The American Journal of Medicine
Volume 123, Issue 5 , Pages 446-453, May 2010

Are Atrial Fibrillation Patients Receiving Warfarin in Accordance with Stroke Risk?

Parts of these data were presented in poster format at the Heart Rhythm Society Annual Meeting, May 13-16, 2009.

  • Peter J. Zimetbaum, MD

      Affiliations

    • Harvard Medical School, Boston, Mass
    • Corresponding Author InformationRequests for reprints should be addressed to Peter J. Zimetbaum, MD, Director, Clinical Cardiology, Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, MA 02215
    • ,
  • Amit Thosani, MD

      Affiliations

    • Harvard Medical School, Boston, Mass
    • ,
  • Hsing-Ting Yu, MPH

      Affiliations

    • Cerner LifeSciences, Beverly Hills, Calif
    • ,
  • Yan Xiong, MS

      Affiliations

    • Cerner LifeSciences, Beverly Hills, Calif
    • ,
  • Jay Lin, PhD

      Affiliations

    • sanofi-aventis, Bridgewater, NJ
    • ,
  • Prajesh Kothawala, MD, MPH

      Affiliations

    • Cerner LifeSciences, Beverly Hills, Calif
    • ,
  • Matthew Emons, MD, MBA

      Affiliations

    • Cerner LifeSciences, Beverly Hills, Calif

Article Outline

Abstract 

Background

Clinical guidelines for the management of atrial fibrillation and atrial flutter provide recommendations for anticoagulation based on patients' overall risk of stroke. To determine the real-world compliance of physicians with these recommendations, we conducted a retrospective cohort study examining the utilization of warfarin in atrial fibrillation/flutter patients by stroke risk level.

Methods

Patients with a qualifying atrial fibrillation/flutter diagnosis during ≥18 months' continuous enrollment between January 2003 and September 2007, and with ≥6 months' eligibility after the first atrial fibrillation/flutter diagnosis, were identified from the US MarketScan database (Thomson Reuters, New York, NY). Warfarin use within 30 days of the first diagnosis was assessed according to stroke risk, estimated using the Congestive heart failure, Hypertension, Age >75 years, Diabetes, Stroke (CHADS2) score.

Results

Of 171,393 patients included in the analysis, 20.0% had a CHADS2 score of 0 (low risk), 61.6% a score of 1-2 (moderate risk), and 18.4% a score of 3-6 (high risk). Warfarin, recommended for high stroke-risk patients, was given to only 42.1% of those with a CHADS2 score of 3-6. A similar percentage of patients with moderate (43.5%) or low stroke risk (40.1%) received warfarin. Only 29.6% of high-risk, 33.3% of moderate-risk, and 34.1% of low-risk patients who were started on warfarin received uninterrupted therapy for 6 months following their initial prescription.

Conclusions

These data suggest that guideline recommendations that anticoagulation should be provided in accordance with stroke risk in atrial fibrillation patients are not routinely followed in clinical practice. The causes and clinical implications of under-utilization of anticoagulation in atrial fibrillation patients with high stroke risk warrant further study.

Keywords: Anticoagulation, Atrial fibrillation, Atrial flutter, Stroke risk, Warfarin

 

Atrial fibrillation/atrial flutter is widespread, affecting an estimated 2.4 million individuals in the US. Furthermore, the prevalence of atrial fibrillation is projected to more than double over the next 50 years.1 Patients with atrial fibrillation frequently have increased rates of cardiovascular and all-cause mortality compared with individuals without atrial fibrillation.2, 3, 4 As a result, health care utilization rates are high, with approximately 350,000 hospitalizations and 5 million office visits attributable to the management of atrial fibrillation in 2001.5 Treatment of atrial fibrillation represents a significant economic burden, with total direct costs estimated at $6.65 billion annually in the US.5

Clinical Significance

 


•Guidelines for the management of atrial fibrillation recommend that warfarin is provided according to patients' stroke risk.

•Less than half of patients with atrial fibrillation/atrial flutter in a large US medical claims database received warfarin following a diagnosis of atrial fibrillation/atrial flutter.

•Similar proportions with low, moderate, and high stroke risk received warfarin, suggesting that warfarin is underused among patients with high stroke risk, while overused in those with low stroke risk.

