AF AND STROKE / Darío C. Sobral Filho, et al 55REVIEW ARTICLE

Atrial Fibrillation and Stroke Risk Prevention in the Elderly

DARIO C. SOBRAL FILHO, MD, PHD, FACC - ANA CRISTIANNE R. LARANJEIRA, MD

Received: 02/12/2009Accepted: 20/01/2010

PROCAPE Hospital – University of Pernambuco - Brazil

INTRODUCTION

Atrial fibrillation (AF) is the most common sustainedarrhythmia observed in medical practice. (1) Itsprevalence increases with age; starting at the age of50, this doubles every decade of life, corresponding toa 5% prevalence in the population older than 60, anda 13% prevalence in the population older than 80. Thismeans that 70% of cases affect people who are between65 and 85 years old. (2) This is due to the increase inthe prevalence of predisposing factors that can triggerthe onset of arrhythmia that occur with age, forexample, changes in tissue and heart structure,systemic diseases, and the use of medications thatinduce arrhythmia, as well as other factors.

When an episode of AF lasts longer than 48 hours,there is an increased risk of intra-atrial thrombiformation. Atrial thrombi can migrate through bloodvessels and cause a thromboembolic phenomenon(TEP) (3), which constitutes the most common andserious type of complication associated with AF. Thesethrombotic events can either affect the central nervoussystem (e.g., causing an ischemic stroke) or they canbe peripheral. Therefore, treatment for the primaryand secondary prevention of such complications basedon risk stratification (2, 4-5) is important and shouldbe prescribed, except in the presence of contraindications,such as organic, psychosocial, and geographical factors.(1, 6-15)

The drugs used for primary and secondaryprevention are oral anticoagulants (OAC) andantiplatelet agents, the latter being less effective thanthe former. However, in daily practice, OACs are oftenunderused, especially among the elderly, due to thefear of possible hemorrhagic complications rather thanthe presence of true contraindications. (3)

Risk factors and thromboembolic phenomenaIn patients with AF, the risk factors for thromboembolicphenomena are considered either cardiac or independent.Among the former are valvulopathy (e.g., prostheticvalves, mitral stenosis), congestive heart failure(CHF), coronary artery disease, a left atrial diametergreater than 5 mm, and hypertrophic cardiomyopathy.The independent risk factors include a previoushistory of cerebral ischemia, systemic arterial

hypertension (SAH), diabetes mellitus (DM),peripheral arterial disease, thyrotoxicosis, femalegender, and advanced age. (2)

These risk factors have different degrees ofinfluence in terms of an individual’s predispositionto TEPs, that is, the presence of one risk factor mayimply a higher risk than another. These risk factorscan be stratified as follows: 1) low risk: female gender,age between 65 and 75 years, coronary artery disease(CAD), and thyrotoxicosis; 2) moderate risk: 75 yearsof age or older, SAH, CHF, a ventricular ejectionfraction (VEF) ≤ 35%, and DM; and 3) high risk: aprevious ischemic stroke or transient ischemic attack(TIA), mitral stenosis, and the presence of a prostheticheart valve. (2)

Age is an independent risk factor that is associatedwith a moderate increase in the risk of the occurrenceof TEPs. In other words, even in the absence of otherrisk factors, an elderly person with AF and no otherrisk factors for TEP may suffer such a complication.TEPs can cause serious sequelae, because, generally,2/3 of them affect the central nervous system andevolve, in 70% of cases, into serious and irreversiblecomplications or death.

Beyond the fact that age is an independent riskfactor and that these patients may have serioussequelae from TEPs, another important point is thatthe incidence of ischemic stroke, like that of AF,increases progressively with age. Until the age of 60,the risk of stroke per year is 1.5%. It increases to 5%per year for individuals between 60 and 75 years old,is 8.3% per year for people who are 75 years old, andcan reach as high as 23% per year for octogenarians.(2) Moreover, the risk of cerebral hemorrhage due tothe use of OACs is greater in patients who are over 80years of age; thus, the age group with the greatest needfor prevention through anticoagulation is the samegroup that is most likely to suffer from its side effects.Therefore, a careful risk-benefit analysis must beperformed is important in order to avoid both unnecessaryanticoagulation and undue contraindications.

