Atrial fibrillation: more bleeds occurred on warfarin than aspirin.

Clinical bottom line (level 1b)

  1. Patients with atrial fibrillation on warfarin compared with aspirin, were at increased risk of major haemorrhage (NNH = 31 at 2.6 years) .
  2. Around 6% of patients on warfarin had a bleed by 2.6 years, and around 2% had an intracranial bleed.
Stroke Prevention in Atrial Fibrillation Investigators : Archives of Internal Medicine 1996; 156: 409-416
Expires June 2003

The study

Unblinded concealed randomised trial with intention-to-treat
Setting: sixteen acute hospitals, USA

1100 patients (aged mean 70 years, ?% male) non-valvular atrial fibrillation

Excluded if
  • creatinine >265 mmol/dl
  • plt<100
  • haemoglobin <10
  • prothrombin ratio >2 on two occasions (~INR 4.0)
  • blood pressure >180/100 mmHg
  • chronic alcoholism
  • AST or ALT >3 times normal
  • faecal occult blood
  • follow-up impossible
  • GU or GI bleed within 6 months
  • on NSAIDs
  • previous intracranial bleed
  • condition predisposing to head injury


    • Control Group: (n = 545, 545 analysed): aspirin 325 mg/day
    Experimental Group: (n = 555, 555 analysed): warfarin adjusted so INR 2.0 to 4.5

    100% followed for 2.6 years
    Outcome notes:
    • major haemorrhage : major bleed if intracranial/spinal (confirmed CT) or admitted, transfused, or required surgery, or lead to permanent functional decrease

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)    		ARR
    (95% CI)    		NNH
    (95% CI)
    major haemorrhage 2.6 years 16
    (2.94%)    		 34
    (6.13%)    		 -109%
    (-274% to -17%)    		 -3.19%
    (-5.64% to -0.74%)    		 31
    (18 to 140)
    intracranial bleed 2.6 years 5
    (0.92%)    		 13
    (2.34%)    		 -155%
    (-611% to 8%)    		 -1.42%
    (-2.92% to 0.07%)    		 70
    (NNT = 1500 to infinity;
    NNH = 34 to infinity)

  • Multivariate analysis was performed on risk factors.
  • Significant risk factors (no numbers given) were:
    • increasing age
    • increasing number of prescription medication
  • 8/18 (44%) of intracranial bleeds were fatal- no patient made a full recovery.

    • Citation

    1. Stroke Prevention in Atrial Fibrillation Investigators , : Bleeding during antithrombotic therapy in patients with atrial fibrillation. Archives of Internal Medicine 1996; 156: 409-416
    Search Terms: atrial fibr* and anticoag* in Cochrane
    Contributor: Chris Ball and Clare Wotton, June 2000
    Reviewer:

    Clinical Question.
    Patient atrial fibrillation
    Intervention or Exposure warfarin
    Comparison aspirin
    Outcome bleeding