The presence of atrial fibrillation or atrial flutter is strongly associated with an increased risk of stroke.3, 4, 6 Atrial fibrillation increases both the risk of stroke by an estimated 5-fold and stroke severity, compared with stroke due to other etiologies.7 This increased stroke risk may be attributed to the hypercoagulable state associated with atrial fibrillation, direct increases in vascular endothelial growth factor levels, and the presence of risk factors such as hypertension and diabetes among atrial fibrillation patients.8, 9, 10

Improving the prevention and management of stroke in atrial fibrillation/atrial flutter patients is critical to reduce hospitalization and mortality. Although a recent analysis has suggested that the antiarrhythmic dronedarone may be associated with a reduced incidence of stroke in atrial fibrillation patients,11 to date most clinical studies of rhythm- and rate-control strategies for atrial fibrillation have demonstrated limited efficacy in decreasing the risk of stroke.12, 13, 14 Oral anticoagulation is currently considered the most effective treatment option for prevention of cerebrovascular events in atrial fibrillation patients.15, 16, 17 However, due to the attendant bleeding risk, it is critical that anticoagulant therapy is prescribed with consideration to the risk-benefit ratio for each patient.15, 17 Patients with atrial fibrillation and a low stroke risk (≤2% chance of stroke/year) do not benefit sufficiently from oral anticoagulation to compensate for the associated risks, whereas anticoagulation is favored in those with a high stroke risk (≥6% chance of stroke/year).15, 16

Various risk stratification schemes have been developed to determine patients' overall stroke risk.15, 18, 19, 20 One such scheme is the Congestive heart failure, Hypertension, Age >75 years, Diabetes, prior Stroke (CHADS2) score system, which estimates the cumulative stroke risk conferred by the presence of these risk factors.19, 20 Guidelines for atrial fibrillation management have incorporated the concept of evaluating a patient's stroke risk when determining antithrombotic requirements.15, 17 Accordingly, oral anticoagulant therapy with a vitamin K antagonist such as warfarin is recommended for atrial fibrillation patients with high stroke risk, while for those with low stroke risk, aspirin is recommended.15, 17 For patients with moderate stroke risk, warfarin or aspirin are recommended, depending on bleeding risks and patient preference.15, 17

Numerous studies have assessed antithrombotic usage among atrial fibrillation patients in clinical practice.21, 22, 23, 24, 25, 26 However, limited data are available on the real-world compliance of physicians in the US with guideline recommendations for prescribing anticoagulation in accordance with patients' stroke risk. We therefore examined the use of warfarin stratified by stroke risk among atrial fibrillation/atrial flutter patients in a large US medical claims database.

Back to Article Outline

Methods 

Study Design 

This retrospective cohort study included adult patients in the US MarketScan Commercial Claims and Encounters Database and Medicare Supplemental Database from Thomson Reuters (New York, NY). These databases are derived from employer- and government-funded (Medicare) health care insurance plans and represent 21.6 million individuals covered under a variety of fee-for-service and capitated provider reimbursement schemes. Data from the MarketScan database have been used previously for a range of research studies.27, 28, 29 The database complied with all aspects of the Health Information Portability and Accountability Act of 1996.

Study Population 

Patients were aged over 18 years with ≥18 months' continuous enrollment between January 1, 2003 and September 30, 2007. Patients with atrial fibrillation or atrial flutter were identified by the presence of one primary inpatient diagnosis of atrial fibrillation/atrial flutter (International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis code 427.3x), or one secondary inpatient and one outpatient atrial fibrillation/atrial flutter diagnosis, or ≥2 outpatient diagnoses of atrial fibrillation/atrial flutter on different dates during at least 18 months' continuous enrollment. The date of an atrial fibrillation/atrial flutter diagnosis preceded by at least 12 months' continuous enrollment was the index date. The first qualifying diagnosis of atrial fibrillation/atrial flutter during at least 18 months of continuous enrollment was the “first recorded atrial fibrillation/atrial flutter diagnosis.” Patients had to have at least 6 months' eligibility following this diagnosis.