Risk stratification and indications for anticoagulationRisk stratification strategies to select the patients whomight benefit the most from anticoagulation have

56 REVISTA ARGENTINA DE CARDIOLOGÍA / VOL 78 NÚMERO ESPECIAL - ARRITMIAS / AÑO 2010

been evaluated in some studies that have used a riskscore for ischemic stroke. This risk score assignspoints to each risk factor and the resultant sumapproximates the risk of ischemic stroke.

The Framingham score (4) considers femalegender, the progression of age, SAH, DM, ischemicstroke, and previous episodes of cerebral ischemia asrisk factors (RF) for ischemic stroke. A Framinghamscore ≤ 7 is considered low-risk, a score between 8and 13 is considered medium risk, and a score between14 and 31 is considered high risk. In practice, thisscore has not been used to determine who shouldreceive anticoagulant therapy.

The CHADS2 score (Cardiac failure, Hypertension,Age, Diabetes, Stroke) (5), assigns the presence ofcongestive heart failure with an ejection fraction lessthan 35%, hypertension, diabetes, and age >75 onepoint each of and a personal history of ischemic strokeand/or transient ischemic attack two points. A scoreof zero points means that a patient is low-risk, a scorebetween one and two means that a patient ismoderate-risk, and a score between three and sixmeans that a patient is high-risk. Anticoagulation isindicated for those who have a CHADS2 score of twoor higher.

CHADS2 ScoreCardiac failure – 1Hypertension – 1Age > 75 – 1Diabetes mellitus – 1Stroke or transient ischemic attack – 2

The international guidelines for AF (2) cite theCHADS2 score, but add valvulopathy and thyrotoxicosisto the risk stratification and suggest the use ofanticoagulants in patients who have one high-risk riskfactor or more than one moderate-risk risk factor. Forthose who have a single moderate risk factor, theguidelines suggest the use of either aspirin or oralanticoagulants. For patients with AF who have no riskfactors, the guidelines recommend using aspirin(acetylsalicylic acid-ASA) for TEP prevention.

According to the current guidelines, there isunanimous agreement that patients older than 75years of age (even those without any other RFs) aswell as patients between 65 and 75 years of age withat least one risk factor, benefit from prevention.

Considering that the diseases that are consideredrisk factors (DM, SAH, CAD, etc.) most commonly affectthe elderly (those >65 years), the majority of patientswith AF would benefit from primary prevention.

Medications for TEP PreventionThe drugs classically tested and recommended for TEPprophylaxis include OACs (warfarin) and antiplateletagents (ASA) (1-2, 6-15).

Randomized studies performed between 1989 and1992 (6-7, 11, 13-15) compared the use of warfarin,

aspirin, and/or a placebo, with regard to efficacy of TEPprevention in patients with atrial fibrillation. (1-7) Themean follow-up period in these studies varied from 15.2to 26.4 months. The results demonstrated that OACsare more effective than ASA in the primary andsecondary prevention of TEPs in patients with AF. (16)

The AFASAK study (7) compared the use ofwarfarin, aspirin, and a placebo in terms of TEPprevention in 1007 patients. In this study, patients’international normalization ratio (INR), whichmeasures the intensity of blood anticoagulation, wascontrolled between 2.8 and 4.2 and the dose of aspirinwas 75 mg. Five ischemic stroke events occurred inthe warfarin group, 20 occurred in the aspirin group,and 21 occurred in the placebo group. In the warfaringroup, 21 patients were withdrawn because of non-fatalbleeding complications compared with 2 on aspirin andnone on placebo. The efficacy of prevention was 59%with anticoagulants and 16% with aspirin (CI 95%).

