Stroke: Anticoagulant therapy has no effect on death, dependence or recurrent stroke

Clinical bottom line (level 1a)

  1. Patients with an acute stroke who receive anticoagulants are not less likely to die or be dependent.
  2. Patients given anticoagulants are less likely to have a DVT (NNT = 3 at unknown) or symptomatic PE (NNT = 277 at unknown) , but have more extracranial bleeds (NNH = 134 at unknown) .
  3. Patients given anticoagulants have fewer ischaemic strokes (NNT = 119 at unknown) , but more symptomatic intracranial bleeds (NNH = 140 at unknown) , so there is no overall difference in the subsequent stroke rate.
Gubitz et al: Cochrane Library 1999; 1: -
Expires November 2002

The study

Systematic review of all randomised controlled trials of
  • Patients: acute presumed or confirmed ischaemic stroke
  • Intervention: anticoagulants: subcutaneous and intravenous unfractionated heparin, low-molecular-weight heparins, subcutaneous and intravenous heparinoids, oral anticoagulants and specific thrombin inhibitors
  • Outcome: death, symptomatic or asymptomatic deep vein thrombosis, symptomatic pulmonary embolism, recurrent stroke, haemorrhage


    • Articles found in all languages using Cochrane Stroke Group trials register and MedStrategy, ? (search terms: detailed in text ) and contacting pharmaceutical companies, and searching trial register of Antithrombotic Therapy Trialists's Collaboration

    Selection criteria: see above and exclusion criteria
    Appraisal criteria: quality based on randomisation method, blinding, intention-to-treat analysis possible, loss to follow-up
    Articles excluded if:
    • unblinded or not randomised
    • patients randomised > 14 days after stroke onset
    • transient ischaemic attacks only
    • only included patients with intracranial haemorrhage


    21 studies found involving 23427 patients (aged 28 to 92). Trials commonly excluded patients thought to be at high-risk of bleeding (clotting disorders, hepatic or renal failure), or severe hypertension
    Only the effect on DVT was found to be heterogeneous - this was resolved by excluding studies with uncertain concealment.

    The evidence

    Outcome Time to outcome CER OR
    (95% CI)    		 NNT
    (95% CI)
    death or dependence 4 weeks 6454/10737
    (61%)    		 0.99
    (0.94 to 1.05)     		418
    (NNT = 87 to infinity;
    NNH = 67 to infinity)
    death from all causes during treatment period unknown 962/11086
    (8.7%)    		 0.99
    (0.90 to 1.09)     		1258
    (NNT = 141 to infinity;
    NNH = 125 to infinity)
    death from all causes 4 weeks 2274/11032
    (21%)    		 1.05
    (0.98 to 1.12)     		-122
    (NNT = 51 to infinity;
    NNH = 300 to infinity)
    DVT during treatment period unknown 204/460
    (44%)    		 0.21
    (0.15 to 0.29)     		3
    (3 to 4)
    symptomatic PE during treatment period unknown 102/11016
    (0.93%)    		 0.61
    (0.45 to 0.83)     		277
    (196 to 637)
    recurrent ischaemic/unknown stroke during treatment period unknown 388/10739
    (3.6%)    		 0.76
    (0.65 to 0.88)     		119
    (81 to 239)
    symptomatic intracranial haemorrhage during treatment period weeks 54/11198
    (0.48%)    		 2.52
    (1.92 to 3.30)     		-139
    (-229 to -92)
    any recurrent stroke or symptomatic intracranial haemorrhage unknown 435/10739
    (4.1%)    		 0.97
    (0.85 to 1.11)     		847
    (NNT = 232 to infinity;
    NNH = 169 to infinity)
    major extracranial haemorrhage during treatment period weeks 42/10927
    (0.38%)    		 2.99
    (2.24 to 3.99)     		-134
    (-214 to -89)

    Comments

    1. 20,000 of the 23,427 patients in this overview came from one trial (the IST). Excluding this trial did not substantially alter the results.
    2. Most trials lacked longer-term follow-up (3 to 6 months), before which an assessment of disability cannot meaningfully be made

    Citation

    1. Gubitz G, Counsell C, Sandercock P, et al: anticoagulants for acute ischaemic stroke (Cochrane Review): Update Software. Cochrane Library 1999; 1: -
    Search Terms: stroke in Cochrane
    Contributor: Chris Ball and Musab Hayatli, November 1999
    Reviewer: Rowan Harwood

    Clinical Question.
    Patient acute stroke
    Intervention or Exposure anticoagulant
    Outcome death, dependence, stroke, haemorrhage, DVT, PE