Patients were classified as having newly diagnosed atrial fibrillation/atrial flutter if there was no evidence of atrial fibrillation/atrial flutter before the index date. Evidence of atrial fibrillation/atrial flutter included: a diagnosis of atrial fibrillation/atrial flutter; hospitalization, urgent care, or an office visit with a diagnosis of palpitations/tachycardia; new prescription for a rate control drug; an echocardiogram or Holter monitoring; use of warfarin with no other indication for anticoagulation (including deep venous thrombosis or pulmonary embolism); and ≥30 days' supply of an antiarrhythmic agent and no other indication for arrhythmias. Newly diagnosed atrial fibrillation/atrial flutter patients included those with a secondary atrial fibrillation/atrial flutter diagnosis that may resolve spontaneously without treatment (eg, postoperative cardiac patients). Patients were classified as having pre-existing atrial fibrillation/atrial flutter if the first recorded diagnosis occurred >30 days before the index date. Patients with indeterminate atrial fibrillation/atrial flutter (first recorded atrial fibrillation/atrial flutter diagnosis <30 days before the index date) were excluded.

Data Collection and Variables 

The use of warfarin within 30 days following the first atrial fibrillation/atrial flutter diagnosis was assessed in terms of the proportion of atrial fibrillation/atrial flutter patients receiving warfarin and the proportion of treated patients with uninterrupted therapy (<30-day gap in prescription coverage) over the 180 days following the first prescription, stratified by stroke risk. Stroke risk was estimated using the CHADS2 score.19, 20

The severity of warfarin-related adverse events was assessed in terms of the frequency of inpatient admissions, outpatient visits, and emergency department (ED) visits attributed to warfarin toxicity. Qualifying encounters included those with a primary diagnosis of gastrointestinal bleeding or intracranial hemorrhage that occurred up to 60 days after the most recent warfarin supply ended.

Back to Article Outline

Statistical Analysis 

All analyses were conducted using SAS version 9 (SAS Institute Inc., Cary, NC). Continuous variables were summarized as mean±SD and median, while binary data were summarized with percentages. The Charlson Comorbidity Index was calculated according to the method of Romano et al.30 The CHADS2 score was calculated as the sum of points allocated to each of the following stroke risk factors: congestive heart failure exacerbation (1 point), hypertension (1 point), age >75 years (1 point), diabetes mellitus (1 point), and prior stroke or transient ischemic attack (2 points).19 Patients were stratified according to their individual CHADS2 score (ranging from 0 to 6) and by CHADS2 category (a CHADS2 score of 0 was considered low stroke risk; 1 to 2 moderate risk, and 3 to 6 high risk). Adverse events associated with warfarin toxicity were expressed as the number of events per person-month of drug exposure ([number of events]/[total days supply of medication/30 days]).

Back to Article Outline

Results 

Characteristics of Overall Study Population 

Of 573,519 patients in the MarketScan database with a diagnosis of atrial fibrillation/atrial flutter during the study period, 171,393 patients were included in the analysis: 51,907 (30.3%) with newly diagnosed atrial fibrillation/atrial flutter and 119,486 (69.7%) with pre-existing atrial fibrillation/atrial flutter (Figure 1). The mean (SD) duration of continuous eligibility across all patient groups was 41.0 (13.7) months. Overall, patients were of mean age 73.5 years, and 54.8% were male (Table 1).

