Article Text

Management of intracranial bleeding associated with anticoagulation: balancing the risk of further bleeding against thromboembolism from prosthetic heart valves
  1. T G PHAN,
  1. Department of Neurology, Mayo Clinic and Foundation, W8A, 200 First Street Southwest, Rochester MN 55905, USA
  1. Professor E F M Wijdicks wijoe{at}
  1. D R WREN
  1. Department of Neurology, Atkinson Morley's Hospital, Copse Hill, Wimbledon SW20 0NE, UK

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    We read with interest the article by Crawleyet al.1 We respectfully take issue with the authors.

    (1) In this article on the management of intracranial bleeding associated with anticoagulation, the authors reported on a patient who had an “intracerebral haematoma” and later developed two new “intracranial haematomas” after heparin therapy. We hope that the authors were referring to either lobar haemorrhage or basal ganglia haemorrhage. If the patient had a subdural haematoma, also classified under intracranial haemorrhage, then a surgical procedure as well as discontinuation of anticoagulation and reversal would have been the preferred treatment.

    (2) In their review of the literature, the authors did not discuss the articles by Wijdicks et al and Babikian et al on the relative safety of discontinuation of oral warfarin after intracranial haemorrhage in patients with mechanical heart valves.2 3 These authors have found that temporary discontinuation of warfarin for 1 to 2 weeks was relatively safe.

    (3) Crawley et al estimated a 0.016% daily risk of embolism or 0.67% over 6 weeks.1 In our experience, although the risk is relatively low, it is in the order of 3% over 30 days.4

    (4) Crawley et al correctly stated that having reversed anticoagulation in patients with prosthetic heart valves, it is uncertain when to restart it.1 None of the patients in our experience had recurrence of intracranial haemorrhage on restarting warfarin (after a short period of discontinuation) during their stay in hospital.4 Thus we do not recommend a prolonged period of warfarin discontinuation in patients who are at high risk of embolisation. Additionally we recommend that these patients be screened with echocardiography in evaluating the risk-benefit ratio of warfarin discontinuation and the urgency of restarting anticoagulation.


    Wren replies:

    We fully agree with Phan and Wijdicks that if a patient has a subdural haematoma while on treatment with warfarin surgical drainage might be required. The particular patient reported had recurrent lobar haemorrhage.1-1 Photographs of the brain CT were supplied but not published.

    We are aware of the articles by Wijdicks et al 1-2 and Babikian et al 1-3 as well as the recently published article by Phan et al, 1-4 which is an extension of the previous article published by Wijdickset al. 1-1 The dilemma of reinstituting anticoagulation for patients with cardioembolic sources and intracranial haemorrhages is discussed by Hacke1-5 in the editorial accompanying the recent paper by Phan et al.1-4 Of particular interest is the apparent paradox between the reported embolic risk without anticoagulation with modern artificial heart valves in the range of 4 per 100 patient-years and observed risk in the order of 3% over 30 days in the retrospective studies of Phanet al going up to 20% in the study of Bertram et al. 1-4 1-6

    We agree with Phan and Wijdicks and Wijdicks et al that patients at high risk of embolisation should have a limited period of warfarin discontinuation and that each patient needs to be assessed individually as suggested by Hacke.1-5We also agree with Hacke's suggestion that a prospective registry would be very useful given the difficulties of setting up any form of randomised control trial. The potential prothrombolic effects of haemorrhage and reversal of anticoagulation are also subjects that merit investigation with, for example, thromboelastography.


    1. 1-1.
    2. 1-2.
    3. 1-3.
    4. 1-4.
    5. 1-5.
    6. 1-6.