The BAATAF study (8) randomized 212 patientsto warfarin (with an INR between 1.2 and 1.5) and208 patients to aspirin. Two ischemic strokes occurredin the first group and thirteen in the second. Theeffectiveness of prevention with anticoagulant use was86% (CI 95%).

The SPAF I study (9-12) compared the use ofwarfarin (with an INR level between 2 and 4.5), aspirin(325 mg), and placebo in TEP prevention in patientswith AF. There was a risk reduction of 69% in thewarfarin group. Six patients in the warfarin grouphad an ischemic stroke, but they were not using themedication correctly during the time of the event. Incontrast to the other studies, which showed thataspirin did little to prevent TEPs, the effectivenessof prevention of aspirin was 42% in this study.

However, there was controversy regarding theselection of patients for this study, because isolatedAF was included and some TEP risk factors were usedas exclusion criteria. Therefore, the patients in thisstudy had a lower risk of ischemic complications thanthe cohorts included in other studies. SPAF II (10,12) followed SPAF I. In SPAF II, there was no placebogroup and patients were split into two groups basedon age (<75 and ≥ 75).

The results of SPAF II showed that warfarin wasmore effective than aspirin in the prevention of TEPs

AF AND STROKE / Darío C. Sobral Filho, et al 57

in both groups and that the older patients on warfarinsuffered more cerebral hemorrhages than those whotook aspirin. Nevertheless, the goal of the intensityof anticoagulation went INR 2.0 – 4.5. (12) SPAF III(11-12) classified patients as high-risk or low-risk withregard to the occurrence of TEPs. Those who hadsystolic BPs greater than 160 mmHg, left ventriculardysfunction, CHF, previous TEPs, as well as those whowere female and those older than 75 years of age wereconsidered high-risk and randomized into two groups:one that received OACs with a goal INR between 2and 3 and another that received aspirin (325 mg)mixed with an OAC with a goal INR between 1.2 and1.5, measured at 3-month intervals. The study endedprematurely because the group with the lower goalINR exhibited more ischemic events, whereas thegroup with an INR between 2 and 3 exhibited asignificant decrease in the risk of TEPs.

Researchers from the CAFA study (13) began adouble-blind, controlled study to evaluate warfarincompared with a placebo for TEP prevention inpatients with AF. However, when the initial results ofthe AFASAK and SPAF studies illustrated thesuperiority of the anticoagulant over aspirin, the studyended early so that the enrolled patients could receivewarfarin therapy.

The researchers from the SPINAF study (14)compared warfarin with a placebo and concluded thatthe anticoagulation exhibited a 79% risk reduction forischemic stroke.

After the aforementioned randomized studiesregarding primary prevention for TEPs had beencompleted, the EAFT study (15) investigated theeffectiveness of warfarin and aspirin for secondaryprevention compared with a placebo. The effectivenessof warfarin was 66% and that of aspirin was 18%. Therate of ischemic stroke in the placebo group was 12%per year, in which was higher than the rates observedin the studies on primary prevention.

In 1999, a meta-analysis of five of the initial studies(6) concluded that anticoagulation is highly effectivein the prevention of ischemic stroke with a riskreduction of 65% (95% CI: 52% - 74%), while aspirinreduced the risk of ischemic stroke by 23% (95% CI:0% - 40%) when compared to placebo. When comparedwith each other, warfarin had a risk reduction of 46%as compared to aspirin (95% CI: 27% - 60%).

In 2002, the authors of the AFFIRM study (17)published their results evaluating what is the bestbetter strategy in patients with AF: rate control orrhythm control.

Throughout the follow-up period, OACs wereappropriately prescribed for TEP prophylaxis basedon the existence of validated studies in the literaturedemonstrating that OAC use decreased the risk ofTEPs in patients with AF. The authors concluded that,independent of the chosen strategy (rate vs. rhythmcontrol), adequate anticoagulation was the most

important factor in the prevention of thromboembolicevents in the patients evaluated.