Table 1. Patient Demographics
All Patients (n=171,393)CHADS2 Score⁎
Low RiskModerate RiskHigh Risk
0(n=34,338)1(n=58,004)2(n=47,559)3(n=20,589)4(n=7711)5(n=2625)6(n=567)
Male, n(%)93,993(54.8)22,620(65.9)32,562(56.1)23,614(49.7)10,150(49.3)3604(46.7)1184(45.1)259(45.7)
Age, years, mean(SD)73.5(12.5)61.1(11.2)73.3(11.7)78.4(9.6)79.8(8.6)80.9(7.2)81.4(7.6)82.3(4.5)
Age category, years, n(%)
≤5516,052(9.4)9558(27.8)4681(8.1)1390(2.9)345(1.7)50(0.6)28(1.1)—
56-6525,258(14.7)10,300(30.0)9443(16.3)3960(8.3)1205(5.9)269(3.5)81(3.1)—
66-7540,986(23.9)14,480(42.2)16,005(27.6)6923(14.6)2650(12.9)693(9.0)235(9.0)—
76-8563,987(37.3)—20,282(35.0)25,402(53.4)11,501(55.9)4833(62.7)1531(58.3)438(77.2)
>8525,110(14.7)—7593(13.1)9884(20.8)4888(23.7)1866(24.2)750(28.6)129(22.8)
Region, n(%)
Northeast15,914(9.3)3288(9.6)5573(9.6)4313(9.1)1786(8.7)669(8.7)239(9.1)46(8.1)
North Central49,143(28.7)9726(28.3)16,413(28.3)13,858(29.1)5927(28.8)2223(28.8)824(31.4)172(30.3)
South51,810(30.2)11,650(33.9)17,780(30.7)13,627(28.7)5705(27.7)2185(28.3)701(26.7)162(28.6)
West54,212(31.6)9573(27.9)18,139(31.3)15,696(33.0)7134(34.7)2626(34.1)859(32.7)185(32.6)
Unknown314(0.2)101(0.3)99(0.2)65(0.1)37(0.2)8(0.1)2(0.1)2(0.4)

CHADS=Congestive heart failure, Hypertension, Age, Diabetes, Stroke.

⁎Calculated as the sum of points allocated to each stroke risk factor: congestive heart failure exacerbation=1 point, hypertension=1 point, age >75 years=1 point, diabetes mellitus=1 point, prior stroke or transient ischemic attack (or both)=2 points.19

Overall, 70.3% of patients received atrial fibrillation-related drug treatments within 30 days of their first recorded atrial fibrillation/atrial flutter diagnosis, with a median of 2 treatments (by drug class) per patient. Over half of patients (55.2%) received rate-control agents and 18.3% received rhythm-control agents.

Characteristics of Study Population Stratified by Stroke Risk 

In total, 34,388 (20.0%) patients were classified as having low stroke risk (CHADS2 score 0), 105,563 (61.6%) moderate stroke risk (CHADS2 score 1-2), and 31,492 (18.4%) high stroke risk (CHADS2 score 3-6) (Figure 2). The most common risk factor contributing to the CHADS2 score across all groups was age >75 years (52.0% of patients), followed by hypertension (47.2% of patients) (Table 2). As expected, mean age, proportion of male patients, and proportion of patients with comorbidities (particularly coronary artery disease) increased with CHADS2 score (Table 1, Table 3).

Table 2. Patients Meeting the CHADS2 Criteria
All Patients (n=171,393)CHADS2 Score⁎
Low RiskModerate RiskHigh Risk
0(n=34,338)1(n=58,004)2(n=47,559)3(n=20,589)4(n=7711)5(n=2625)6(n=567)
CHF exacerbation, n(%)33,773(19.7)—3262(5.6)11,590(24.4)12,472(60.6)3927(50.9)1955(74.5)567(100)
Hypertension, n(%)80,921(47.2)—21,318(36.8)33,854(71.2)16,287(79.1)6505(84.4)2390(91.1)567(100)
Age >75 years, n(%)89,097(52.0)—27,875(48.1)35,286(74.2)16,389(79.6)6699(86.9)2281(86.9)567(100)
Diabetes mellitus, n(%)32,763(19.1)—5549(9.6)12,384(26.0)9327(45.3)3687(47.8)1249(47.6)567(100)
Stroke or TIA (or both),n(%)12,853(7.5)——1002(2.1)3646(17.7)5013(65.0)2625(100)567(100)

CHADS=Congestive heart failure, Hypertension, Age, Diabetes, Stroke; CHF=congestive heart failure; TIA=transient ischemic attack.

