What is the efficacy of a) anticoagulation versus placebo; b) anticoagulation versus antiplatelet therapy; and c) antiplatelet therapy versus placebo for stroke prevention in permanent AF patients?

Index Table
Study ReferenceIDHierarchy
van Walraven C et al, Oral anticoagulants vs aspirin in nonvalvular atrial fibrillation: an individual patient meta-analysis. JAMA: 288: 2441 – 2448. 2002 Ref 17451+
Taylor F et al, Systematic review of long term anticoagulation or antiplatelet treatment in patients with non-rheumatic atrial fibrillation. BMJ: 322: 321 – 326. 2001 Ref 175171+
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. Ref 171201+
Segal JB eal, Anticoagulants or antiplatelet therapy for non-rheumatic atrial fibrillation and flutter. The Cochrane Library. 2004. Ref 172311+
Benavente O. et al, Antiplatelet therapy for preventing stroke in patients with non-valvular atrial fibrillation and no previous history of stroke or transient ischemic attacks. The Cochrane Library. 2004 Ref 178331+
Benavente O et al., Oral anticoagulants for preventing stroke in patients with non-valvular atrial fibrillation and no previous history of stroke or transient ischemic attacks. The Cochrane Library 2004. Ref 173351+
Perret-Guillaume C and Wahl DG., Low-dose warfarin in atrial fibrillation leads to more thromboembolic events without reducing major bleeding when compared to adjusted-dose. Thrombosis & Haemostasis: . 91: 394 – 402. 2004 Ref 1774181+
Hylek E et al, Effect of intensity of oral anticoagulation on stroke severity and mortality in atrial fibrillation. New England Journal of Medicine: . 349: 1019 – 1026. 2003 Ref 1764582+
Perez-Gomez F, Alegria E, Berjon J, Iriarte JA, Zumalde J, Salvador A et al. Comparative effects of antiplatelet, anticoagulant, or combined therapy in patients with valvular and nonvalvular atrial fibrillation A randomized multicenter study. J Am Coll Cardiol 2004;44:1557–66. Ref 17918681+
NICE LEVELS OF EVIDENCE HIERARCHY
1++High-quality meta-analysis, systematic reviews of RCTs or RCTs with a very low risk of bias.
1+Well-conducted meta-analysis, systematic reviews of RCTs or RCTs with a low risk of bias.
1−Meta-analysis, systematic reviews of RCTs or RCTs with a high risk of bias.
2++High-quality systematic review of case-control or cohort studies. High-quality case-control or cohort studies with a very low risk of confounding, bias or chance and a high probability that the relationship is causal.
2+Well-conducted case-control or cohort studies with a low risk of confounding, bias or chance and a moderate probability that the relationship is causal.
2−Case-control or cohort studies with a high risk of confounding, bias or chance and a significant risk that the relationship is not causal.
3Non-analytic studies (e.g. case reports and case series).
4Expert opinion, formal consensus.
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[5]: van Walraven C et al, Oral anticoagulants vs aspirin in nonvalvular atrial fibrillation: an individual patient meta- analysis. JAMA: 288: 2441 – 2448. 2002
Study typeMeta-analysis: “Pooled analysis of patient-level data from 6 published, randomized clinical trials” (see Included studies below)
Study objective“To compare the risk of vascular and bleeding events in patients with nonvalvular AF treated with vitamin K –inhibiting oral anticoagulants or acetylsalicylic acid (aspirin).”
Evidence level1+
Number of patientsTreatment GroupN
Anticoagulation1939
Aspirin2113
Total:4052
Patient characteristicsInclusion Criteria: “A total of 4052 patients with AF randomly assigned to receive therapeutic doses of oral anticoagulant or aspirin with or without low-dose oral anticoagulants.”
AnticoagulationAspirin
Mean age (SD), years71.5 (8.9)71.9 (8.7)
% Male59%62%
% Non-paroxysmal AF82%85%
% AF > 1 year66%67%
% Previous stroke/TIA24%24%
% High stroke risk165%65%
Notes:
  1. Stroke risk is based on published Atrial Fibrillation Investigator’s criteria
InterventionAnticoagulation (international normalised ratio > 2.0)
ComparisonAspirin (75mg to 325mg)
Length of follow-up1.9 years
Outcome measuresStrokeAll, ischaemic, haemorrhagic
Cardiovascular eventsMyocardial infarction, systemic emboli, vascular death
BleedingMajor bleeding, lethal bleeding
Mortality
Effect sizeAnticoagulation1Aspirin1P
All strokes2.44.5<0.001
Ischaemic strokes2.04.3<0.001
Haemorrhagic strokes0.50.3NS2
Major Bleeding2.21.3<0.05
Mortality4.95.2NS
Notes:
  1. Effect size measured as frequency of events per 100 patient-years
  2. NS = Not significant (P>0.05)
Source of fundingNot reported. One author receives financial support from commercial organisations
Additional commentsPrimary studies included both blinded and unblinded trials. Therefore, overall grade is 1+.
NCC CC ID (Ref Man)5
Included studiesStudy NameIDPopulationComparison
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
EAFT[851]age >25 years with prior stroke or transient ischemic attackwarfarin; aspirin(300mg); placebo
PATAF[1055]age >60 years chronic or paroxysmal AFwarfarin; aspirin(150mg);
SPAF II <=75 year[1216]age< 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
SPAF II >75 years[1216]age> 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
SPAF III[880],[1214]patients at high-risk for thromboembolic event, chronic or paroxysmal AFwafarin; aspirin(325mg) and low-dose warfarin
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[17]: Taylor F et al, Systematic review of long term anticoagulation or antiplatelet treatment in patients with non rheumatic atrial fibrillation. BMJ: 322: 321 – 326. 2001
Study typeSystematic Review with Meta-analysis of 6 trials/subtrials (see Included References below)
Study objectiveComparison of anticoagulation with antiplatelet therapy in the prevention of fatal and non-fatal outcomes in atrial fibrillation
Evidence level1+
Number of patientsThere were a total of 3298 participants in the 6 trials, but the numbers in each treatment arm are not reported.
Patient characteristicsInclusion Criteria:
  • Trials of patients with non-rheumatic atrial fibrillation randomised to either long-term anticoagulation (>12 months) or antiplatelet treatment. Trials that included patients with co-morbid/precipitating thyrotoxicosis and patients with prosthetic heart valves were excluded.
InterventionWarfarin (international normalised ratio (INR) target ranges varied from 2.0 to 2.8 for the lower bound and 3.0 to 4.5 for the upper bound). Warfarin was the only anticoagulant used in all of the studies.
ComparisonAspirin (dosage varied from 75mg/day to 325 mg/day) or Indobufen (100mg BID to 200mg BID)
Length of follow-up12 to 42 months; mean not reported.
Outcome measuresFatal OutcomesFatal stroke, fatal vascular event excluding stroke, all-cause mortality
Non-Fatal OutcomesNon-fatal stroke, non-fatal myocardial infarction
Other OutcomesCombined fatal and non-fatal outcomes1, trial-defined primary outcome, major bleeds
Note: 1) * Combined fatal and non-fatal outcomes comprises vascular deaths and all non-fatal vascular events.
Effect sizeFatal outcomesStrokeVascularAll cause
Pooled effects (95% CI)0.74 (0.39 to 1.40)0.86 (0.63 to 1.17)0.94 (0.72 to 1.21)
Non-fatal outcomesStrokeMyocardial infarction
Pooled effects (95% CI)0.68 (0.46 to 0.99)0.83 (0.46 to 1.50)
Other outcomesCombined fatal and non-fatal*Trial defined primary outcomeMajor bleeds
Pooled effects (95% CI)0.79 (0.61 to 1.02)0.73 (0.52 to 1.02)1.45 (0.93 to 2.27)
Notes:
3.

Effect sizes expressed as odds ratios. Lower values favour anticoagulation group.