The ACTIVE study (a multicenter, double-blind,randomized trial) compared warfarin and clopidogrelin patients with AF and risk factors for stroke in its“W” trial. Clopidogrel was observed to be inferior inrelation to warfarin for TEP prevention, leading toan early discontinuation of this trial. (16)

A 2007 meta-analysis (18) evaluated 13 studies andconcluded that warfarin reduces the risk of ischemicstroke by 64% and aspirin reduces the risk of ischemicstroke by 22%; thereby concluding that warfarin wasmore effective than aspirin with a relative risk reductionof 39%. The increase in the risk of intracranialhemorrhages was 0.3% per year in the warfarin group.

Some studies evaluated the use of ximelagatran,an OAC with protection indices and hemorrhagiccomplication rates similar to those of warfarin thathas the advantage of not needing to monitor andcontrol patients’ INRs. However, it was found to causesevere hepatotoxicity, which has become animpediment in its clinical use. (19)

The continuation of the ACTIVE study with its“A” trial evaluated the combination of aspirin andclopidogrel in comparison to aspirin alone in patientswith AF and an increased risk of thromboembolismthat were unable to receive a prescription for oralanticoagulants for various reasons. The combinationwas found to prevent thromboembolic events(especially ischemic strokes), albeit with a significantincrease in the risk of hemorrhagic complications. (20)

Recently, the RE-LY study (21) was published, whichcompared the use of the thrombin inhibitor dabigatranwith warfarin in a large multicenter randomized studyinvolving 18,113 patients with AF who had risk factorsfor the development of TEPs. This study showed thatdabigatran (administered at a dosage of 110 mg) wasassociated with a rate of systemic and cerebralthromboembolic events that was similar to thatobserved in patients taking warfarin. Of note, therewere fewer serious hemorrhagic complications observedat the 110 mg dosage. However, when the dosage ofdabigatran was increased to 150 mg, lower rates ofthromboembolic events were observed in the group ofpatients taking dabigatran (as compared to the groupof patients taking warfarin), but the rate of serioushemorrhagic complications increased to a rate that wassimilar to that of patients taking warfarin. Theannouncement of these results has had a great impactto the medical community because dabigatran is thefirst oral drug to have a better performance thanwarfarin and it has the advantage of no requirementto monitor and control patients’ INRs.

The intensity of anticoagulation andassociated complicationsINR monitoring and control in anticoagulated patientsis important because patients should be anticoagulated

58 REVISTA ARGENTINA DE CARDIOLOGÍA / VOL 78 NÚMERO ESPECIAL - ARRITMIAS / AÑO 2010

to the degree that their risk for ischemic events islowered, but not to the degree that increases theirrisk of hemorrhagic complications. An INR between2 and 3 represents the therapeutic window in whichthese two objectives are obtained. (22) Generally,hemorrhagic events occur with an INR of >4. Someauthors report that an INR as low as 1.6 provideseffective protection. (23) Based on these data, otherpublications (2-3) recommend a goal INR of between1.7 and 2.5 in the elderly.

Adjusted odds ratios for ischemic stroke andintracranial hemorrhage in relation to intensity ofanticoagulation (based on INR) in randomized trialsof antithrombotic therapy for patients with atrialfibrillation (2)

Some studies have reported a higher incidence ofischemic stroke in patients with INRs of <2. This wasalso associated with more severe neurological sequelaeand a higher mortality rate in those patients ascompared to patients with an INR of e”2 (24-25). Alater study (26) concluded that the following are riskfactors for intracranial hemorrhage: advanced age,uncontrolled SAH, previous ischemic stroke, andanticoagulation intensity. Blood pressure control anda goal INR of between 2 and 2.5 seem to reduce therisk of TEP, however there is a clear need forrandomized studies to better define the goal INR inthese patients.

In 2006, a new analysis was done using the datapublished since 2003 regarding the intensity ofanticoagulation. (27) It concluded that anticoagulationwith a goal INR of between 2 and 2.5 for patients withnonvalvular atrial fibrillation are just as effective asmore intensive anticoagulation.