⁎Calculated as the sum of points allocated to each stroke risk factor: congestive heart failure exacerbation=1 point, hypertension=1 point, age >75 years=1 point, diabetes mellitus=1 point, prior stroke or transient ischemic attack (or both)=2 points.19

Table 3. Patient Comorbidities
Comorbidity, n(%)All Patients (n=171,393)CHADS2 Score⁎
Low RiskModerate RiskHigh Risk
0(n=34,338)1(n=58,004)2(n=47,559)3(n=20,589)4(n=7711)5(n=2625)6(n=567)
CAD19,850(11.6)2509(7.3)5815(10.0)5994(12.6)3456(16.8)1392(18.1)541(20.6)143(25.2)
AMI7856(4.6)839(2.4)2056(3.5)2365(5.0)1571(7.6)671(8.7)263(10.0)91(16.1)
Angina12,318(7.2)1619(4.7)3806(6.6)3705(7.8)2034(9.9)788(10.2)305(11.6)61(10.8)
Cardiac dysrhythmias/conduction disorders51,504(30.1)8231(24.0)15,526(26.8)15,195(32.0)7982(38.8)3107(40.3)1188(45.3)275(48.5)
PAD11,384(6.6)864(2.5)2922(5.0)3806(8.0)2310(11.2)962(12.5)401(15.3)119(21.0)
Hypothyroidism17,377(10.1)2748(8.0)5463(9.4)5153(10.8)2522(12.3)1029(13.3)381(14.5)81(14.3)
Emphysema/COPD27,345(16.0)2943(8.6)7905(13.6)8596(18.1)5016(24.4)1982(25.7)727(27.7)176(31.0)

AMI=acute myocardial infarction; CAD=coronary artery disease; COPD=chronic obstructive pulmonary disease; PAD=peripheral artery disease.

⁎Calculated as the sum of points allocated to each stroke risk factor: congestive heart failure exacerbation=1 point, hypertension=1 point, age >75 years=1 point, diabetes mellitus=1 point, prior stroke or transient ischemic attack (or both)=2 points.19

Trends in Warfarin Utilization by Stroke Risk 

Warfarin was prescribed to 42.6% of the overall study population, with coverage being slightly higher in patients with newly diagnosed atrial fibrillation/atrial flutter (49.6%) than in those with pre-existing atrial fibrillation/atrial flutter (39.5%).

Fewer than half of atrial fibrillation/atrial flutter patients in each CHADS2 score group were prescribed warfarin: 42.1% of patients at high risk of stroke (CHADS2 score 3-6), 43.5% of those at moderate risk, and 40.1% of those at low risk (Figure 3). No differences were observed in warfarin utilization across individual CHADS2 score categories (Figure 3).

  • View full-size image.
  • Figure 3. 

    Warfarin use within 30 days of first recorded atrial fibrillation/flutter diagnosis, by stroke risk level. *Hatched area represents the proportion of patients with uninterrupted therapy over 180 days following their initial warfarin prescription. CHADS2=Congestive heart failure, Hypertension, Age >75 years, Diabetes, Stroke score.

Low utilization of warfarin was observed consistently upon stratification according to newly diagnosed and pre-existing atrial fibrillation/atrial flutter. Across the high stroke-risk group, a marginally higher proportion of newly diagnosed atrial fibrillation/atrial flutter patients (44.7%) than pre-existing atrial fibrillation/atrial flutter patients (41.4%) received warfarin, although for the group with a CHADS2 score of 5, this trend was reversed. For patients with low or moderate stroke risk, more newly diagnosed atrial fibrillation/atrial flutter patients (50.2% across both risk groups) than pre-existing atrial fibrillation/atrial flutter patients (39.0%) received warfarin (Figure 3).

Among patients initiating warfarin, only 32.8% received uninterrupted therapy over the first 6 months following their initial prescription (13.0% of the total study population). Overall, a higher proportion of patients with newly diagnosed atrial fibrillation/atrial flutter had uninterrupted treatment over 6 months (46.0% of those initiating therapy, 18.3% of the total population) than patients with pre-existing atrial fibrillation/atrial flutter (27.0% of those initiating therapy, 10.7% of the total population). Among newly diagnosed atrial fibrillation/atrial flutter patients initiating warfarin, a marginally higher proportion of those with moderate (46.3%) or high stroke risk (46.6%) received uninterrupted therapy for 6 months compared with those with low risk (44.7%). For patients with pre-existing atrial fibrillation/atrial flutter who initiated treatment, those in the high stroke-risk group were slightly less likely to have 6 months' uninterrupted treatment (25.8%) than the moderate (27.5%) and low stroke-risk (26.8%) groups (Figure 3).

