Stroke: clinical features may help predict thromboembolism.

Clinical bottom line (level 1b)

  1. Less than a tenth of patients with nonrheumatic atrial fibrillation had a thromboembolic event.
  2. Patients were at increased risk of a thromboembolic event if they had:
    • history of hypertension (NNF = 10 for unknown)
    • previous thromboembolism (NNF = 11 for unknown)
    • recent congestive heart failure (NNF = 8 for unknown)
    • diabetes (NNF = 11 for unknown)
    • diuretic use (NNF = 18 for unknown)
    • systolic blood pressure >160 at entry (NNF = 12 for unknown)
The Stroke Prevention in Atrial Fibrillation Investigators : Annals of Internal Medicine 1992; 116 (1): 1-5
Expires October 2003

The study

Prospective cohort study with objective outcomes, adjusted for confounding factors, validated in an independent set of patients.

Setting: 15 medical centres, USA

568 patients (aged mean 67y, 70% male) nonrheumatic atrial fibrillation

Excluded if
  • mitral stenosis
  • transient ischaemic events were not monitored



    • Factors studied:
  • thromboembolic event
  • history of hypertension
  • previous thromboembolism
  • recent congestive heart failure
  • diabetes
  • diuretic use
  • systolic blood pressure >160 at entry




    • Multivariate analysis was used to adjust for confounding factors.

    100% followed for mean 1.3y
    Outcomes studied:
  • thromboembolic event (ischaemic stroke or systemic emboli)
  • thromboembolism (exploratory variables)

    • Data was taken from a randomised controlled trial, studying the effects of warfarin, aspirin or placebo on the prevention of ischaemic stroke and systemic embolism. The patients studied here were from the placebo arm.
    • A second analysis was carried out, not validating previously described factors but assessing for other clinical factors which may lead to thromboembolism.

    The evidence

    outcome time to outcome number of patients/total number %
    (95% CI)
    thromboembolic event mean 1.3y 46/568 8.1%
    (5.9% to 10%)
    thromboembolism (exploratory variables) mean 1.3y 46/568 8.1%
    (5.9% to 10%)

    prognostic factor for
    thromboembolic event     		time to outcome adjusted RR
    (95% CI)    		 NNF+
    (95% CI)
    history of hypertension mean 1.3y 2.2
    (1.1 to 4.3)     		10
    (4 to 120)
    previous thromboembolism mean 1.3y 2.1
    (1.0 to 4.2)     		11
    (4 to infinity)
    recent congestive heart failure mean 1.3y 2.6
    (1.2 to 5.4)     		8
    (3 to 62)

    prognostic factor for
    thromboembolism (exploratory variables)     		time to outcome unadjusted RR
    (95% CI)    		 NNF+
    (95% CI)
    diabetes mean 1.3y 2.1
    (1.1 to 3.9)     		11
    (4 to 120)
    diuretic use mean 1.3y 1.7
    (1.0 to 3.1)     		18
    (6 to infinity)
    systolic blood pressure >160 at entry mean 1.3y 2.0
    (1.0 to 3.9)     		12
    (4 to infinity)

    • thromboembolism rate (% per year (95% CI)) with risk factors of hypertension, recent congestive heart failure and previous thromboembolism:
      • no risk factors 2.5% (1.3 to 5.0)
      • one risk factor 7.2% (4.8 to 10.8)
      • two or three risk factors 17.6% (10.5 to 29.9)

    Comments

    1. The results should be viewed with some caution as the data was taken from a randomised controlled trial.

    Citation

    1. The Stroke Prevention in Atrial Fibrillation Investigators , : Predictors of thromboembolism in atrial fibrillation: I. Clinical features of patients at risk. Annals of Internal Medicine 1992; 116 (1): 1-5
    Contributor: Clare Wotton and Bob Phillips, October 2000
    Reviewer:

    Clinical Question.
    Patient nonrheumatic atrial fibrillation
    Intervention or Exposure risk factors
    Outcome arterial thromboembolism