The Problem of the Underutilization of OACs inthe ElderlyA controversial factor in TEP prophylaxis in theelderly is the risk of inducing hemorrhagiccomplications through the use of anticoagulant drugs,especially in patients over 75. (16, 31) On the otherhand, is it exactly this same age group that has an

increased risk of ischemic stroke, or rather, this wouldbe the group most adversely affected by not using OACin terms of ischemic stroke risk. In reality, somestudies have found an increased risk of cerebralhemorrhage in this age group (32), but it has not beenproven that patients with AF should not undergoprophylaxis for that reason alone. In the end, the riskof cerebral hemorrhage corresponds to 0.6% per year,which is much smaller annual risk than that conferredby thromboembolic complications (22) if adequateprophylaxis is not prescribed.

In spite of the innumerable randomized studiesthat have clarified that OACs are the best choice forTEP prevention, other articles have evaluatedwhether or not the guidelines established by thesestudies were actually being followed. They have calledattention to the underutilization of OACs in theelderly (29-38): just 30% of elderly patients with aformal indication for OAC use are anticoagulated,characterizing a clinical underutilization. Patientswith no contraindications to OAC use do not receivethe recommended therapy for more than a decade inmany cases, simply because there is a fear ofhemorrhagic complications from anticoagulation, inspite of the fact that the literature has already proventhat the goal INR levels espoused in the literatureare both safe and effective. Often, anticoagulation isonly started after the first ischemic stroke.

CONCLUSIONS

The best way to prevent TEPs in patients with atrialfibrillation has been clearly defined. Of course, thereare situations in which formal contraindications toOAC use exist for organic, psychosocial, andgeographical reasons, which should be respectedbecause of the implication that the risk is greater thanthe benefit for the elderly with AF in these cases.However, often, despite the absence of thesecontraindications, OAC therapy is not initiated,predisposing patients to the unnecessary risk ofthromboembolic complications that can lead toirreversible neurological sequelae or death.

It is up to clinicians to try, whenever possible, tocollect data on patients’ medical history, to evaluatepatients’ organic risk factors, and to educate bothpatients and their family about the risks of TEPs inan attempt to establish the best therapy in a givenpatient. The advent of the use of new anticoagulantagents with a better safety and effectiveness profileswill make it easier to approach the issue oralanticoagulation, especially in elderly patients.

REFERENCES

1. Zimerman LI, Fenelon G, Martinelli Filho M, Grupi C, Atié J, LorgaFilho A, et al. Sociedade Brasileira de Cardiologia. DiretrizesBrasileiras de Fibrilação Atrial. Arq Bras Cardiol 2009;92:1-39.2. Fuster V, Rydén LE, Cannom DS, Crijns HJ, Curtis AB, EllenbogenKA, et al; Task Force on Practice Guidelines, American College of

