Return to PJ Online Home Page
The Pharmaceutical Journal Vol 264 No 7097 p773-777
May 20, 2000 Continuing education

Atrial fibrillation

(3) Antithrombotic therapy

By Gregory Y. H. Lip, MD, FRCPE, and Sridhar Kamath, MRCP

Patients with atrial fibrillation are at risk of stroke and thromboembolism from thrombus formation in the atria. Antithrombotic therapy is therefore a key part of the management of the condition
The first article in the series discusses the condition and the second article covers antiarrhythmic agents

Atrial fibrillation (AF) is the most common cardiac disorder leading to stroke and thromboembolism. The presence of non-valvular AF increases the risk of stroke approximately 4- to 5-fold. The importance of AF as a risk factor for stroke substantially increases with age.1 AF is usually present in about 15-20 per cent of patients with acute stroke, and is associated with a 1.5 to 3.0 fold higher mortality than that for stroke patients who are in sinus rhythm.2 Strokes attributed to AF also tend to be more severe, with greater disability, longer hospital stay and lower rate of discharge to the patient's own home.3,4 There is also a high recurrence rate and silent cerebral infarcts frequently occur.
Most of the stroke and thromboembolic events in patients with AF are due to thrombus formation in the fibrillating left atria. The loss of atrial systole and increased atrial stasis (visualised as the phenomenon of spontaneous echocontrast using transoesophageal echocardiography), in association with abnormalities of haemostasis and platelet activation in AF, all predispose to thrombus formation (thrombogenesis). Risk factors for stroke, such as valve disease, hypertension and heart failure, are additive to the risk associated with AF per se.5 For example, in patients with mitral valve stenosis, the risk of thrombogenesis is increased 17-fold.6

Prevention of thromboembolism in AF

In the past two decades, there have been substantial advances in the provision of thromboprophylaxis in AF, with publication of randomised controlled trials establishing the benefits from anticoagulation for the prevention of stroke and thromboembolism.5,7 Many different antithrombotic agents have been used to reduce the risk of thromboembolism in AF, including coumarin anticoagulants (eg, warfarin), antiplatelet drugs (eg, aspirin) and parenteral anticoagulants (eg, heparin).

Warfarin Warfarin, which is administered orally, is an antagonist of vitamin K, which is a cofactor for a microsomal enzyme that activates coagulation factors II, VII, IX, and X by carboxylation of their glutamic residues.8 Administration of warfarin in therapeutic doses decreases the synthesis of these coagulation factors in the liver by approximately 40 per cent, resulting in hypocoagulability of the blood and reducing the propensity towards thrombus formation in AF.
Warfarin has no influence on the already activated circulating coagulation factors in the blood, and thus the stabilisation of the prothrombin time ratio depends on the clearance rate of the already formed coagulation factors in the blood, which may take two to three days. Monitoring of the intensity of anticoagulation by warfarin is best done by determining the prothrombin time (PT) in comparison to that of a control plasma sample. The desired level of international normalised ratio (INR) should be the one that would provide greatest benefit in terms of reducing the thromboembolic risk while subjecting the patient to the least bleeding risk.
Traditionally, measurement of a patient's anticoagulation has been in busy, hospital-based clinics, despite some evidence that more efficient anticoagulation control is achieved if general practitioners undertake anticoagulation monitoring.9 Computer programs have been developed to manage warfarin dosing, and their reliability appears to be either as good or better than dosing by experienced medical staff.10 Self-testing11 and even self-management of anticoagulation intensity12 by the patient results in as good or even better control of anticoagulation than physician-managed anticoagulation, and may become widespread practice in the future, especially with increasing numbers of patients taking anticoagulant therapy for conditions such as AF.