Adverse Events Associated with Warfarin 

For patients with newly diagnosed atrial fibrillation/atrial flutter, per patient-month of warfarin exposure there were (mean) 0.11 inpatient admissions (ie, 1 admission per 9 patient-months of exposure), 0.21 outpatient visits, and 0.11 ED visits for gastrointestinal bleeding or intracranial hemorrhage presumed to be related to warfarin toxicity. For patients with pre-existing atrial fibrillation/atrial flutter, per patient-month of exposure there were (mean) 0.09 inpatient admissions, 0.19 outpatient visits, and 0.09 ED visits for adverse events consistent with warfarin toxicity.

Back to Article Outline

Discussion 

This study demonstrated that less than half of patients with newly diagnosed or pre-existing atrial fibrillation/atrial flutter in a large US medical claims database received warfarin therapy following a diagnosis of atrial fibrillation/atrial flutter, and only a third of patients starting treatment had 6 months of uninterrupted therapy. Furthermore, the low utilization of warfarin was consistently observed across patients with high, moderate, or low stroke risk.

Despite guideline recommendations that anticoagulation should be provided in accordance with patients' stroke risk,15, 17 similar proportions of patients in the low, moderate, and high stroke-risk groups received warfarin. These data suggest that anticoagulant therapy remains underused among atrial fibrillation patients with an elevated risk of stroke. Thus, patients who would benefit the most from anticoagulation due to a high risk of stroke are not receiving treatment, while those who have the least to benefit may perhaps be unnecessarily exposed to the inconvenience and risks associated with anticoagulant therapy. Furthermore, among warfarin-treated patients with high stroke risk, less than half had uninterrupted treatment for 6 months following therapy initiation. This indicates that even among high-risk patients in whom anticoagulant therapy is initiated, treatment is not routinely continued over the long term.

A number of studies have demonstrated similarly low rates of anticoagulation among atrial fibrillation patients to those observed in this analysis, with the lowest treatment rates occurring among elderly patients.21, 22, 24, 25 The reasons for the low anticoagulation rates in patients with moderate to high stroke risk are not captured in this claims database; however, it is likely that they reflect in part physicians' concerns about the risks associated with long-term anticoagulation. Bleeding complications are particularly common among elderly patients,31 who represented 82.4% of the high stroke-risk group. Other risk factors for bleeding complications, such as uncontrolled hypertension and pre-existing ischemic heart disease, also are commonly found in atrial fibrillation patients, especially those with higher CHADS2 scores,32, 33 and this may compound physicians' concerns over prescribing anticoagulation in this setting.

For certain patients, anticoagulation may have been initiated following electrical cardioversion, in line with treatment guidelines.15 For patients with low stroke risk, anticoagulation would be recommended only for short-term use (∼4 weeks) following cardioversion. However, the proportion of patients experiencing interruption of warfarin therapy was similarly high among those with high stroke risk, again indicating that stroke risk receives insufficient consideration when determining anticoagulant treatment strategies. By way of explanation for the uniform use of anticoagulation across stroke-risk categories and, in particular, the high prescription rates among low-risk patients, it is possible that physicians prescribed anticoagulation based predominantly on the severity of atrial fibrillation, or the presence of atrial fibrillation symptoms, rather than considering stroke risk in terms of a patient's overall clinical profile. In fact, stroke risk has been shown to be similar for patients with both paroxysmal and sustained atrial fibrillation/atrial flutter.34

Notably, patients with newly diagnosed atrial fibrillation/atrial flutter were more likely to receive anticoagulation than patients with pre-existing atrial fibrillation/atrial flutter across all CHADS2 categories. This would suggest that physicians are reluctant to continue anticoagulation over the long term and to prescribe anticoagulants to high-risk patients with multiple risk factors for bleeding. In addition, suboptimal anticoagulant treatment rates may stem from patient noncompliance with inconvenient dosing regimens, dietary restrictions, requirements for regular monitoring of their international normalized ratio, or an under-appreciation of their risk of stroke.35