AF AND STROKE / Darío C. Sobral Filho, et al 59

Cardiology/American Heart Association; Committee for PracticeGuidelines, European Society of Cardiology; European Heart RhythmAssociation; Heart Rhythm Society. ACC/AHA/ESC 2006 guidelinesfor the management of patients with atrial fibrillation-executivesummary: a report of the American College of Cardiology/AmericanHeart Association Task Force on Practice Guidelines and theEuropean Society of Cardiology Committee for Practice Guidelines(Writing Committee to Revise the 2001 Guidelines for theManagement of Patients with Atrial Fibrillation). Eur Heart J2006;27:1979-2030.3. Risk factors for stroke and efficacy of antithrombotic therapy inatrial fibrillation. Analysis of pooled data from five randomizedcontrolled trials. Arch Intern Med 1994;154:1449–57.4. Gage BF, van Walraven C, Pearce L, Hart RG, Koudstaal PJ, BoodeBS, et al. Selecting patients with atrial fibrillation for anticoagulation:stroke risk stratification in patients taking aspirin. Circulation2004;110:2287-92.5. Inoue H, Nozawa T, Hirai T, Iwasa A, Okumura K, Lee JD, et al.Accumulation of risk factors increases risk of thromboembolic eventsin patients with nonvalvular atrial fibrillation. Circ J 2006;70:651-6.6. Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotictherapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med 1999;131:492-501.7. Petersen P, Boysen G, Godtfredsen J, Andersen ED, Andersen B.Placebo-controlled, randomized trial of warfarin and aspirin forprevention of thromembolic complications in crohinic Atrial Fibrilationtrial. The Copenhagen AFASAK Study. Lancet 1989;1:175-9.8. The effect of low-dose warfarin on the risk of stroke in patientswith nonrheumatic atrial fibrillation. The Boston AreaAnticoagulation Trial for Atrial Fibrillation Investigators. N Engl JMed 1990;323:1505-11.9. Stroke Prevention in Atrial Fibrillation Study. Final results.Circulation 1991;84:527-39.10. Warfarin versus aspirin for prevention of thromboembolism inatrial fibrillation: Stroke Prevention in Atrial Fibrillation II Study.Lancet 1994;343:687-91.11. Adjusted-dose warfarin versus low-intensity, fixed-dose warfarinplus aspirin for high-risk patients with atrial fibrillation: StrokePrevention in Atrial Fibrillation III randomised clinical trial. Lancet1996;348:633-8.12. Hart RG, Halperin JL, Pearce LA, Anderson DC, Kronmal RA,McBride R, et al; Stroke Prevention in Atrial FibrillationInvestigators. Lessons from the Stroke Prevention in AtrialFibrillation trials. Ann Intern Med 2003;138:831-8.13. Connolly SJ, Laupacis A, Gent M, Roberts RS, Cairns JA, JoynerC. Canadian Atrial Fibrillation Anticoagulation (CAFA) Study. J AmColl Cardiol 1991;18:349-55.14. Ezekowitz MD, Bridgers SL, James KE, Carliner NH, CollingCL, Gornick CC, et al. Warfarin in the prevention of stroke associatedwith nonrheumatic atrial fibrillation. Veterans Affairs StrokePrevention in Nonrheumatic Atrial Fibrillation Investigators. N EnglJ Med. 1992;327:1406-12.15. EAFT The European atrial Fibrilation Trial Group . SecondaryPrevention in nonrheumatic Atrial Fibrillation After TransietIschaemic Attack or Minor Stroke. Lancet 1993;342:1256-62.16. CTIVE Writing Group of the ACTIVE Investigators, Connolly S,Pogue J, Hart R, Pfeffer M, Hohnloser S, Chrolavicius S.ClopidogrelPlus Aspirin Versus Oral Anticoagulation for Atrial Fibrillation inthe Atrial Fibrillation Clopidogrel Trial With Ibesartan for Preventionof Vascular Events (ACTVE W) a randomized controlled trial. Lancet2006; 367:1903-12.17. Wyse DG, Waldo AL, DiMarco JP, Domanski MJ, Rosenberg Y,Schron EB, et al; Atrial Fibrillation Follow-up Investigation ofRhythm Management (AFFIRM) Investigators. A comparison of ratecontrol and rhythm control in patients with atrial fibrillation. N EnglJ Med. 2002;347:1825-33.18. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotictherapy to prevent stroke in patients who have nonvalvular atrialfibrillation. Ann Intern Med. 2007;146:857-67.