Source of fundingNorth Thames Research and Development Programme
Additional comments
NCC CC ID (Ref Man)17
Included studiesStudy NameIDPopulationComparison
AFASAK II[396]age >18 years with chronic or paroxysmal AFwarfarin; aspirin (300mg); aspirin and low- dose warfarin;
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
PATAF[1055]age >60 years chronic or paroxysmal AFwarfarin; aspirin(150mg);
SIFA[1206]age >30 years with chronic or paroxysmal AFwarfarin, indobufen (200mg bid)
SPAF II <=75 year[1216]age< 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
SPAF II >75 years[1216]age> 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
Evidence Table: What is the efficacy of anticoagulation versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[20]: 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.
Study typeMeta-Analysis (see Included studies below)
Study ObjectiveComparison of warfarin with placebo/control for the prevention of stroke in atrial fibrillation
Evidence level1+
Number of patientsComparisonTreatmentControl
Adjusted-dose warfarin versus placebo14501450
Adjusted-dose warfarin versus control1454439
Notes: 1) controls included: low/fixed-dose warfarin with/without aspirin; indobufen.
Patient characteristicsRandomized trials testing long-term (>3 months) use of antithrombotic agents in patients with atrial fibrillation.”
Doubleblind and nonblinded trials were included.”
InterventionAdjusted-dose warfarin (achieved target INR range: 2.0 to 2.9)
ComparisonPlacebo, or control (low-/fixed-dose warfarin (achieved target INR range: 1.1 to 1.4) with or without adjunctive aspirin)
Length of follow-upComparisonTreatment1Control1
Adjusted-dose warfarin versus placebo2396 (1.65)2207 (1.52)
Adjusted-dose warfarin versus control938 (2.07)907 (2.07)
Notes: 1) Mean follow-up not reported. Figures represent the total patient years and a calculation of the mean follow-up period in brackets (total patient-years divided by total number of participants)
Outcome measuresStroke frequency
Effect sizeComparisonRRR (95% CI)1ARR2
Adjusted-dose warfarin versus placebo62 (48 to 72)3.1
Adjusted-dose warfarin versus control338 (220 to 68)1.0
Notes:
4.

RRR = Relative Risk Reduction (in favour of warfarin)

5.

ARR = Absolute Risk Reduction (in favour of warfarin)

6.

RRR and ARR refer only to low/fixed-dose warfarin comparison

Source of fundingNational Institute of Neurologic Disorders and Stroke, Bethesda, Maryland.
Additional comments
NCC CC ID (Ref Man)20
Included studiesStudy NameIDPopulationComparison
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
BAATAF[930]chronic or paroxysmal AFwarfarin; placebo
CAFA[127]age >19 years with chronic or paroxysmal AFwarfarin; placebo
EAFT[851]age >25 years with prior stroke or transient ischemic attackwarfarin; aspirin(300mg); placebo
SPAF I - group1[898]adults, chronic or paroxysmal AF,warfarin-eligible warfarin; aspirin(325mg)
SPINAF[1008]chronic AFwarfarin; placebo
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic referenceHart 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.
Study typeMeta-Analysis (see Included studies below)
Study ObjectiveComparison of warfarin with aspirin for the prevention of stroke in atrial fibrillation
Evidence level1+
Number of patientsComparisonTreatmentControl
Adjusted-dose warfarin versus aspirin14161421
Notes: 1) controls included: low/fixed-dose warfarin with/without aspirin; indobufen.
Patient characteristicsRandomized trials testing long-term (>3 months) use of antithrombotic agents in patients with atrial fibrillation.”
Doubleblind and nonblinded trials were included.”
InterventionAdjusted-dose warfarin (target international normalised ratio > 2.0)
ComparisonAspirin (75mg to 325mg/day)
Length of follow-upComparisonTreatment1Control1
Adjusted-dose warfarin versus aspirin3169 (2.24)3103 (2.18)
Notes: 1) Mean follow-up not reported. Figures represent the total patient years and a calculation of the mean follow-up period in brackets (total patient-years divided by total number of participants)
Outcome measuresIschaemic or haemorrhagic stroke frequency
Effect sizeComparisonRRR (95% CI)1ARR2
Adjusted-dose warfarin versus aspirin36 (14 to 52)0.8
Notes:
7.

RRR = Relative Risk Reduction (in favour of warfarin)

8.

ARR = Absolute Risk Reduction (in favour of warfarin)

Source of fundingNational Institute of Neurologic Disorders and Stroke, Bethesda, Maryland.
Additional comments
  • This meta-analysis included studies of both patients with and without a previous stroke or TIA
NCC CC ID (Ref Man)20
Included studiesStudy NameIDPopulationComparison
AFASAK II[396]age >18 years with chronic or paroxysmal AFwarfarin; aspirin (300mg); aspirin and low- dose warfarin;
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
EAFT[851]age >25 years with prior stroke or transient ischemic attackwarfarin; aspirin(300mg); placebo
PATAF[1055]age >60 years chronic or paroxysmal AFwarfarin; aspirin(150mg);
SIFA[1206]age >30 years with chronic or paroxysmal AFwarfarin, indobufen (200mg bid)
MWNAF[1215]age >60 years with chronic AFwarfarin; low-dose warfarin
SPAF II <=75 year[1216]age< 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
SPAF II >75 years[1216]age> 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
SPAF III[880], [1214]patients at high-risk for thromboembolic event, chronic or paroxysmal AFwafarin; aspirin(325mg) and low-dose warfarin
Evidence Table: What is the efficacy of antiplatelet versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[20]: 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.
Study typeMeta-Analysis (see Included studies below)
Study ObjectiveComparison of antiplatelet drugs with placebo for the prevention of stroke (ischaemic and haemorrhagic) in atrial fibrillation
Evidence level1+
Number of patientsComparisonTreatmentControl
Aspirin versus placebo16241601
Dipyridamole versus placebo114107
All antiplatelets versus placebo18421815
Notes: 1) controls included: low/fixed-dose warfarin with/without aspirin; indobufen.
Patient characteristicsRandomized trials testing long-term (>3 months) use of antithrombotic agents in patients with atrial fibrillation.”
Doubleblind and nonblinded trials were included.”
InterventionAntiplatelet (aspirin: 75mg to 325mg/day; dipyridamole 200mg BID)
ComparisonPlacebo
Length of follow-upComparisonTreatment1Control1
Aspirin versus placebo2539 (1.56)2363 (1.48)
Dipyridamole versus placebo127 (1.1)111 (1.04)
All antiplatelets versus placebo2799 (1.52)2585 (1.42)
Notes: 1) Mean follow-up not reported. Figures represent the total patient years and a calculation of the mean follow-up period in brackets (total patient-years divided by total number of participants)
Outcome measuresStroke frequency
Effect sizeComparisonRRR (95% CI) 1ARR2
Aspirin versus placebo22 (2 to 38)1.7
Dipyridamole versus placebo22 (−60 to 62)5.7
All antiplatelets versus placebo24 (7 to 39)1.9
Notes:
9.

RRR = Relative Risk Reduction (in favour of antiplatelet agent)

10.

ARR = Absolute Risk Reduction (in favour of antiplatelet agent)

