Venous thromboembolism: vitamin K antagonists reduce recurrent episodes, but the benefits decrease over time

Clinical bottom line (level 1a)

  1. Patients with venous thromboembolism who take vitamin K antagonists for longer periods compared with shorter are less likely to have another episode (NNT = 9 at 2 years) , but are more likely to bleed (NNH = 69 at 2 years) .
  2. The reduction in risk continues for as long as vitamin K antagonists are used.
Hutten and Prins: Cochrane Library 2000; 3 : -
Expires November 2003

The study

Systematic review of all randomised controlled trials of
  • Patients: symptomatic venous thromboembolism
  • Intervention: different durations of treatment with vitamin K antagonists
  • Outcome: recurrent venous thromboembolism, death, major bleeding

    Articles found in using Medline, Embase, (search terms: Cochrane review gropu on peripheral vascular diseases ) and hand-searching relevant journals, and contacting colleagues

    Selection criteria: by 2 independent reviewers - see above and below
    Appraisal criteria: randomisation and concealment of allocation
    Articles excluded if:
    • different target INR ranges used
    • continuous use of another anticoagulant or antiplatelet drug
    • duplicate reports or preliminary reports of data later presented in full
    • no objective tests to confirm diagnosis of DVT or PE
    • no data on thromboembolic events or bleeding available
    • studies not blinded
    • no breakdown for major and minor bleeding


    4 studies found involving 1500 patients followed for 1 to 4 years

    The evidence

    Long v. short anticoagulation
    Outcome Time to outcome CER OR
    (95% CI)    		 NNT
    (95% CI)
    current venous thromboembolism 2 years 137/742
    (18.5%)    		 0.37
    (0.28 to 0.51)     		9
    (8 to 12)
    major bleeding 2 years 4/742
    (0.54%)    		 3.75
    (1.63 to 8.62)     		-69
    (-300 to -25)
    death 2 years 50/742
    (6.7%)    		 0.70
    (0.45 to 1.09)     		52
    (NNT = 180 to infinity;
    NNH = 28 to infinity)

    Comments

    1. On ceasing anticoagulation in the long arm, there was no difference in recurrent rates of venous thromboembolism between the two groups.
    2. The authors noted that the absolute risk of recurrent venous thromboembolism declines over time, so the benefits of anticoagulation will fall (i.e. the NNT rises over time). However the risk of bleeding from anticoagulation remains constant. Therefore for each patient a point will be reached when the risks outweigh the benefits.

    Citation

    1. Hutten BA, and Prins MH: duration of treatment with vitamin K antagonists in symptomatic venous thromboembolism (Cochrane review). Cochrane Library 2000; 3 : -
    Search Terms: thromboembol* in Cochrane Library
    Contributor: Chris Ball, November 2001
    Reviewer:

    Clinical Question.
    Patient
    Intervention or Exposure
    Outcome