19. Testa L, Andreotti F, Biondi Zoccai GG, Burzotta F, Bellocci F,Crea F. Ximelagatran/melagatran against conventionalanticoagulation: a meta-analysis based on 22,639 patients. Int JCardiol. 2007;122:117-24.20. ACTIVE Investigators, Connolly SJ, Pogue J, Hart RG,Hohnloser SH, Pfeffer M, Chrolavicius S, et al. Effect of clopidogreladded to aspirin in patients with atrial fibrillation. N Engl J Med2009;360: 2066-78.21. 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 EnglJ Med. 2009;361:1139-51.22. Wittkowsky. AK. Effective Anticoagulation Therapy:Defining theGap Between Clinical Studies and Clinical Practice. Am J ManagCare 2004;10: s297- 306.23. Ono A, Fujita T. Low-intensity anticoagulation for strokeprevention in elderly patients with atrial fibrillation: efficacy andsafety in actual clinical practice. J Clin Neurosci 2005;12:891-424. Hylek EM, Go AS, Chang Y, Jensvold NG, Henault LE, SelbyJV, et al. Effect of intensity of oral anticoagulation on stroke severityand mortality in atrial fibrillation. N Engl J Med. 2003;349:1019-26.25. Fang MC, Chang Y, Hylek EM, Rosand J, Greenberg SM, Go AS,et al. Advanced age, anticoagulation intensity, and risk for intracranialhemorrhage among patients taking warfarin for atrial fibrillation.Ann Intern Med. 2004;141:745-52.26. Hart RG, Tonarelli SB, Pearce LA. Avoiding central nervoussystem bleeding during antithrombotic therapy: recent data and ideas.Stroke. 2005;36:1588-93.27. Odén A, Fahlén M, Hart RG. Optimal INR for prevention of strokeand death in atrial fibrillation: a critical appraisal. Thromb Res.2006;117:493-9.28. Torn M, Bollen WL, van der Meer FJ, van der Wall EE, RosendaalFR. Risks of oral anticoagulant therapy with increasing age. ArchIntern Med. 2005;165:1527-32.29. Gage BF, Fihn SD, White RH. Warfarin therapy for an octogenarianwho has atrial fibrillation. Ann Intern Med. 2001;134:465-74.30. Kamath S, Lip GY. Atrial fibrillation in the elderly:anticoagulation strategies and indications in the very elderly. Am JGeriatr Cardiol. 2002;11:357-62.31. Fihn SD, Callahan CM, Martin DC, McDonell MB, Henikoff JG,White RH. The risk for and severity of bleeding complications inelderly patients treated with warfarin. The National Consortium ofAnticoagulation Clinics. Ann Intern Med. 1996;124:970-9.32. Hylek EM, Singer DE. Risk factors for intracranial hemorrhagein outpatients taking warfarin. Ann Intern Med. 1994;120:897-902.33. White RH, McBurnie MA, Manolio T, Furberg CD, Gardin JM,Kittner SJ, et al. Oral Anticoagulation in patients with atrialfibrillation: adherence with guidelines in an elderly cohort.Am J Med1999;106:165-71.34. Gayoso Diz P, Calle Custodio R, Prieto Maroto A, Herrera Calvo D,Sala López AI, Gómez Mosquera MD. Auricular Fibrillation as a riskfactor of cerebrovascular events in the over-65 s.Is clinical practice inanticoagulante prophylaxis adequate. Aten primaria 2005;36:198-203.35. Sobral Filho D. Principais causas de subutilização deanticoagulantes em pacientes com fibrilação atrial associada afatores de risco para tromboembolismo. Replampa 2005;18:197(Resumen).36. Tapson VF, Hyers TM, Waldo AL, Ballard DJ, Becker RC, CapriniJA, et al; NABOR (National Anticoagulation Benchmark andOutcomes Report) Steering Committee. Antithrombotic therapypractices in US hospitals in an era of practice guidelines. Arch InternMed 2005;165:1458-64.37. Cabral NL, Volpato D, Ogata TR, Ramirez T, Moro C, Gouveia S.Atrial fibrillation, stroke and anticoagulation: under-use of warfarin?.Arq Neuropsiquiatr 2004;62:1016-21.38. Mesas CE, Veloso HH, De Paola AA. Anticoagulation for atrialfibrillation: underutilization in a Brazilian tertiary outpatient clinic.Clin Cardiol. 2004;27:592-3.