Source of fundingNational Institute of Neurologic Disorders and Stroke, Bethesda, Maryland.
Additional comments
  • This meta-analysis included studies of both patients with and without a previous stroke or TIA
NCC CC ID (Ref Man)20
Included studiesStudy NameIDPopulationComparison
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
EAFT[851]age >25 years with prior stroke or transient ischemic attackwarfarin; aspirin(300mg); placebo
ESPSII[1226]atrial fibrillation with previous stroke/TIAdipyridamole (200mg bid); dipyridamole with aspirin (25mg); placebo
SPAF I - group1[898]adults, chronic or paroxysmal AF,warfarin-eligible warfarin; aspirin(325mg)
UK-TIA[33]atrial fibrillation with previous stroke/TIAaspirin (300mg, 1200mg); placebo
LASAF[664]cardiology clinic patientsaspirin(125mg); alternate day
aspirin(125mg); placebo
Evidence Table: What is the efficacy of anticoagulation versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[31]: Segal JB, McNamara RL, Miller MR, Powe NR, Goodman SN, Robinson KA, Bass EB. Anticoagulants or antiplatelet therapy for non-rheumatic atrial fibrillation and flutter (Cochrane Review). In: The Cochrane Library, Issue 3, 2004. Chichester, UK: John Wiley & Sons, Ltd.
Study typeSystematic Review with Meta-Analysis (see Included studies below)
Study objectiveComparison of warfarin versus placebo/control in the prevention of stroke in participants with atrial fibrillation
Evidence level1+
Number of patientsComparisonN TreatmentN Control
Warfarin versus Placebo14291425
Adjusted-dose warfarin versus low-dose warfarin153150
Adjusted-dose warfarin versus low-dose warfarin with aspirin693692
Patient characteristicsStudyMean age (years)male, N(%)stroke, N(%)paroxysmal AF, N(%)diabetes, N(%)CHF, N(%)hypertension, N(%)
SPAF I(warf/plac)66303(72)34(8.1)151(36)66(16)70(17)219(52)
SPAF III71629(60)381(37)167(16)191(18)234(32)329(32)
BATAAF69304(72)14(3.3)70(17)63(15)109(24)214(51)
CAFA67282(75)14(3.7)26(6.9)45(12)83(22)146(39)
SPINAF67525(100)0097(18)160(30)305(58)
AFASAK I74540(54)58(5.8)084(8.3)521(52)323(32)
AFASAK II73102(60)14(9)023(13)120(71)72(42)
MWNAF74137(45)0040(13)39(13)182(60)
InterventionInterventionDosage
Warfarin (adjusted dose)INR = 2.0 to 4.0
Low-dose WarfarinINR = 1.2 to 1.6 or 1.25mg/day for fixed minidose
Low-dose Warfarin + AspirinINR = 1.2 to 1.6 or 1.25mg/day for fixed minidose warfarin + 300–325mg/day aspirin
ComparisonSee Number of patients above for details of comparisons.
Length of follow-upNot reported.
Outcome measures01 Warfarin vs Placebo
Outcome titleNo. of studiesNo. of participants
01 Stroke62854
02 Major bleeding62854
03 Total mortality41658
04 Total Mortality: Target INR <=4.031237
05 Total Mortality: Target INR >4.01421
06 Stroke: Target INR <=4.041762
07 Stroke: Target INR >4.021092
08 Major Bleeding: Target INR <=4.041762
09 Major Bleeding: Target INR >4.021092
06 Adjusted-Dose Warfarin vs Low-dose warfarin with Aspirin
Outcome titleNo. of studiesNo. of participants
01 Stroke21385
02 Major bleeding21385
03 Total mortality21385
07 Adjusted -Dose Warfarin vs Low-dose Warfarin
Outcome titleNo. of studiesNo. of participants
01 stroke1303
02 major bleeding1303
03 total mortality1303
Effect size01 Warfarin vs Placebo
Outcome titleEffect size1 (lower values favour warfarin)
01 Stroke0.34 [0.25, 0.46]
02 Major bleeding2.35 [1.30, 4.24]
03 Total mortality0.74 [0.53, 1.04]
04 Total Mortality: Target INR <=4.00.74 [0.51, 1.05]
05 Total Mortality: Target INR >4.00.75 [0.26, 2.17]
06 Stroke: Target INR <=4.00.33 [0.23, 0.48]
07 Stroke: Target INR >4.00.35 [0.19, 0.65]
08 Major Bleeding: Target INR <=4.02.74 [1.42, 5.32]
09 Major Bleeding: Target INR >4.01.26 [0.34, 4.70]
06 Adjusted-Dose Warfarin vs Low-dose warfarin with Aspirin
Outcome titleEffect size1 (lower values favour adjusted-dose warfarin)
01 Stroke0.38 [0.24, 0.60]
02 Major bleeding1.14 [0.55, 2.36]
03 Total mortality1.02 [0.68, 1.52]
07 Adjusted -Dose Warfarin vs Low-dose Warfarin
Outcome titleEffect size1 (lower values favour adjusted-dose warfarin)
01 stroke0.13 [0.02, 0.75]
02 major bleeding3.80 [0.76, 19.10]
03 total mortality0.83 [0.28, 2.53]
Notes: 1) Effect size measured as Peto Odds Ratio [95% CI]
Source of fundingAgency for Healthcare Research and Quality; Rockville, MD; contract 290-97-006 USA
Additional comments
  • Some of the studies included in this meta-analysis did not exclude peri-/post-stroke patients, others included only such patients.
  • Authors did not report reasons for excluding (RCT) studies in all cases. Some relevant studies were not included from the evidence search
NCC CC ID (Ref Man)31
Included studiesStudy NameIDPopulationComparison
AFASAK II[396]age >18 years with chronic or paroxysmal AFwarfarin; aspirin (300mg); aspirin and low-dose warfarin;
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
BAATAF[930]chronic or paroxysmal AFwarfarin; placebo
CAFA[127]age >19 years with chronic or paroxysmal AFwarfarin; placebo
EAFT[851]age >25 years with prior stroke or transient ischemic attackwarfarin; aspirin(300mg); placebo
MWNAF[1215]age >60 years with chronic AFwarfarin; low-dose warfarin
SPAF I - group1[898]adults, chronic or paroxysmal AF,warfarin-eligible warfarin; aspirin(325mg)
SPAF III[880],[1214]patients at high-risk for thromboembolic event, chronic or paroxysmal AFwafarin; aspirin(325mg) and low-dose warfarin
SPINAF[1008]chronic AFwarfarin; placebo
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[31]: Segal JB, McNamara RL, Miller MR, Powe NR, Goodman SN, Robinson KA, Bass EB. Anticoagulants or antiplatelet therapy for non-rheumatic atrial fibrillation and flutter (Cochrane Review). In: The Cochrane Library, Issue 3, 2004. Chichester, UK: John Wiley & Sons, Ltd.
Study typeSystematic Review with Meta-Analysis of 14 studies. Only those (6) studies comparing warfarin with antiplatelet therapies are reported here.
Study objectiveComparison of warfarin versus aspirin and other antiplatelet drugs in the prevention of stroke in participants with atrial fibrillation
Evidence level1+
Number of patientsComparisonN TreatmentN Control
Warfarin versus Antiplatelet (Aspirin or Indobufen)16451653
Warfarin versus Aspirin11911191
Warfarin versus Indobufen454462
Aspirin versus low-dose Warfarin319279
Patient characteristicsStudyMean age (years)male, N(%)stroke, N(%)paroxysmal AF, N(%)diabetes, N(%)CHF, N(%)hypertension, N(%)
SPAF II (<75 years)65539(75)43(6.0)233(33)122(17)125(17)378(53)
SPAF II (> 75 years)80225(58)37(9.6)90(23)50(13)98(25)200(52)
AFASAK I74540(54)58(5.8)084(8.3)521(52)323(32)
AFASAK II73102(60)14(9)023(13)120(71)72(42)
SIFA73434(47)916(100)251(27)160(17)302(33)509(56)
PATAF(warf/asa)70125(46)068(25)46(17)11(4)99(36)
PATAF (low warf/asa)75269(45)092(15)76(13)25(4)243(41)
InterventionInterventionDosage
Warfarin (adjusted dose)INR = 2.0 to 4.0
Low-dose WarfarinINR = 1.2 to 1.6 or 1.25mg/day for fixed minidose
Aspirin75mg/day to 325mg/day. 2 studies has < 300mg/day
Indobufen100–200mg/day
ComparisonSee Number of Patients above for list of comparisons
Length of follow-upNot reported.
Outcome measures03 Warfarin versus Antiplatelet Drugs
Outcome titleNo. of studiesNo. of participants
01 stroke63298
02 major bleeding63298
03 total mortalitiy52627
04 Warfarin vs Aspirin
Outcome titleNo. of studiesNo. of participants
01 Stroke52382
02 Major bleeding52382
03 Total mortality41711
05 Warfarin vs Indobufen
Outcome titleNo. of studiesNo. of participants
01 Stroke1916
02 Major Bleeding1916
03 Total mortality1916
08 Aspirin vs Low-dose Warfarin
Outcome titleNo. of studiesNo. of participants
01 Stroke1598
02 Bleeding1598
03 Mortality1598
Effect size04 Warfarin vs Aspirin
Outcome titleEffect size1 (lower odds ratio favours warfarin)
01 Stroke0.65 [0.44, 0.97]
02 Major bleeding1.56 [0.77, 3.18]
03 Total mortality0.95 [0.60, 1.50]
05 Warfarin vs Indobufen
Outcome titleEffect size1 (lower odds ratio favours warfarin)
01 Stroke0.57 [0.27, 1.20]
02 Major Bleeding3.89 [0.78, 19.36]
03 Total mortality0.93 [0.56, 1.52]
08 Aspirin vs Low-dose Warfarin
Outcome titleEffect size1 (lower odds ratio favours warfarin)
01 Stroke0.99 [0.49, 2.02]
02 Bleeding1.09 [0.29, 4.09]
03 Mortality1.50 [0.99, 2.28]
Notes: 1) Effect size measured as Peto Odds Ratio [95% CI]
Source of fundingAgency for Healthcare Research and Quality; Rockville, MD; contract 290-97-006 USA
Additional comments
  • Some of the studies included in this meta-analysis did not exclude peri-/post-stroke patients, others included only such patients
  • Authors did not report reasons for excluding (RCT) studies in all cases. Some relevant studies were not included from the evidence search
  • Authors did not distinguish between studies that used efficacious doses of aspirin from those that did not
NCC CC ID (Ref Man)31
Included studiesStudy NameIDPopulationComparison
AFASAK II[396]age >18 years with chronic or paroxysmal AFwarfarin; aspirin (300mg); aspirin and low- dose warfarin;
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
PATAF[1055]age >60 years chronic or paroxysmal AFwarfarin; aspirin(150mg);
SIFA[1206]age >30 years with chronic or paroxysmal AFwarfarin, indobufen (200mg bid)
SPAF II <=75 year[1216]age< 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
SPAF II >75 years[1216]age> 75 years, chronic or paroxysmal AFwarfarin; aspirin(325mg)