In contrast to our results, an analysis of the Euro Heart Survey observed an increasing proportion of atrial fibrillation patients prescribed oral anticoagulants with increasing stroke risk, as assessed using a variety of scoring systems, including CHADS2. However, the variation in drug use across the risk categories was marginal, and a high proportion of patients with low stroke risk also received treatment (40%-50%).23 A prospective survey among atrial fibrillation patients in Geneva also showed an increase in anticoagulation with CHADS2 score, with as many as 88% (403/458) of patients with an indication for oral anticoagulation receiving therapy overall.26 While this may reflect a difference in the practice patterns in these European populations compared with the US managed care organizations in our study, it also is likely that in the European studies, physicians' knowledge that they were participating in a survey prompted them to take more than usual care in following guideline recommendations.

The retrospective, observational nature of this analysis could be a potential limitation. For the purposes of our analysis, we estimated the CHADS2 score retrospectively based on reported comorbidities; although some of the physicians may have estimated patients' CHADS2 score before prescribing anticoagulation, this information was not available in the database. Another limitation could arise from the inherent definition of the CHADS2 risk stratification method; atrial fibrillation patients with previous stroke/transient ischemic attack/thromboembolism are known to be at high risk of recurrent stroke, yet in the absence of other risk factors, these patients would have a CHADS2 score of only 2 and would be classified as moderate risk. However, only 2.2% of patients with a CHADS2 score of 2 had prior stroke/transient ischemic attack, and thus, this is unlikely to have contributed significantly to the results observed.

Back to Article Outline

Conclusion 

This study indicates that physicians currently fail to take sufficient account of patients' overall risk of stroke when considering anticoagulant therapy, and that anticoagulation is generally underused in atrial fibrillation patients. There is a need for continued research in this area to help physicians to balance the risks and benefits of anticoagulation in atrial fibrillation patients, especially in relation to stroke risk. However, until further evidence is available, physicians should ensure that they follow guideline recommendations and carefully evaluate overall stroke risk when considering anticoagulant therapy in patients with atrial fibrillation.

