In summary, patients who received early anticoagulation for acute ischemic stroke:

Benefits in NNT

  • None were helped (no death or dependence prevented)
  • 0.86% reduced risk of recurrent ischemic stroke
  • 0.34% reduced risk pulmonary emboli
  • None were helped (no death or dependence prevented)
  • 1 in 117 were helped (recurrent ischemic stroke prevented)
  • 1 in 295 were helped (pulmonary emboli prevented)

Harms in NNT

  • 0.97% increased risk of symptomatic intracranial hemorrhage
  • 0.89% increased risk of extra-cranial hemorrhage
  • 1 in 103 were harmed (symptomatic intracranial hemorrhage)
  • 1 in 112 were harmed (extra-cranial hemorrhage)

View As:

Source: Sandercock PAG, Counsell C, Kane EJ. Anticoagulants for acute ischaemic stroke. Cochrane Database Syst Rev. 2015;(3):CD000024.


Study Population: 28 trials comprising 24,025 patients

Efficacy Endpoints: Reduction in death or dependence, death from all causes, recurrent ischemic stroke, pulmonary emboli, deep vein thrombosis

Harm Endpoints: Symptomatic intracranial hemorrhage or extra-cranial hemorrhage

Narrative: Stroke is a leading cause of death and disability worldwide, and the global incidence is increasing due to aging populations and lifestyle changes.1, 2, 3 The most common cause of stroke is ischemia due to acute arterial occlusion from a blood clot.3, 4 Theoretically, anticoagulants (blood thinners) including heparins, vitamin K antagonists, factor Xa inhibitors, and specific thrombin inhibitors reduce clot propagation and decrease damaged cerebral tissue, which may improve survival rates and patient recovery.5 Anticoagulants may also reduce the risk of recurrent stroke, deep vein thrombosis (DVT), and pulmonary embolism (PE), though they are associated with an increased risk of intracranial and extra-cranial hemorrhage.5 Current guidelines recommend against urgent anticoagulation for patients with acute ischemic stroke, and the guidelines state that the safety and efficacy of argatroban, other thrombin inhibitors, and oral factor Xa inhibitors is not well established for the treatment of acute ischemic stroke.3

The Cochrane review discussed here included 28 trials (n = 24,025 patients).5 Authors included randomized controlled trials (RCTs) of patients with presumed or confirmed acute ischemic stroke who received anticoagulants or controls (included no treatment, placebo, aspirin, standard dose thrombolytic). Dates of publication from the included trials ranged between 1960 to 2020, and anticoagulation was initiated within 4.5 hours to 2 weeks from time of symptom onset depending on the individual study. Patients randomized greater than 14 days after stroke onset were excluded. Anticoagulants included subcutaneous and intravenous standard unfractionated heparin, low-molecular-weight heparins, subcutaneous and intravenous heparinoids, oral vitamin K antagonists, factor Xa inhibitors, and specific thrombin inhibitors. Patients with cerebral venous thrombosis and transient ischemic attack (TIA) were excluded, as well as those with intracranial hemorrhage (ICH) before entry.

The primary outcome was death or dependency (dependent on help from other people for activities of daily living) at least one month after stroke. Secondary outcomes included death from any cause during the scheduled treatment period, death from any cause during the scheduled follow-up period, objective evidence of DVT on imaging, recurrent stroke during treatment and during the follow-up period, at least one confirmed PE during life or at autopsy within the treatment or follow-up period, symptomatic ICH, and major extra-cranial hemorrhage (defined by the original article, or if not present, any fatal bleed or bleed requiring transfusion or operation).

Early anticoagulation did not reduce the risk of death or dependence at the end of follow-up (12 RCTs; 22,428 participants; high-certainty evidence) or the risk of death from all causes (22 RCTs; 22,602 participants; low-certainty evidence) during the treatment period. However, treatment was associated with an increase in symptomatic ICH (OR: 2.47; 95% CI, 1.90 to 3.21; absolute risk difference [ARD]: 0.97%; number needed to harm [NNH]: 103; 20 RCTs; 23,221 participants; moderate-certainty evidence) and extracranial hemorrhage (OR: 2.99; 95% CI, 2.24 to 3.99; ARI: 0.89%; NNH: 112; 18 RCTs; 22,255 participants; moderate-certainty evidence).