Heparin The prime role of heparin in AF is when immediate anticoagulation is desired until adequate INR is attained with warfarin or the patient needs to undergo urgent cardioversion. Heparin antagonises the action of antithrombin III, which in turn rapidly inhibits the activated coagulation factors (such as thrombin), leading to an anticoagulant effect. Therapy is administrated by an intravenous infusion and monitored by measurement of the activated partial thromboplastin time (APTT), which is sensitive to the levels of thrombin, factor Xa, and factor IXa. The therapeutic range of APTT depends on the commercial reagent used to measure it. On average, a value of over 1.5 has a therapeutic effect. The dose of heparin needed to maintain APTT within therapeutic levels varies widely and different methods of deciding the right dose for the patient have been proposed. These include dose-adjustment nomograms, weight-adjusted nomograms, and computer algorithms.
Recently, low molecular weight heparin (LMWH) preparations, which are administered subcutaneously and act by facilitating the inhibition of coagulation factor Xa by antithrombin, have been introduced with good results in AF.13
LMWH does not require monitoring by measurement of the APTT, and its ease of administration has resulted in increasing use in place of the traditional unfractionated heparin infusion.
The risk of bleeding is increased in patients who are on heparin. Heparin-induced thrombocytopenia is another troublesome complication of unfractionated heparin therapy. Other complications with the latter include hypersensitivity reactions, skin necrosis and, after prolonged use, alopecia and osteoporosis.

Aspirin Agents such as aspirin exert their antiplatelet action by irreversibly inhibiting the platelet cyclo-oxygenase enzyme, which in turn results in the inhibition of conversion of platelet arachidonic acid to prostaglandin intermediates. This deficiency of prostaglandin intermediates exerts anti-aggregatory effects on the platelets during their entire lifetime. This results in a deficiency of platelet plug formation and inactivated platelets fail to facilitate the coagulation process.
Aspirin has widespread use in cardiovascular and cerebrovascular disease. The meta-analyses by the Antiplatelet Trialists' Collaboration14 have confirmed significant reductions in thrombotic events by using aspirin. Trials in patients with AF have largely confirmed a beneficial effect of aspirin, especially in low risk patient groups.15

Antihrombotic therapy for AF in the clinical setting

The provision of thromboprophylaxis in AF includes the use of heparin in acute AF, the use of warfarin for cardioversion of persistent AF, and the use of aspirin or warfarin in paroxysmal or permanent AF.

Acute AF In patients presenting de novo with AF, a clear history of arrhythmia onset is often necessary. If the duration of AF is clearly less than 48 hours, cardioversion may be performed without the need for long-term anticoagulation post-cardioversion,16 although there are no randomised trials specifically answering this question. Nevertheless, in one series, intra-atrial thrombus was detected by transoesophageal echocardiography in approximately 15 per cent of patients presenting with acute onset (less than 48 hours) AF,17 although the possibility remains that many of these patients had developed AF asymptomatically.
Manning et al18 reported that left atrial thrombi were identified in 43 per cent of patients with acute thromboembolism and newly recognised AF, compared with 10 per cent of controls, while spontaneous echocontrast (indicating stasis) was identified in 87 per cent and 48 per cent, respectively. However, as discussed above, many patients may have developed AF asymptomatically and, in cases of uncertainty, anticoagulation is warranted. Thus, patients presenting de novo with AF should be started on intravenous heparin following admission, achieving an APTT ratio of 2-3, which should reduce the risk of thrombus formation.

Anticoagulation and cardioversion of persistent AF In cases where the duration of AF is over 48 hours, anticoagulation with warfarin (INR 2-3) should be started for a minimum of three weeks and, if appropriate, cardioversion attempted and anticoagulation continued for at least four weeks. Recommendations from the American College of Chest Physicians19 regarding anticoagulation for cardioversion of AF are summarised in Panel 1. Recently, transoesophageal echocardiography has evolved as an attractive alternative to prolonged anticoagulation before DC cardioversion.20 One pilot trial (the ACUTE pilot study) of 126 patients with AF that had been present for over two days who were having electrical cardioversion reported that atrial thrombi were detected in 13 per cent; the use of transoesophageal echocardiography may allow cardioversion to be done earlier, decreasing the risk for embolism, and may be associated with less clinical instability than conventional therapy.20
A recent review suggested that the risk of thromboembolism following cardioversion for AF without anticoagulation is 5 per cent; after transoesophageal echocardiography-guided cardioversion this risk is 1.3 per cent, compared with 0.3-0.8 per cent following a conventional anticoagulation regime.21