Evidence Table: What is the efficacy of antiplatelet therapy versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[31]: Segal JB, McNamara RL, Miller MR, Powe NR, Goodman SN, Robinson KA, Bass EB. Anticoagulants or antiplatelet therapy for non-rheumatic atrial fibrillation and flutter (Cochrane Review). In: The Cochrane Library, Issue 3, 2004. Chichester, UK: John Wiley & Sons, Ltd.
Study typeSystematic Review with Meta-Analysis of 14 studies. Only those (4) studies comparing aspirin with placebo are reported here.
Study ObjectiveComparison of aspirin versus placebo in the prevention of stroke in participants with atrial fibrillation
Evidence level1+
Number of patientsComparisonN TreatmentN Control
Aspirin versus Placebo13961373
Patient characteristicsStudyMean age (years)male, N(%)stroke, N(%)paroxysmal AF, N(%)diabetes, N(%)CHF, N(%)hypertension, N(%)
SPAF I(asp/plac)67790(71)73(6.5)376(34)196(18)224(20)588(52)
AFASAK I74540(54)58(5.8)084(8.3)521(52)323(32)
EAFT(asp/plac)73438(56)782(100)192(25)101(13)89(11)376(48)
LASAF66151(51)055(19)21(7)13(4)146(50)
InterventionAspirin (75mg to 300mg per day, depending on study – see below)
ComparisonPlacebo
Length of follow-upNot reported
Outcome measures02 Aspirin vs Placebo
Outcome titleNo. of studiesNo. of participants
01 Stroke52769
02 Major Bleeding32574
03 Total mortality32097
Effect size02 Aspirin vs Placebo
Outcome titleEffect size1 (lower values favour aspirin)
01 Stroke0.80 [0.62, 1.03]
02 Major Bleeding0.94 [0.48, 1.84]
03 Total mortality0.87 [0.68, 1.12]
Notes: 1) Effect size measured as Peto Odds Ratio [95% CI]
Source of fundingAgency for Healthcare Research and Quality; Rockville, MD; contract 290-97-006 USA
Additional comments
  • Some of the studies included in this meta-analysis did not exclude peri-/post-stroke patients, others included only such patients.
  • Authors did not report reasons for excluding (RCT) studies in all cases. Some relevant studies were not included from the evidence search
NCC CC ID (Ref Man)31
Included studiesStudy NameIDPopulationComparison
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
EAFT[851]age >25 years with prior stroke or transient ischemic attackwarfarin; aspirin(300mg); placebo
LASAF[664]cardiology clinic patientsaspirin(125mg); alternate day
aspirin(125mg); placebo
SPAF I - group2[898]adults, chronic or paroxysmal AF,warfarin-, ineligible aspirin(325mg)
Evidence Table: What is the efficacy of antiplatelet therapy versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[33]: Benavente O. et al, Antiplatelet therapy for preventing stroke in patients with non-valvular atrial fibrillation and no previous history of stroke or transient ischemic attacks. The Cochrane Library. 2004
Study typeSystematic Review with Meta-Analysis of two RCTs (AFASK I and SPAF I)
Study objectiveComparison of Antiplatelet Therapy with Placebo in the prevention of thromboembolic complications of chronic non-valvular atrial fibrillation
Evidence level1+
Number of patientsTreatment GroupN
Aspirin838
Placebo842
Total:1680
Patient characteristicsInclusion Criteria:
  • Diagnosis: Diagnosed paroxysmal or persistent AF. Patients with mitral stenosis or prosthetic cardiac valves were excluded.“
InterventionAspirin (325 mg/d in SPAF I, 75 mg/d in AFASAK I)
ComparisonPlacebo (double-masked)
Length of follow-upMean follow-up of 1.3 years/participant
Outcome measuresOutcome titleNo. of studiesNo. of participants
01 All ischemic stroke (fatal and non-fatal)21680
02 All ischemic stroke or intracranial haemorrhage21680
03 All disabling or fatal ischemic stroke or intracranial haemorrhage21680
04 Myocardial infarction11046
05 All systemic emboli21680
06 All intracranial haemorrhage21680
07 All vascular death21680
08 All major extracranial bleeds21680
09 Composite: stroke (ischemic or hemorrhagic), MI or vascular death21680
Effect sizeOutcome titleEffect size1
01 All ischemic stroke (fatal and non-fatal)0.71 [0.46, 1.10]
02 All ischemic stroke or intracranial haemorrhage0.70 [0.45, 1.08]
03 All disabling or fatal ischemic stroke or intracranial haemorrhage0.88 [0.48, 1.58]
04 Myocardial infarction0.61 [0.23, 1.64]
05 All systemic emboli0.67 [0.19, 2.33]
06 All intracranial haemorrhage1.02 [0.14, 7.23]
07 All vascular death0.88 [0.56, 1.38]
08 All major extracranial bleeds1.14 [0.44, 2.98]
09 Composite: stroke (ischemic or haemorrhagic), MI or vascular death0.76 [0.54, 1.05]
Notes: 1) Effect size measured as Peto Odds Ratio [95% CI]. Lower odds favours Aspirin
Source of fundingNot reported
Additional comments
  • The study combined clinically efficacious with non-efficacious doses of aspirin into a single meta-analysis
NCC CC ID (Ref Man)33
Included studiesStudy NameIDPopulationComparison
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
SPAF I - group2[898]adults, chronic or paroxysmal AF,warfarin-, ineligible aspirin(325mg)
Evidence Table: What is the efficacy of anticoagulation versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[35]: Benavente O, Hart R, Koudstaal P, Laupacis A, McBride R Oral anticoagulants for preventing stroke in patients with non-valvular atrial fibrillation and no previous history of stroke or transient ischemic attacks (Cochrane Review). In: The Cochrane Library, Issue 3, 2004. Chichester, UK: John Wiley & Sons, Ltd.
Study typeSystematic Review with Meta-Analysis of 5 RCTs (see Included Studies below)
Study Objective“The objective of this review was to characterize the efficacy and safety of oral anticoagulation (OAC) with vitamin K antagonists for the primary prevention of stroke in patients with chronic AF.”
Evidence level1+
Number of patientsTreatment GroupN
OAC1154
Control1159
Total:2313
Patient characteristicsInclusion Criteria:
  • “Participants with AF documented by EKG, either intermittent (i.e. paroxysmal) or sustained (i.e. constant) were included. Those with mitral stenosis or prosthetic cardiac valves were not included.”
InterventionAdjusted-dose warfarin (INR range not reported)
ComparisonAspirin (various doses) or no treatment or placebo
Length of follow-upMean follow-up varied from 1.2–2.3 years between trials, with a mean of 1.5 years/participant overall
Outcome measuresOutcome titleNo. of studiesNo. of participants
01 All ischemic stroke (fatal and non-fatal)52313
02 All ischemic stroke or intracranial haemorrhage52313
03 All disabling or fatal ischemic stroke or intracranial haemorrhage52313
04 Myocardial infarction31318
05 All systemic emboli52313
06 All intracranial haemorrhage52313
07 All major extracranial bleeds52313
08 All vascular death52313
09 Composite outcome: stroke (ischemic or hemorrhagic), MI or vascular death52313
Effect sizeOutcome titleEffect size1
01 All ischemic stroke (fatal and non-fatal)0.34 [0.23, 0.52]
02 All ischemic stroke or intracranial haemorrhage0.39 [0.26, 0.59]
03 All disabling or fatal ischemic stroke or intracranial haemorrhage0.47 [0.28, 0.80]
04 Myocardial infarction0.87 [0.32, 2.42]
05 All systemic emboli0.45 [0.13, 1.57]
06 All intracranial haemorrhage2.38 [0.54, 10.50]
07 All major extracranial bleeds1.07 [0.53, 2.12]
08 All vascular death0.84 [0.56, 1.27]
09 Composite outcome: stroke (ischemic or hemorrhagic), MI or vascular death0.57 [0.42, 0.77]
Notes: 1) Effect size measured as Peto Odds Ratio [95% CI]
Source of fundingNot reported
Additional comments
NCC CC ID (Ref Man)35
Included studiesStudy NameIDPopulationComparison
AFASAK[141]age >18 years with chronic AFwarfarin; aspirin (75mg); placebo
BAATAF[930]chronic or paroxysmal AFwarfarin; placebo
CAFA[127]age >19 years with chronic or paroxysmal AFwarfarin; placebo
SPAF I - group1[898]adults, chronic or paroxysmal AF,warfarin-eligible warfarin; aspirin(325mg)
SPINAF[1008]chronic AFwarfarin; placebo
Evidence Table: What is the efficacy of anticoagulation versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[418]: Perret-Guillaume C and Wahl DG., Low-dose warfarin in atrial fibrillation leads to more thromboembolic events without reducing major bleeding when compared to adjusted-dose. Thrombosis & Haemostasis: . 91: 394 – 402. 2004
Study typeMeta-analysis of 4 RCTs (see Included Studies below)
Study ObjectiveComparison of rates of thromboembolic and haemorrhagic events between low-dose and adjusted-dose warfarin
Evidence level1+
Number of patientsTreatment GroupN
Adjusted-dose warfarin977
Low-dose warfarin (with and without aspirin)1131
Total:2108
Patient characteristicsDiagnosis: Non-rheumatic atrial fibrillation (with or without prior stroke or transient ischaemic attack)
InterventionAdjusted-dose warfarin (INR > 2.0)
ComparisonLow/minidose-warfarin (authors combined low- and minidose-warfarin groups based on a narrow INR range of 1.12 to 1.4) or low/minidose-warfarin with aspirin
Length of follow-upNot reported
Outcome measuresThrombotic eventsIschaemic stroke, systemic embolism, vascular death, all thrombotic events
Haemorrhagic eventsMajor haemorrhage, haemorrhagic death,
Effect sizeAdjusted-Dose Warfarin versus Low-Dose Warfarin:
Thrombotic OutcomeRR1 (95% CI)
Ischaemic stroke0.46 (0.2 to 1.07)
Systemic embolism1.18 (0.33 to 4.21)
All thrombotic events20.50 (0.25 to 0.97)
Vascular death1.1 (0.72 to 1.67)
Haemorrhagic OutcomeRR1 (95% CI)
Major haemorrhage1.23 (0.67 to 2.27)
Haemorrhagic death0.97 (0.27 to 3.54)
Adjusted-Dose Warfarin versus Low-Dose Warfarin with Aspirin3:
OutcomeRR1 (95% CI)
Ischaemic stroke0.67 (0.33 to 1.36)
All thrombotic events20.63 (0.38 to 1.04)
Vascular death1.36 (0.64 to 2.89)
Major haemorrhage1.62 (0.58 to 4.54)
Notes:
11.