Early anticoagulation was associated with fewer recurrent ischemic strokes (odds ratio [OR]: 0.75; 95% confidence interval [CI], 0.65 to 0.88; absolute risk difference [ARD]: 0.86%; number needed to treat [NNT]: 117; 12 RCTs; 21,665 participants; moderate-certainty evidence), reduced frequency of symptomatic PE (OR: 0.60; 95% CI, 0.44 to 0.81; ARD: 0.34%; NNT: 295; 14 RCTs; 22,544 participants; high-certainty evidence), and a reduced risk of DVT (OR: 0.21; 95% CI, 0.15 to 0.29; ARD: 29.2%; NNT: 4; 10 RCTs; 916 participants; low-certainty evidence). Since the studies reporting the rate of DVT were small and possessed significant heterogeneity (low-certainty evidence), we do not report this outcome in the summary table.

Another systematic review published in 2011 arrived at a similar conclusion for early anticoagulation for acute ischemic stroke and found no survival or dependence benefit.6, 7 A patient-level meta-analysis of large trials published in 2013 sought to identify subgroups of patients most likely to derive net benefit from heparin; no benefit from anticoagulation in high-risk individuals was reported.8

Caveats: There are several important limitations to this systematic review. While the age of included patients ranged from 28 to 92 years, a significant proportion of patients were over the age of 70. Most trials excluded patients at high bleeding risk (e.g., liver or kidney failure, clotting disorders) and those with significant hypertension (>180 mm Hg systolic blood pressure or >120 mm Hg diastolic blood pressure). The time from stroke onset to initiation of anticoagulation significantly varied, with most studies using less than 48 hours. However, several trials enrolled subjects up to 2 weeks. This potentially limits the applicability to emergency clinicians. The scheduled period of anticoagulant treatment was 1-2 weeks in 24 trials and 1 month in 4 trials, a relatively short time when compared with anticoagulation for other conditions (e.g., PE or DVT). Imaging with head computed tomography (CT) was obtained prior to randomization for all patients in 18 trials; imaging time varied in the remaining trials. Two trials were conducted before CT was available, and in 3 trials almost no CT scans were performed even though it was available; thus, some patients with ICH may have been included. The follow-up period was generally short for most trials, with 4 RCTs following patients for only 14 days. The absence of long-term follow-up in some of the trials could have resulted in missed outcomes after the follow-up period. Accurate assessment for disability requires time for recovery, which is typically at least several months, and few trials evaluated long-term functional status. This is an important outcome, as treatments that prevent death can lead to living with significant disability. However, the IST trial followed patients for 6 months.9 This trial was conducted over 20 years ago and enrolled over 80% of the included patients in the meta-analysis.9

Moreover, approximately one-quarter of studies had incomplete blinding of participants, study personnel, or outcome assessors. Another important limitation is that there only limited information was reported regarding subgroups including the type and dose of the anticoagulants used. Other measures to prevent clot formation such as aspirin, early mobilization, and pneumatic compression were not routinely included, though these may be beneficial in reducing this risk.3, 6, 7, 8, 10 While the risk of DVT was reduced with anticoagulation, trials including patients receiving anticoagulation to prevent DVT had severe strokes with a high likelihood of paralysis of a lower extremity. Consequently, this may reflect a cohort at much greater risk of DVT than the average patient with a stroke. Moreover, the studies reporting a reduction in DVT were small with high heterogeneity and primarily diagnosed DVT through 125-I-labeled fibrinogen scanning, rather than ultrasound. These DVTs were generally subclinical and asymptomatic, and thus the clinical benefit of treating these DVTs is unknown. Recent evidence suggests the overall prevalence of symptomatic DVT after acute stroke ranges from 1-10%,11, 12, 13 so it is likely that the benefit may be exaggerated in the current review. Finally, many of the trials were conducted over 20 years ago, and many patients did not receive comprehensive care in dedicated stroke units, which is a vital component of current stroke care.3

Based on the existing data, anticoagulation for those with acute ischemic stroke does not appear to reduce death or disability. While anticoagulation may reduce the risk of recurrent stroke, DVT, and PE, it is also associated with an increased risk of ICH and extracranial hemorrhage. Therefore, we have assigned a color recommendation of Yellow (unclear if benefits) to this intervention. Further studies should include patients treated in dedicated stroke centers and managed in combination with other interventions (e.g., aspirin, early mobilization, rehabilitation, etc.) and with longer follow-up periods.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author: Brit Long, MD; Michael Gottlieb, MD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated: February 14, 2022

  1. The Title Bar

    The title bar is color-coded with our overall recommendation.

    • Green: Benefits outweigh risks.
    • Yellow: Unclear risk/benefit profile.
    • Red: Benefits do not outweigh risks.
    • Black: Obvious harms, no clear benefits.
  2. Tip content...