Anticoagulation for paroxysmal and permanent AF In a recent meta-analysis of antithrombotic therapy in AF, the use of anticoagulation has been shown to reduce the risk of stroke by 62 per cent (95 per cent confidence intervals [CI] 48-72 per cent), while aspirin reduces the risk by 22 per cent (95 per cent CI 2-38 per cent).22 Absolute risk reductions with warfarin were 1.5 per cent per year for primary prevention and 2.5 per cent per year for secondary prevention. Furthermore, adjusted dose warfarin reduced the risk by 36 per cent (95 per cent CI 14-52 per cent) relative to aspirin.22 In the pooled meta-analysis by the AF investigators of the five initial primary prevention trials, an annual rate of all strokes of 4.5 per cent was found in the control group and a rate of 1.4 per cent in the warfarin group. This difference represents an overall risk reduction of 68 per cent (95 per cent CI 50 to 79 per cent).23
Warfarin prophylaxis also resulted in a decrease in mortality of 33 per cent, and a decrease in the combined adverse outcome (stroke, systemic embolism or death) of 48 per cent.23 Warfarin usage was particularly beneficial as secondary prevention in the EAFT study, where the risk of stroke was reduced from 12 per cent to 4 per cent per year (hazard ratio 0.34, 95 per cent CI 0.20-0.57).24 Despite these impressive figures, the efficacy of warfarin prophylaxis is probably underestimated, because most strokes in patients randomised to warfarin occurred while the patients were not in fact taking warfarin or while they were significantly underanticoagulated. Overall, the data from the trials suggest that treating 1,000 patients with warfarin over 12 months will prevent 30 strokes. A similar reduction in relative risk for secondary prevention translates into the prevention of 80 strokes a year for every 1,000 patients treated. In these trials, the overall incidence of haemorrhage was 1.3 per cent a year for those receiving warfarin compared with 0.9 per cent a year for the control group.25,26
Patients with paroxysmal AF are at similar stroke risk to patients with sustained AF.23,27 The most recent analysis from the Stroke Prevention in AF Investigators27 reported that 460 patients with paroxysmal AF had stroke rates and risk factors similar to patients in sustained AF, and that these patients were likely to benefit from anticoagulation. There is also a theoretical risk that embolisation of pre-existing thrombus is more likely during onset (and reversion) of paroxysms of AF, and indeed, some data have shown an increased risk of stroke during this time.28 Thus, patients with paroxysmal AF should be managed in a similar way to patients with sustained AF, with long-term anticoagulation for high risk patients.29 If these patients are not anticoagulated when they present with an acute episode of AF, they should be managed with intravenous heparin and (if the duration of AF is over 48 hours) warfarin, as described above, prior to cardioversion.

Panel 1: Recommendations for anticoagulation for cardioversion

  • The administration of warfarin for three weeks before elective cardioversion of AF of 48 hours' or more duration; continuation of warfarin therapy for four weeks after cardioversion
  • Administration of intravenous heparin followed by warfarin if cardioversion cannot be postponed for three weeks
  • Treat atrial flutter similarly
  • No anticoagulant therapy for supraventricular tachycardia or atrial fibrillation of > 48 hours' duration

Based on the 5th ACCP consensus conference on antithrombotic therapy19

Anticoagulation in AF patients presenting with stroke Some patients with acute AF present with an acute stroke or thromboembolic event. The debate over the role of thrombolytic therapy for early acute ischaemic stroke is beyond the scope of this review. Nevertheless, in cases of established thromboembolic strokes related to acute AF, warfarin (target INR 2-3) could be started in non-hypertensive patients with small sized strokes (after a CT scan or MRI scan done 48 hours or more after symptom onset confirms the absence of intracranial haemorrhage), although this recommendation is not based on any evidence from randomised controlled trials.30 Anticoagulation should however be postponed for one to two weeks in patients with large thromboembolic strokes due to the potential risk of haemorrhagic transformation.30 The use of aspirin in the International Stroke Trial31 and the Chinese Acute Stroke Trial32 led to a marginal benefit in mortality but this was less among patients with AF at time of presentation, possibly because of embolisation of pre-existing thrombi rather than the prevention of new thrombus formation per se.33

