Anticoagulation and intracranial haemorrhage: increased risk with INR>4.0, previous stroke or artificial heart valves.

Clinical bottom line (level 3b)

  1. Patients on warfarin were at an increased risk of intracranial bleed with:
    • increasing INR>4.0 (per 0.5 increase) (NNH = 37 at unknown)
    • history of cerebrovascular disease (NNH = 98 at unknown)
    • prosthetic heart valves (NNH = 120 at unknown)
    • increasing age (per decade) (NNH = 320 at unknown)
Hylek and Singer: Annals of Internal Medicine 1994; 120 (11): 897-902
Stroke prevention in atrial fibrillation investigators : Lancet 1996; 348: 633-638
Expires June 2003

The study

Case-control study with objective outcomes, adjusted for confounding factors, not validated in an independent set of patients.

Setting: anticoagulation clinic, university hospital, USA

352 patients (aged 42; aged 65 to 79 years, 54% male) on warfarin

Excluded if
  • <18 years
  • haemorrhage from head injury
  • bleeding diathesis
  • anatomical abnormality


  • Cases: 121 patients (52% male, mean age 42): intracranial haemorrhage (diagnosed by CT, lumbar puncture or autopsy)
    Controls: 231 patients (52% male, mean age 42): randomly selected from clinic

    Factors studied:
  • intracranial haemorrhage


  • Factors summarised:
  • PT ratio nearest to admission (risk increases per unit increase)
  • history of cerebrovascular disease
  • prosthetic heart valve
  • age (risk increase per decade)


  • Logistic regression analysis was used to adjust for confounding factors.

    Outcomes studied:
  • intracranial haemorrhage (from ref. 2)

  • The evidence

    Patient expected event rate for intracranial haemorrhage (from ref. 2): 0.8%
    risk factor for
    intracranial haemorrhage (from ref. 2)
    adjusted OR
    (95% CI)
    NNH
    (95% CI)
    PT ratio nearest to admission (risk increases per unit increase) 4.5
    (2.6 to 8.0)
    37
    (19 to 80)
    history of cerebrovascular disease 2.3
    (1.4 to 3.7)
    98
    (48 to 320)
    prosthetic heart valve 2.1
    (1.2 to 3.8)
    120
    (46 to 630)
    age (risk increase per decade) 1.4
    (1.2 to 1.7)
    320
    (180 to 630)

    • 64% of cases had an intracerebral bleed (of which 46% were fatal); the rest were subdural (of which 20% were fatal).
    • Prothrombin ratio was approximately 2 x INR.
    • Risks of intracranial bleed increase sharply when PT ratio >2.0 (ie. INR>4.0).

    Citation

    1. Hylek EM, and Singer DE: Risk factors for intracranial hemorrhage in outpatients taking warfarin. Annals of Internal Medicine 1994; 120 (11): 897-902
    2. Stroke prevention in atrial fibrillation investigators , : Adjusted-dose warfarin versus low intensity fixed-dose warfarin and aspirin for high-risk patients with atrial fibrillation. Lancet 1996; 348: 633-638
    Search Terms: atrial fibr* and anticoag* in Cochrane
    Contributor: Chris Ball and Clare Wotton, June 2000
    Reviewer:

    Clinical Question.
    Patient on warfarin
    Intervention or Exposure risk factors
    Outcome intracranial haemorrhage