RR = Relative Risk

12.

All thrombotic events excludes vascular death

13.

Comparison with low-dose warfarin with aspirin based on results of 3 trials

Source of fundingNot reported
Additional comments
  • Study selection includes studies not included in other meta-analyses. Coverage is more comprehensive than in Lip 2004 ([873])
NCC CC ID (Ref Man)418
Included studiesStudy NameIDPopulationComparison
AFASAK II[396]age >18 years with chronic or paroxysmal AFwarfarin; aspirin (300mg); aspirin and low-dose warfarin;
PATAF[1055]age >60 years chronic or paroxysmal AFwarfarin; aspirin(150mg);
MWNAF[1215]age >60 years with chronic AFwarfarin; low-dose warfarin
SPAF III[880],[1214]patients at high-risk for thromboembolic event, chronic or paroxysmal AFwafarin; aspirin(325mg) and low-dose warfarin
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[458]: Hylek E et al, Effect of intensity of oral anticoagulation on stroke severity and mortality in atrial fibrillation. New England Journal of Medicine: . 349: 1019 – 1026. 2003
Study typeRetrospective cohort (using ATRIA patient population)
Study objectiveComparison of stroke severity against various antithrombotic therapies and dosage levels.
Evidence level2+
Number of patientsTreatment GroupN
No Treatment248
Aspirin160
Warfarin188
Total:596
Patient characteristicsInclusion Criteria:
No TreatmentAspirinWarfarin
Mean Age, years798075
% Male453952
% Heart Failure353539
% Hypertension705870
% Prior Ischaemic Stroke1222740
Notes:
2.

The rates of prior ischaemic stroke appear to be significantly different for Warfarin group.

InterventionVarious levels of anticoagulation (median recorded INR was 1.7)
ComparisonAspirin or no treatment
Length of follow-up18 months
Outcome measures
  1. Risk from suffering a severe (fatal or severely disabling) ischaemic stroke.
Effect sizeIntervention1Odds Ratio (95% CI)P
No treatment2.2 (1.3 to 3.8)0.004
Aspirin1.3 (0.7 to 2.3)NS
Warfarin (INR < 2.0)1.9 (1.1 to 3.4)0.03
Warfarin (INR > 2.0)1.02
Notes:
14.

Intervention was determined upon presentation with ischaemic stroke

15.

Warfarin with INR > 2.0 was used as the reference category

Source of funding
  • National Institute of Aging
  • Bristol-Myers Squibb
Additional commentsAlthough the apparently significant difference between the cohorts in terms of prior stroke would normally be a source of unacceptable bias for this kind of study, it has been included since this bias would tend to be in the opposite direction to the results.
NCC CC ID (Ref Man)458
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[1005]: Edvardsson N et al, Effects of low-dose warfarin and aspirin versus no treatment on stroke in a medium-risk patient population with atrial fibrillation. Journal of Internal Medicine: 254: 95 – 101. 2003
Study typeRCT (open label)
Study objectiveComparison of low-dose warfarin in combination with aspirin in comparison with no anticoagulation
Evidence level1+
Number of patientsTreatment GroupN
Warfarin+Aspirin334
No Anticoagulation334
Total:668
Patient characteristicsInclusion Criteria:
Warfarin+AspirinNo Anticoagulation
Mean Age (SD), years72 (7)73 (7)
Age Range, years53 to 9058 to 90
% Male6461
InterventionWarfarin 1.25mg/day + aspirin 75mg/day
ComparisonNo anticoagulation
Length of follow-up33 months (mean)
Outcome measuresPrimary End PointsStroke
Secondary End PointsTransient ischaemic attack (TIA), all-cause mortality, cardiovascular mortality, acute myocardial infarction, peripheral embolism, stroke or TIA, bleeding event
Notes:
2.

Only those end points that are significantly associated with the intervention or are of particular importance will be reported here

Effect sizeWarfarin+AspirinNo AnticoagulationP
% Stroke (ischaemic or haemorrhagic)9.612.3NS
% Stroke or TIA11.716.50.09
% All cause mortality9.310.8NS
% Bleeding event5.71.20.003
Source of fundingBristol Myers Squibb
Additional comments
NCC CC ID (Ref Man)1005
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[1047]: Hart RG et al, Cardioembolic vs. noncardioembolic strokes in atrial fibrillation: frequency and effect of antithrombotic agents in the stroke prevention in atrial fibrillation studies. Cerebrovascular Diseases: 10: 39 – 43. 2000
Study typeRCT with Cohort using SPAF 3 data and Meta-Analysis pooling results of SPAF 1, 2 and 3
Study ObjectiveComparison of rates of cardioembolic versus non-cardioembolic strokes during warfarin and aspirin prophylaxis
Evidence level2++
Number of patientsSPAF 3 Treatment GroupN
Low-Risk Aspirin Cohort892
High-Risk Aspirin+Warfarin521
High-Risk Warfarin523
Total:1936
Total number of patients included in meta-analysis (pooled participants from SPAF 1–3): 3,950. Note that previous SPAF trials also included placebo/no treatment groups
Patient characteristicsAF patients with either a high-risk of stroke (randomised to either combination therapy or warfarin) and AF patients with a lower risk of stroke assigned to an aspirin cohort.
InterventionAdjusted-dose warfarin or low-dose warfarin with aspirin
ComparisonAspirin
Length of follow-upNot reported
Outcome measuresCardioembolic strokeRate/frequency of cardioembolic stroke
Non-cardioembolic strokeRate/frequency of non-cardioembolic stroke
Undetermined strokeRate/frequency of stroke with undetermined aetiology
Notes:
3.

