kim J, Friedman M, et al. Patterns of use of antithrombotic therapy and quality of anticoagulation among patients with non-valvular atrial fibrillation in clinical practice. Int J Clin Pract 2006;60:258-264. Nieuwlaat R, Capucci A, Lip GY, et al. Antithrombotic treatment in real-life atrial fibrillation patients: a report from the Euro Heart Survey on Atrial Fibrillation. Eur Heart J 2006;27:3018-3026. Goto S, Bhatt DL, Rother J, et al. Prevalence, clinical profile, and cardiovascular outcomes of atrial fibrillation patients with atherothrombosis. Am Heart J 2008; 156:855-863. 863.e2. Hopps S, Marcy TR. Warfarin versus aspirin: using CHADS2 to guide therapy for stroke prevention in nonvalvular atrial fibrillation. Consult Pharm 2009;24:841-844. Lee BH, Park JS, Park JH, et al. The effect and safety of the antithrombotic therapies in patients with atrial fibrillation and CHADS score 1. J Cardiovasc Electrophysiol 2010;21:501-507. Bungard TJ, Ghali WA, Teo KK, et al. Why do patients with atrial fibrillation not receive warfarin? Arch Intern Med 2000;160:41-46. Partington SL, Abid S, Teo K, et al. Pre-admission warfarin use in patients with acute ischemic stroke and atrial fibrillation: the appropriate use and barriers to oral anticoagulant therapy. Thromb Res 2007;120:663-669. Man-Son-Hing M, Laupacis A. Anticoagulant-related bleeding in older persons with atrial fibrillation: physicians’ fears often unfounded. Arch Intern Med 2003; 163:1580-1586. Mant J, Hobbs FD, Fletcher K, et al. Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised controlled trial. Lancet 2007;370:493-503. Becker RC, Meade TW, Berger PB, et al. The primary and secondary prevention of coronary artery disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008;133: 776S-814S.