Risk stratification

The risk of stroke and thromboembolic events in AF varies widely depending on the presence or absence of associated risk factors. Thus, risk stratification of a patient with AF is useful in assessing the risk and benefit of starting anti-thrombotic therapy. Various clinical and echocardiographic parameters predict the risk of stroke in patients with AF (Panel 2).34 It should therefore be feasible to identify "high risk" patients who would benefit most from anticoagulation.
Aspirin should be administered to patients who have contraindications to warfarin therapy. Nevertheless, patient preference plays a vital role in choosing the treatment strategy and the patients should be well informed about the benefits and the risks of therapy. Importantly, risk stratification for thromboprophylaxis in AF is not a static process and patients who are judged to be at low risk of stroke and are not given warfarin should be periodically evaluated for the possible development of high risk features favouring anticoagulation.

Panel 2: Risk stratification and anticoagulation in non-valvular atrial fibrillation (NVAF)*

ASSESS RISK
1. High risk (annual risk of cerebrovascular accident [CVA] = 8-12 per cent)

  • All patients with NVAF and previous transient ischaemic attack or CVA
  • All patients aged 75 and over with NVAF and diabetes and/or hypertension
  • All patients with NVAF and clinical evidence of valve disease, heart failure, thyroid disease and/or impaired left ventricular function on echocardiography**

2. Moderate risk (annual risk of CVA = 4 per cent)

  • All patients under 65 with NVAF and clinical risk factors: diabetes, hypertension, peripheral arterial disease, ischaemic heart disease
  • All patients over 65 with NVAF who have not been identified in high risk group

3. Low risk (annual risk of CVA = 1 per cent)

  • All other patients under 65 with NVAF with no history of embolism, hypertension, diabetes or other clinical risk factors

TREATMENT
High risk: Warfarin (target INR 2.0-3.0) if no contraindications and possible in practice
Moderate risk: Either warfarin or aspirin. In view of insufficient clearcut evidence, treatment may be decided on individual basis. Referral and echocardiography may help
Low risk: Aspirin 75-300mg daily

REASSESS RISK FACTORS AT REGULAR INTERVALS

*adapted from reference 34
**Echocardiogram not needed for routine risk assessment but refines clinical risk stratification in cases of impaired left ventricular function and valve disease (see 1 above). A large left atrium per se is not an independent risk factor on multivariate analysis

Risks of antithrombotic therapy

The annual risks of intracranial haemorrhage increase from 0.1 per cent in controls to 0.3 per cent in warfarin groups in the pooled analysis of the AF investigators.23 This represents an excess of two intracranial bleeds per annum per thousand patients treated. The bleeding risk with anticoagulation was particularly associated with an INR greater than 3.0, fluctuating INRs and uncontrolled hypertension.35 A risk of falling, especially in the elderly population, is also perceived by some physicians as a contraindication to antithrombotic therapy due to the fear of intracranial haemorrhage, although a recent study suggests that this concern may be overstated.36
An association between intracranial haemorrhage and increasing age has also been found. In the SPAF-II study,37 for example, the benefits of warfarin were nearly counteracted by an increased risk of intracranial haemorrhage in the warfarin group, which was unexpectedly high in patients over 75 years (1.8 per cent/year); however, the rate of intracranial haemorrhage in this group was substantially higher than that reported in the original five trials (on average 0.3 per cent/year),which may reflect the higher intensity of anticoagulation in the SPAF-II study (INR up to 4.5).
Randomised trials with target INRs between 2.7 and 4.8 show a 10-fold increase in intracranial bleeding among patients given anticoagulants compared with placebo.26 Bleeding usually occurs into the skin, soft tissues, oropharynx, gastrointestinal tract and the urinary tract. Even though intracranial bleeding accounts for only 2 per cent of anticoagulation-related bleeding, this serious complication could be fatal.
Careful patient selection (assessing for bleeding risk and compliance), patient education, INR monitoring, and regular medical supervision play vital roles in reducing the bleeding risks of anticoagulation therapy for AF.
Aspirin, even in low doses, could increase the rate of major haemorrhage by 0.5 per cent/year in elderly people,38 although overall risk of major bleeding associated with antiplatelet therapy generally appears to be low, at less than 1 per cent per year. Nevertheless, the adverse effects of aspirin on the gastrointestinal tract, which are often dose-related, could vary from dyspepsia to life-threatening haemorrhage.