Stroke classification was based on published classification system

4.

Only ischaemic strokes are considered

Effect sizeTherapy1CardioembolicNon-cardioembolicUndetermined
Placebo/no treatment32198
Aspirin+warfarin2702038
Warfarin37144
Uncertain/other401
Total (%)113 (52%)53 (24%)51 (24%)
Summary:
  • Overall level of significance of the interaction between stroke aetiology and treatment: p < 0.001, SPAF 1 to 3 data
  • Adjusted-dose warfarin resulted in a disproportionate reduction in the rate of cardioemolic strokes in comparison to placebo/no treatment (p<0.02, SPAF 1 to 3 data)
  • Aspirin therapy (including warfarin+aspirin treatment group) resulted in a trend towards a disproportionate reduction in the rate of non-cardioembolic strokes in comparison to other treatment groups (p=0.06, SPAF 1 to 3 data)
  • Warfarin therapy had a dramatic reduction on the number of cardioembolic strokes in comparison to aspirin/aspirin+warfarin therapy (83% less in warfarin group, p<0.001), but there was no significant reduction in the number of non-cardioembolic strokes (SPAF 2 and 3 data)
Notes:
16.

Therapy groupings are stated by the authors as varying greatly, and hence rates are not displayed. Significant comparisons are based on patterns of frequency

17.

Aspirin+warfarin group received ineffective doses of warfarin

18.

Adjusted dose warfarin to INR 2.0 to 3.0

Source of fundingDivision of Stroke and Trauma, Institute of Neurological Disorders and Stroke, Bethesda, Md., USA
Additional comments
NCC CC ID (Ref Man)1047
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[1093]: Lechat P et al and FFAACS (Fluindione FAAeCSI., Anticoagulant (fluindione)-aspirin combination in patients with high-risk atrial fibrillation. A randomized trial (Fluindione, Fibrillation Auriculaire, Aspirin et Contraste Spontane; FFAACS). Cerebrovascular Diseases: 12: 245 – 252. 2001
Study typeDouble-blind RCT (multi-centre)
Study objectiveComparison of a combination of anticoagulation and antiplatelet treatment regimen with anticoagulation alone
Evidence level1++
Number of patientsTreatment GroupN
Anticoagulation + Antiplatelet (A+A)76
Anticoagulation + Placebo (A+P)81
Total:157
Patient characteristicsInclusion Criteria:
  • Diagnosis: Permanent or paroxysmal atrial fibrillation within the previous 3 months, and at least one of the following risk factors:
  • Patients with contraindications, recent cardiac surgery, NYHA grade IV or thyroid dysfunction were excluded.
A+AA+P
Mean Age (SD), years73 (6)74 (7)
% Male4456
% Permanent AF5658
% Thromboembolic history6565
InterventionA+A: 100mg aspirin/day + fluindione administered to a target international normalised ratio of 2.0 to 2.6
ComparisonA+P: Placebo + fluindione administered to a target international normalised ratio of 2.0 to 2.6
Length of follow-up0.84 years
Outcome measuresPrimary EndpointThromboembolic event or vascular death
Thromboembolic Event
MortalityAll-cause mortality, vascular death, haemorrhagic death
Bleeding EventMajor bleeding event, minor bleeding event
Effect sizeA+AA+PP
Primary Endpoint (%)7.92.9NS
Thromboembolic Event (%)3.21.4NS
Mortality (%)4.74.3NS
Bleeding Event – Minor (%)15.81.4<0.05
Bleeding Event – Major (%)4.81.4NS
Bleeding Event – Total (%)20.62.9<0.05
Source of funding
  • Assistance Publique – Hospitaux de Paris
  • Bayer Company, Procter and Gamble Pharmaceuticals
  • French Drug Agency
  • French Federation of Cardiology
  • French Ministry of Health
Additional commentsStudy was stopped prematurely, resulting in low recruitment and follow up and underpowered
NCC CC ID (Ref Man)1093
Evidence Table: What is the efficacy of anticoagulation versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[1213]: Executive Steering Committee on behalf of the SPORTIF III Investigators, Stroke prevention with the oral direct thrombin inhibitor ximelagatran compared with warfarin in patients with non-valvular atrial fibrillation (SPORTIF III): Lancet 2003; 362: 1691–98
Study typeRCT
Study objectiveComparison of Ximelagatran with warfarin for safety and efficacy in the prevention of stroke in AF patients
Evidence level1+
Number of patientsTreatment GroupN
Ximelagatran1704
Warfarin1703
Total:3407
Patient characteristicsInclusion Criteria:
  1. Age 18 years or older
  2. Persistent or paroxysmal non-valvular atrial fibrillation verified by at least two electrocardiogram recordings, one of which was made within 2 weeks before randomisation
  3. One or more of the following risk factors for stroke:
    1. Hypertension (raised blood pressure needing antihypertensive treatment but <180/100 mm Hg)
    2. Age 75 years or older
    3. Previous stroke, transient ischaemic attack, or systemic embolism
    4. Left ventricular dysfunction (left-ventricular ejection fraction <40% or symptomatic congestive heart failure)
    5. Age 65 years or older and coronary artery disease
    6. Age 65 years or older and diabetes mellitus
XimelagatranWarfarin
Men1158 (68%)1196 (70%)
Age (years, mean [SD])70.3 (8.6)70.1 (8.6)
Aspirin at entry345 (20%)359 (21%)
Systolic blood pressure (mm Hg, mean [SD])139 (18)139 (18)
Atrial fibrillation onset less than 1 year368 (22%)347 (20%)
Paroxysmal atrial fibrillation160 (9%)124 (7%)
Previous stroke, TIA, or both417 (24%)405 (24%)
Previous non-CNS embolism74 (4%)77 (5%)
Age 75 years or older581 (34%)565 (33%)
Hypertension1229 (72%)1230 (72%)
Left ventricular dysfunction574 (34%)584 (34%)
Notes:
3.

Data are number of patients (%) unless otherwise stated. TIA=transient ischaemic attack. *Includes five patients in each group incorrectly randomised without risk factors.

InterventionXimelagatran (fixed-dose, 36 mg twice daily)
ComparisonOpen-label warfarin (adjusted-dose, international normalised ratio [INR] 2·0–3·0)
Length of follow-upMean 17·4 months
Outcome measuresPrimary events (ITT)
Composite Outcome 1Primary event or death (ITT)
Composite Outcome 2Composite of mortality, stroke, systemic embolism, and myocardial infarction (OT)
Composite Outcome 3Composite of ischaemic stroke, systemic embolism, and transient ischaemic attack (OT)
Bleeding Events
  • Major bleeding (OT)
  • Major or minor bleeding (OT)
Notes:
5.

ITT = Outcome is reported on an intention-to-treat basis

6.

OT = Outcome is reported on an on-treatment basis

Effect sizeXimelagatranWarfarinARR1 (95% CI)
Primary events (ITT)40 (1.6%)56 (2.3%)−0.7% (−1.4 to 0.1)
Composite Outcome 1103 (4.2%)124 (5.1%)−0.9% (−2.1 to 0.3)
Composite Outcome 296 (4.2%)116 (4.9%)−0.7% (−1.9 to 0.5)
Composite Outcome 348 (2.1%)67 (2.9%)−0.8% (−1.7 to 0.2)
Major bleeding (OT)29 (1.3%)41 (1.8%)−0.5% (−1.2 to 0.2)
Major or minor bleeding (OT)478 (25.8%)547 (29.8%)−4.0% (−6.9 to −1.1)
Notes:
19.