Back to Article Outline

References 

  1. Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study. JAMA. 2001;285:2370–2375
  2. Benjamin EJ, Wolf PA, D'Agostino RB, et al. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation. 1998;98:946–952
  3. Wolf PA, Mitchell JB, Baker CS, et al. Impact of atrial fibrillation on mortality, stroke, and medical costs. Arch Intern Med. 1998;158:229–234
  4. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991;22:983–988
  5. Coyne KS, Paramore C, Grandy S, et al. Assessing the direct costs of treating nonvalvular atrial fibrillation in the United States. Value Health. 2006;9:348–356
  6. Biblo LA, Yuan Z, Quan KJ, et al. Risk of stroke in patients with atrial flutter. Am J Cardiol. 2001;87:346–349A349
  7. Dulli DA, Stanko H, Levine RL. Atrial fibrillation is associated with severe acute ischemic stroke. Neuroepidemiology. 2003;22:118–123
  8. Lip GY. Does atrial fibrillation confer a hypercoagulable state?. Lancet. 1995;346:1313–1314
  9. Chung NA, Belgore F, Li-Saw-Hee FL, et al. Is the hypercoagulable state in atrial fibrillation mediated by vascular endothelial growth factor?. Stroke. 2002;33:2187–2191
  10. 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;154:1449–1457
  11. Connolly SJ. ATHENA: The effect of dronedarone on cardiovascular outcomes and stroke in patients with atrial fibrillation. Program and abstracts from the European Society of Cardiology Congress 2008. Germany: Munich; 2008;Clinical trials update 3
  12. Van Gelder IC, Hagens VE, Bosker HA, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med. 2002;347:1834–1840
  13. Wyse DG, Waldo AL, DiMarco JP, et al. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med. 2002;347:1825–1833
  14. Carlsson J, Miketic S, Windeler J, et al. Randomized trial of rate-control versus rhythm-control in persistent atrial fibrillation: the Strategies of Treatment of Atrial Fibrillation (STAF) study. J Am Coll Cardiol. 2003;41:1690–1696
  15. Fuster V, Ryden LE, Cannom DS, et al. ACC/AHA/ESC 2006 Guidelines for the Management of Patients with Atrial Fibrillation: 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): developed in collaboration with the European Heart Rhythm Association and the Heart Rhythm Society. Circulation. 2006;114:e257–e354
  16. Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med. 1999;131:492–501
  17. Singer DE, Albers GW, Dalen JE, et al. Antithrombotic therapy in atrial fibrillation: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest. 2008;133(6 Suppl):546S–592S
  18. van Walraven C, Hart RG, Wells GA, et al. A clinical prediction rule to identify patients with atrial fibrillation and a low risk for stroke while taking aspirin. Arch Intern Med. 2003;163:936–943
  19. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA. 2001;285:2864–2870
  20. Hughes M, Lip GY. Stroke and thromboembolism in atrial fibrillation: a systematic review of stroke risk factors, risk stratification schema and cost effectiveness data. Thromb Haemost. 2008;99:295–304
  21. Lewis WR, Fonarow GC, LaBresh KA, et al. Differential use of warfarin for secondary stroke prevention in patients with various types of atrial fibrillation. Am J Cardiol. 2009;103:227–231
  22. Fang MC, Stafford RS, Ruskin JN, Singer DE. National trends in antiarrhythmic and antithrombotic medication use in atrial fibrillation. Arch Intern Med. 2004;164:55–60
  23. 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
  24. 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
  25. Waldo AL, Becker RC, Tapson VF, Colgan KJ. Hospitalized patients with atrial fibrillation and a high risk of stroke are not being provided with adequate anticoagulation. J Am Coll Cardiol. 2005;46:1729–1736
  26. Meiltz A, Zimmermann M, Urban P, Bloch A. Atrial fibrillation management by practice cardiologists: a prospective survey on the adherence to guidelines in the real world. Europace. 2008;10:674–680
  27. Goldman HH, Frank RG, Burnam MA, et al. Behavioral health insurance parity for federal employees. N Engl J Med. 2006;354:1378–1386
  28. Chernew M, Gibson TB, Yu-Isenberg K, et al. Effects of increased patient cost sharing on socioeconomic disparities in health care. J Gen Intern Med. 2008;23:1131–1136
  29. Mark TL, Axelsen KJ, Mucha L, Sadkova Y. Racial differences in switching, augmentation, and titration of lipid-lowering agents by Medicare/Medicaid dual-eligible patients. Am J Manag Care. 2007;13(Suppl 3):S72–S79
  30. Romano PS, Roos LL, Jollis JG. Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol. 1993;46:1075–1079discussion 1081-1090
  31. Hylek EM, Evans-Molina C, Shea C, et al. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation. 2007;115:2689–2696
  32. DiMarco JP, Flaker G, Waldo AL, et al. Factors affecting bleeding risk during anticoagulant therapy in patients with atrial fibrillation: observations from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Am Heart J. 2005;149:650–656
  33. Hughes M, Lip GY. Risk factors for anticoagulation-related bleeding complications in patients with atrial fibrillation: a systematic review. QJM. 2007;100:599–607
  34. Hohnloser SH, Pajitnev D, Pogue J, et al. 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;50:2156–2161
  35. Crystal E, Connolly SJ. Role of oral anticoagulation in management of atrial fibrillation. Heart. 2004;90:813–817

 Funding: Financial support for the preparation of this manuscript was provided by sanofi-aventis US, Inc. This funding was used for provision of editorial support by Elizabeth Harvey, PhD and Andrew Fitton, PhD of UBC Scientific Solutions.

 Conflict of Interest: Dr. Peter Zimetbaum is a research consultant to sanofi-aventis. Dr. Jay Lin is an employee of sanofi-aventis. Hsing-Ting Yu and Matthew Emons are employees of Cerner LifeSciences, which provided consulting services to sanofi-aventis. Yan Xiong and Prajesh Kothawala were employed by Cerner LifeSciences at the time the research was conducted.

 Authorship: All authors had access to the study data, and participated in analysis or interpretation of the data (or both), and preparation of the manuscript.

PII: S0002-9343(09)01114-0

doi:10.1016/j.amjmed.2009.11.015

The American Journal of Medicine
Volume 123, Issue 5 , Pages 446-453, May 2010

Article Tools

* Image Usage & Resolution