Other strategies in providing thromboprophylaxis

Other antiplatelet agents One possible strategy in providing thromboprophylaxis is the use of new antiplatelet agents, or combination therapy of agents that inhibit the platelet via different pathways. For example, a single study compared the efficacy of indobufen, a new antiplatelet agent, against warfarin in secondary prevention of stroke in AF. Fewer strokes occurred in patients taking warfarin, with reduced bleeding risk in the indobufen group.39
In the second European Stroke Prevention Study (ESPS-2),40 which was a secondary prevention study, low-dose aspirin (50mg) alone, sustained release dipyridamole (400mg/day) alone, or a combination of the two agents, were compared with placebo over 24 months. This study found that dipyridamole reduced the risk of stroke by 22 per cent relative to placebo, while the combination of dipyridamole and aspirin reduced the risk of stroke by 43 per cent relative to placebo; combination therapy was superior to either agent alone in preventing stroke.
Although only a small proportion of patients in ESPS-2 had AF, a subgroup analysis suggested non-significant trends towards benefits in patients with AF taking aspirin-dipyridamole combination therapy.

Low-intensity warfarin-aspirin combination therapy and fixed dose warfarin These treatment regimes were tested in a series of randomised trials conducted in the 1990s which compared fixed low-dose warfarin regimes (eg, warfarin 1.25mg), or low-intensity warfarin (INR less than 1.5) in combination with aspirin, as alternatives to conventional anticoagulation (INR2-3) or aspirin in patients with AF (as reviewed by Lip34).
The failure of these regimes to provide adequate thromboprophylaxis in AF comes as little surprise as analyses of the optimal anticoagulation intensity for stroke prevention in AF demonstrated that stroke risk appeared to be substantially increased at INR levels less than 2.0. In a case-controlled study by Hylek et al,41 the INR was a powerful and independent determinant of the risk of stroke. For example, when compared with patients with an INR of 2.0, those with an INR of 1.7 had a two-fold increase in the risk of stroke.41 The risk of bleeding substantially increases at INRs greater than 3.0 and, thus, the optimal INR range for patients with AF is 2.0-3.0.

From trials to clinical practice

Care should be employed when interpreting the results of the clinical trials, which demonstrate unequivocal benefit of anticoagulation in AF. These trials were conducted on a carefully selected patient population and anticoagulation was strictly monitored in well-motivated investigators and patients, who essentially received careful follow up and "packages of care".
Whether these results could be fully extrapolated to general clinical practice is not entirely clear. While efforts have been made to attain optimum management of anticoagulation as shown in the trials, many of the observational studies in United Kingdom indicate that there is considerable variation in the antithrombotic management of AF and still a great proportion of eligible patients are not adequately managed.42,43
Until we can develop more effective, efficient and safe methods of providing adequate thromboprophylaxis, efforts at careful risk stratification based on clinical risk factors (with some refinement from echocardiography) are needed.
This would allow the highest risk patients with AF to be identified and treated with warfarin, while lower risk individuals are prescribed aspirin. Concurrent co-morbidity, especially in the elderly, needs to be carefully evaluated and considered in any treatment strategy.