ARR = absolute risk reduction (in favour of Ximelagatran)

Source of fundingAstraZeneca, Mölndal, Sweden
Additional comments
NCC CC ID (Ref Man)1213
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[1225]: Singer DE, Albers GW, Dalen JE, Go AS, Halperin JL, Manning WJ. Antithrombotic Therapy in Atrial Fibrillation: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126:429S–56S.
Study typeSystematic Review without Meta-analysis
Study objectiveReview of the evidence relating to the effectiveness and optimal dosage of anticoagulation therapy
Evidence level4
Number of patientsNot applicable
Patient characteristicsInclusion Criteria:
  • RCTs and observational studies involving the use of anticoagulation therapies and/or antiplatelet therapies for prevention of thromboembolic events in patients with atrial fibrillation
InterventionOral anticoagulation at conventional doses (INR > 2.0)
ComparisonPlacebo, no treatment or low-dose anticoagulation
Length of follow-upNot applicable
Outcome measuresPrimary outcome thromboembolic events and rates of major bleeding reported in primary studies
Effect sizePRIMARY OUTCOME EVENT RATES:
OAC vs ControlOACControlRRR, %¶Reported p Values
AFASAK†2.76.256<0.05
SPAF I2.37.4670.01
BAATAF0.43860.002
CAFA3.44.6260.25
SPINAF0.94.3790.001
EAFT8.516.5470.001
Aspirin vs controlAspirinControlRRR, %¶Reported p Values
AFASAK†5.26.216NS
SPAF I3.66.3420.02
EAFT15.519170.12
ESPS‡13.820.7330.16
LASAF
 125 mg qd2.62.2−15NS
 125 mg every other day0.72.2680.05
OAC vs AspirinOACAspirinRRR, %¶Reported p Values
AFASAK†2.75.248<0.05
SPAF I
SPAF II
 <751.31.9330.24
 >753.64.8270.39
EAFTNANA400.008
AFASAK3.42.7−21NS
PATAF2.53.119NS
OAC vs Low-dose OAC plus aspirinOACOAC + aspirinRRR, %¶Reported p Values
SPAF III1.97.974< 0.0001
AFASAK3.43.2−6NS
NASPEAF (triflusal, not aspirin)
 Higher risk4.62.3−500.03
 Lower risk2.50.92−630.04
OAC vs low-dose OACOACLow-dose OACRRR, %¶Reported p Values
AFASAK3.43.913NS
PATAF2.52.2−12NS
Pengo et al3.66.2420.29
Japanese study1.11.735NS
OAC vs IndobufenOACIndobufenRRR, %¶Reported p Values
SIFA910.615NS
OAC vs OAC plus aspirinOACOAC plus aspirinRRR, %¶Reported p Values
FFAACS2.97.9630.21
OAC vs ximelagatranOACXimelagatranRRR, %¶Reported p Values
SPORTIF III2.31.6−30||
SPORTIF V1.21.625||
Notes:
*NS = not significant
†Based on intention-to-treat analysis.
‡ESPS 2 had two additional treatment arms: dipyridamole, 400 mg qd (annual stroke rate, 15.1%), and dipyridamole, 400 mg qd, plus aspirin, 50 mg qd (annual stroke rate, 11.0%).
§NASPEAF lower-risk group treated with triflusal, 600, mg/d alone, had an annual rate of primary outcome events of 3.8 per 100.
RRR favours the OAC intervention.
|| Noninferiority criterion met; standard p values not applicable.
BLEEDING EVENT RATES:
ComparisonMajor Heamorrhage (Annualised Rate), all sitesIntracerebral Heamorrhage (Annualised Rate), all sites
OAC vs controlOACControlOACControl
AFASAK0.600.30
SPAF I1.51.60.80.8
BAATAF0.40.20.20
CAFA2.10.40.40
SPINAF1.30.900
EAFT2.60.700.2
Aspirin vs controlAspirinControlAspirinControl
AFASAK0.3000
SPAF I1.41.90.30.3
EAFT0.70.60.20.1
ESPS2‡0.90.4NANA
LASAF§
 125 mg qdNANANANA
 125 mg every other dayNANANANA
OAC vs aspirinOACAspirinOACAspirin
AFASAK0.60.30.30
SPAF INANANANA
SPAF II
<751.70.90.50.2
>754.21.61.80.8
AFASAK1.71.60.60.3
PATAF0.20.30.20.3
OAC vs aspirin plus low-dose OACOACAspirin plus OACOACAspirin plus OAC
SPAF III2.12.40.50.9
AFASAK1.70.30.60
NASPEAF (triflusal, not aspirin)||
 Higher risk2.12.10.80.3
 Lower risk1.80.90.70.2
OAC vs low-dose OACOACLow-dose OACOACLow-dose OAC
AFASAK1.10.80.60.3
PATAF0.20.30.20.3
Pengo et al2.610.50
Japanese study6.601.10
OAC vs indobufenOACIndobufenOACIndobufen
SIFA0.9000
OAC vs OAC plus aspirinOACOAC plus aspirinOACOAC plus aspirin
FFAACS1.44.8NANA
OAC vs ximelagatranOACXimelagatranOACXimelagatran
SPORTIF III1.81.30.40.2
SPORTIF V
Notes:
†BAATAF criteria for serious bleeding were different from those in other trials
‡ESPS 2 also included two other treatment groups: (1) modified-release dipyridamole, 200 mg bid; (2) aspirin, 25 mg bid, plus modified-release
dipyridamole, 200 mg bid.
§One fatal hemorrhagic stroke in aspirin, 125 mg qd, group but nonfatal ICH and major non-CNS bleeds not reported.
||NASPEAF lower-risk group treated with triflusal 600, mg/d, alone experienced annual rates of 0.35 per 100 for all severe bleeds and for ICH.
¶Specific rates not given in abstract, but text states that there was no significant difference in major bleeding or in hemorrhagic stroke.

EVIDENCE STATEMENTS:
  • A case-control study based in a large anticoagulation unit found that the risk of stroke increased at INR levels <2.0. For example, the odds of stroke doubled at an INR of 1.7 and tripled at an INR of 1.5 compared to an INR of 2.0, and increased even more dramatically if the INR was <1.5. A second hospital-based case-control study also found a sharp increase in risk of stroke among patients with AF and INR values <2.0. Post hoc analyses of the SPAF III trial were consistent with these epidemiologic analyses.
  • In two studies, the risk of ICH was fairly low at INR values <4.0 but was sharply higher at greater INR levels.
  • A recent report from a large cohort study indicates that INR levels <2.0 not only increase the risk of stroke but also markedly raise the risk of severe or fatal stroke should such an event occur. Since randomized trials have successfully used INR targets of 2.0 to 3.0, this target range seems an appropriate standard. There is currently no direct evidence indicating that this range should be changed for the very elderly (patients >75 years old), who have higher risks than younger patients of both stroke and bleeding on oral anticoagulants.
Source of fundingNot reported
Additional commentsTrials listed included both paroxysmal and persistent atrial fibrillation participants, as well as permanent atrial fibrillation participants. All effect sizes and evidence statements are quoted directly from the study text; references to studies from within the text have been removed.
NCC CC ID (Ref Man)1225
Evidence Table: What is the efficacy of anticoagulation versus placebo for stroke prevention in patients with permanent AF?
Bibliographic reference[1232]: 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. Seminars in Thrombosis & Hemostasis 1993;19:116–21.
Study typeRCT
Study objectiveComparison of low-molecular-weight heparin (LMWH) as an effective antithrombotic agent for long-term stroke prevention
Evidence level1+
Number of patientsTreatment GroupN
LMWH35
Control40
Total:75
Patient characteristicsInclusion Criteria:
  • Patients with chronic non-rheumatic atrial fibrillation
  • Patients with acute stroke were excluded.
LMWHControlP
Mean Age (SD), years79 (7)84 (7)NS1
% Male2028NS
% Previous Embolism4317.5<0.05
Notes:
4.