Dr Lip is consultant cardiologist and reader in medicine and Dr Kamath is research fellow in the Haemostasis, Thrombosis and Vascular Biology Unit, University Department of Medicine,City Hospital, Birmingham

References

1. Kannel WB, Wolf PA, Benjamin EJ, Levy D. Prevalence, incidence, prognosis, and predisposing conditions for atrial fibrillation: population-based estimates. Am J Cardiol 1998;82:2N-9N.
2. Sandercock P, Bamford J, Dennis M, Burn J, Slattery J, Jones L, et al. Atrial fibrillation and stroke: prevalence in different types of stroke and influence on early and long term prognosis (Oxfordshire community stroke project) BMJ 1992;305:1460-5.
3. Jorgensen HS, Nakayama H, Reith J, Raaschou HO, Olsen TS. Stroke recurrence: predictors, severity, and prognosis. The Copenhagen Stroke Study. Neurology 1997;48:891-5.
4. Lin HJ, Wolf PA, Kelly-Hayes M, Belser AS, Kase CS, Benjamin EJ, et al. Stroke severity in atrial fibrillation. The Framingham Study. Stroke 1996;27:1760-4.
5. Atrial fibrillation investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomised controlled trials. Arch Intern Med 1994;154:1449-57.
6. Beppu S, Park YD, Sakakibara H, Nagata S, Nimura Y. Clinical features of intracardiac thrombosis based on echocardiographic observation. Jpn Circ J 1984:48:75-82.
7. 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.
8. Hirsh J, Dalen JE, Anderson DR, Poller L, Bussey H, Ansell J, et al. Oral anticoagulants: Mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest 1998;114:445S-69S.
9. Pell JP, McIver B, Stuart P, Malone DN, Alcock J. Comparison of anticoagulant control among patients attending general practice and a hospital anticoagulant clinic. Br J Gen Pract 1993;43:152-4.
10. Poller L, Shiach CR, MacCallum P, Johansen AM, Munster AM, Magalhaes A, Jespersen J. Computerised anticoagulant dosage - a multicentre randomised study. European Concerted Action on Anticoagulation. Lancet 1998;352:1505-9.
11. Beyth RJ, Landefeld CS. Prevention of major bleeding in older patients treated with warfarin:results of a randomised trial. J Gen Intern Med 1997;12:66.
12. Horstkotte D, Piper C, Wiemer M, Schulte HD, Schultheiss HP. Improvement of prognosis by home prothrombin estimation in patients with life-long anticoagulant therapy. Eur Heart J 1996;17(Suppl):230 (abstract).
13. Harenberg J, Weuster B, Pfitzer M, Dempfle CE, Stehle G, Kubler W, et al. Prophylaxis of embolic events in patients with atrial fibrillation using low molecular weight heparin. Semin Thromb Hemost 1993;19 Suppl 1:116-21.
14. Antiplatelet Trialists' Collaboration. Collaborative overview of randomized trials of antiplatelet therapy. I. Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ 1994;308:81-106.
15. Atrial Fibrillation Investigators. The efficacy of aspirin in patients with atrial fibrillation. Analysis of pooled data from 3 randomized trials. Arch Intern Med 1997;157:1237-40.
16. Final consensus statement of the Royal College of Physicians of Edinburgh consensus conference on atrial fibrillation in hospital and general practice, September 3-4, 1998. Br J Haematol 1999;104:195-6.
17. Stoddard ME, Dawkins PR, Prince CR, Ammash NM. Left atrial appendage thrombus is not uncommon in patients with acute atrial fibrillation and a recent embolic event: a transesophageal echocardiographic study. J Am Coll Cardiol 1995;25:452-9.
18. Manning, WJ, Silverman, DI, Keighley, CS, Dettigen P, Douglas PS. Transesophageal echocardiographically facilitated early cardioversion from atrial fibrillation using short-term anticoagulation: Final results of a prospective 4.5 year study. J Am Coll Cardiol 1995;25:1354.
19. Laupacis A, Albers G, Dalen J, Dunn MI, Jacobson AK, Singer DE. Antithrombotic therapy in atrial fibrillation. Chest 1998;114:579S-89S.
20. Klein AL, Grimm RA, Black IW, Leung DY, Chung MK, Vaughn SE, et al. Cardioversion guided by transesophageal echocardiography: the ACUTE Pilot Study. A randomized, controlled trial. Assessment of Cardioversion Using Transesophageal Echocardiography. Ann Intern Med 1997;126:200-9.
21. Chambers J. Is it safe to cardiovert atrial fibrillation without anticoagulation if the tranoesophageal echocardiogram is normal? Int J Cardiol 1998;63:107-9.
22. 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.
23. Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy atrial fibrillation. Arch Intern Med 1994;154:1449-5.
24. EAFT (European Atrial Fibrillation Trial) Study Group. Secondary prevention in non-rheumatic atrial fibrillation after transient ischaemic attack or minor stroke. Lancet 1993;342:1255-62.
25. Hardman SM, Cowie MR. Fortnightly review: anticoagulation in heart disease. BMJ 1999;318:238-44.
26. Hart RG, Boop B, Anderson DC. Oral anticoagulants and intra-cranial haemorrhage: facts and hypothesis. Stroke 1995;26:1471-7.
27. Hart RG, Pearce LA, Rothbart RM, McAnulty JH, Asinger RW, Halperin JL. Stroke with intermittent atrial fibrillation: incidence and predictors during aspirin therapy. J Am Coll Cardiol 2000;35:183-7.
28. Petersen P, Godfresen J. Embolic complications in paroxysmal atrial fibrillation. Stroke. 1986;17:622-5.
29. Lip GYH. Does paroxysmal atrial fibrillation confer a paroxysmal thromboembolic risk? Lancet 1997;349:1565-6.
30. Sherman DG, Dyken ML Jr, Gent M, Harrison JG, Hart RG, Mohr JP. Antithrombotic therapy for cerebrovascular disorders: an update. Chest 1995;108:444s-56s
31. International Stroke Trial (IST): a randomised trial of aspirin, subcutaneous heparin, both, or neither among 19435 patients with acute ischaemic stroke. International Stroke Trial Collaborative Group. Lancet 1997;349:1569-81.
32. CAST: randomised placebo-controlled trial of early aspirin use in 20,000 patients with acute ischaemic stroke. CAST (Chinese Acute Stroke Trial) Collaborative Group. Lancet 1997;349:1641-9.
33. Lip GYH, Beevers DG. Interpretation of IST and CAST stroke trials. Lancet 1997;350:442-3.
34. Lip GYH. Thromboprophylaxis for atrial fibrillation. Lancet 1999;353:4-6.
35. Gullov AL, Koefoed BG, Petersen P. Bleeding during warfarin and aspirin therapy in patients with atrial fibrillation. The AFASAK 2 study. Arch Intern Med 1999;159:1322-8.
36. Man-Son-Hing M, Nichol G, Lau A, Laupacis A. Choosing antithrombotic therapy for elderly patients with atrial fibrillation who are at risk for falls. Arch Intern Med 1999;159:677-85.
37. Stroke Prevention in Atrial Fibrillation Investigators. Warfarin versus aspirin for prevention of thromboembolism in atrial fibrillation: Stroke Prevention in Atrial Fibrillation II Study. Lancet 1994;343:687-91.
38. Hart RG, Sherman DG, Easton JD, Cairns JA. Prevention of stroke in patients with nonvalvular atrial fibrillation. Neurology 1998;51:674-81.
39. Morocutti C, Amabile G, Fattapposta F, Nicolosi A, Matteoli S, Trappolini M, et al. Indobufen versus warfarin in the secondary prevention of major vascular events in nonrheumatic atrial fibrillation. SIFA (Studio Italiano Fibrillazione Atriale) investigators. Stroke 1997;28:1015-21.
40. Diener HC, Cunha L, Forbes C, Sivenius J, Smetsa P, Lowenthal A. European Stroke Prevention Study. 2. Dipyridamole and acetylsalicylic acid in the secondary prevention of stroke. J Neurol Sci 1996;143:1-13.
41. Hylek EM, Skates SJ, Sheehan MA, Singer DE. An analysis of the lowest effective intensity of prophylactic anticoagulation for patients with nonrheumatic atrial fibrillation. N Engl J Med 1996;335:540-6.
42. Lip GYH, Tean KN, Dunn FG. Treatment of atrial fibrillation in a district general hospital. Br Heart J 1994;71:92-5.
43. Lip GYH, Zarifis J, Watson RD, Beevers DG. Physician variation in the management of patients with atrial fibrillation. Heart 1996;7:200-5.