NS = not significant

Intervention36mg LMWH self-administered subcutaneously daily with pre-loaded disposable syringes
ComparisonNo treatment
Length of follow-up6 months
Outcome measures
Effect sizeLMWHControlP
Embolism Rate (%) – All patients8.620NS
Embolism Rate (%) – Patients with no previous embolic event1015NS
Mortality Rate (%) – All patients1728NS
Notes:
  1. Calculation of statistical significance was not performed in the study. All measures of statistical significance are therefore estimates based on the presented published date.
  2. There were no adverse events in the LMWH group
Source of fundingNot reported
Additional commentsThe significant difference in rate of previous embolism between the two groups suggests that the patients were not, in fact, randomised despite a statement to the contrary. Nonetheless, the bias would be assumed to be in the opposite direction of the actual results.
Results for the subgroup of patients with a previous embolism are reported separately.
Patients in this study were older than in other anticoagulation studies.
NCC CC ID (Ref Man)1232
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[1868]: Perez-Gomez F, Alegria E, Berjon J, Iriarte JA, Zumalde J, Salvador A et al. Comparative effects of antiplatelet, anticoagulant, or combined therapy in patients with valvular and nonvalvular atrial fibrillation A randomized multicenter study. J Am Coll Cardiol 2004;44:1557–66.
Study typeRCT
Study objectiveComparison of effectiveness of Triflusal, Acenocoumarol and their combination in the prevention of stroke in a moderate-risk atrial fibrillation population
Evidence level1+
Number of patientsTreatment GroupN
Triflusal235
Acenocoumarol232
Combination222
Total:689
Patient characteristicsInclusion Criteria:
TriflusalAnticoagulationCombination
Age, yrs (SD)69.9 (8)69.6 (7)69.8 (7)
Men5754.657.4
NYHA functional class II to IV595044
LV ejection fraction (SD)61.0 (11)60.0 (12)60.0 (10)
LA diameter, mm (SD)46.2 (8)47.4 (8)47.0 (8)
Persistent/permanent atrial fibrillation89.989.693.3
InterventionTriflusal: 600mg trifusal daily OR;
Anticoagulation: Anticoagulation (acenocoumarol) administered to a target INR range of 2 to 3.
ComparisonCombination: 600mg trifusal daily AND anticoagulation (acenocoumarol) administered to a target INR range of 1.25 to 2.00.
Length of follow-up2.4 years (approximated mean)
Outcome measuresPrimary Outcome
  • Vascular death, TIA, non-fatal stroke (ischaemic or haemorrhagic) or systemic embolism
Bleeding
Composite Outcome
  • Primary outcome or severe bleeding1
Mortality
Cerebral Embolism
Notes:
7.

Haemorrhagic stroke was excluded from Composite Outcome to avoid double counting.

Effect sizeTriflusalAnticoagulationCombinationP
Primary outcome22 (3.82)15 (2.70)5 (0.92)<0.05
Vascular death8 (1.39)11 (1.98)2 (0.37)<0.01
Ischaemic stroke and TIA15 (2.60)7 (1.26)3 (0.55)NR
Severe bleeding2 (0.35)10 (1.80)5 (0.92)NR
Primary Outcome + Severe bleeding22 (3.82)21 (3.78)8 (1.48)<0.05
Nonvascular death794NR
Nonsevere bleeding51516NR
Notes:
20.

Figures in brackets refers to incidence per year

21.

Statistical hypothesis test results were not reported in all cases in the study.

Source of fundingSpanish Society of Cardiology, Uriach Foundation
Additional comments
NCC CC ID (Ref Man)1868
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[2]: Man-Son-Hing, M and Laupacis, A. Anitcoagulant-Related Bleeding in Older Persons With Atrial Fibrillation. Arch Intern Med 163. 2003
Study typeSystematic review of meta-analyses and primary (RCT and risk-assessment) studies.
Study Objective“To critically appraise whether the presence of additional clinical factors that increase the risk of bleeding affects the chance of anticoagulant-related hemorrhage, and to develop an approach to the use of anticoagulant agents in older patients with atrial fibrillation who have any of these factors”
Evidence level2-
Number of patientsNot reported
Patient characteristicsInclusion Criteria:
  • Comparisons: Studies comparing the risk of bleeding events for different antithrombotic therapies (warfarin, aspirin or placebo) for the prevention of stroke
  • Diagnosis: Atrial fibrillation with identified risk factor for gastrointestinal (GI) bleeding
Risk Stratification Categories:
  • Age: 65–75 years; ≥75 years
  • Stroke Risk: No risk factors; ≥1 risk factor (RF)
  • GI Risk: GI Risk 1: “Recent resolved GI tract bleeding (with Helicobacter pylori testing and treatment)”; GI Risk 2: “Concurrent NSAID1 and misoprostol or PPI2 use or COX-23-specific NSAID use”; GI Risk 3: “Concurrent conventional NSAID use”
Notes:
  1. NSAID = Non-steroidal anti-inflammatory drug;
  2. PPI = proton pump inhibitor;
  3. COX-2 = cyclooxygenase-2
InterventionWarfarin or aspirin
ComparisonAspirin or placebo/no treatment
Length of follow-upNot reported.
Outcome measuresRisk of Intracranial Bleeding
Risk of Gastrointestinal Bleeding
Risk of Stroke
Quality-adjusted Life Years (QALYs)
Effect sizeGI RiskAgeStroke RiskWarfarin1Aspirin1No Treatment1
GI Risk 1:65–75No RF11.1310.5210.12
≥1 RF10.689.709.18
75+No RF8.087.717.47
≥1 RF7.366.436.02
GI Risk 2:65–75No RF10.7510.359.98
≥1 RF10.279.559.06
75+No RF7.847.607.39
≥1 RF7.096.355.96
GI Risk 3:65–75No RF10.1210.029.71
≥1 RF9.629.258.82
75+No RF7.447.397.21
≥1 RF6.666.195.83
Notes:
  1. Columns for warfarin, aspirin and no treatment groups contain QALY data.
Source of fundingNot Reported
Additional comments
  • Not clear in all cases whether results are for general populations receiving stroke prophylaxis or AF populations.
NCC CC ID (Ref Man)2
Evidence Table: What is the efficacy of anticoagulation versus antiplatelet therapy for stroke prevention in patients with permanent AF?
Bibliographic reference[969]: Blackshear JL. Et al, Fecal hemoglobin excretion in elderly patients with atrial fibrillation: combined aspirin and low-dose warfarin vs conventional warfarin therapy. Archives of Internal Medicine: 156: 658–660. 1996
Study typeRCT with auxiliary non-randomised cohort (study participants were split into a high-risk and low-risk cohort and subsequently randomised within each cohort)
Study ObjectiveComparison of faecal haemoglobin levels between high-risk patients receiving warfarin therapy and low-risk patients receiving aspirin
Evidence level1- (with a non-randomised cohort group that if included would imply a 2- grade)
Number of patientsN
Low-dose warfarin + aspirin (LDW+A)31
Adjusted-dose warfarin (ADW)32
Aspirin (A)54
Total:117
Patient characteristicsInclusion Criteria:
  • All Participants: Participation in the SPAF III study
  • Warfarin Group: classification as high-risk of stroke based on presence of one risk factor and currently receiving aspirin therapy
  • Aspirin Group: Classification as low-risk of stroke based on absence of any stroke risk factor
LDW+AADWA
Age170 (10)76 (7)70 (8)
Notes:
5.

There was a significant difference noted between the mean ages. This is partly a reflection of the intentional difference in stroke risk (age was defined as a risk factor). Age is expressed in years (standard deviation).

InterventionLDW+A: 1 to 3 mg per day warfarin and 325mg per day enteric-coated aspirin
ADW: warfarin to international normalised ratio (INR) 2.0 to 3.0
ComparisonA: 325mg per day enteric-coated aspirin
Length of follow-up1 month
Outcome measuresHaemoglobin content in stool, measured by HemoQuant instrument (Mayo Medical Laboratory, Rochester MI)
Effect sizeGroupMg Haemoglobin1
LDW+A1.7 (3.3)2
ADW1.0 (1.9)2
A0.8 (0.7)
Notes:
8.

Mg Haemoglobin = Milligrams of haemoglobin per gram of stool

9.

Mg Haemoglobin between LDW+A and ADW is significant at p=0.003

Source of fundingNot reported
Additional comments
  • Because A group was not included in randomisation process, comparisons of A with either warfarin group should be treated with caution
  • This study does not consider stroke as an outcome, and thus has been graded as 1-
NCC CC ID (Ref Man)969

From: Section 7 Evidence Tables

Cover of Atrial Fibrillation
Atrial Fibrillation: National Clinical Guideline for Management in Primary and Secondary Care.
NICE Clinical Guidelines, No. 36.
National Collaborating Centre for Chronic Conditions (UK).
Copyright © 2006, Royal College of Physicians of London.

All rights reserved. No part of this publication may be reproduced in any form (including photocopying or storing it in any medium by electronic means and whether or not transiently or incidentally to some other use of this publication) without the written permission of the copyright owner. Applications for the copyright owner’s written permission to reproduce any part of this publication should be addressed to the publisher.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.