Practice Practice Pointer

Practical management of coagulopathy associated with warfarin

BMJ 2010; 340 doi: (Published 19 April 2010) Cite this as: BMJ 2010;340:c1813
  1. David Garcia, associate professor1,
  2. Mark A Crowther, professor2,
  3. Walter Ageno, associate professor3
  1. 1Department of Internal Medicine, University of New Mexico, MSC08 4630, Albuquerque, NM 87131, USA
  2. 2Department of Internal Medicine, McMaster University, Hamilton, ON, Canada
  3. 3University of Insubria, Varese, Italy
  1. Correspondence to: D Garcia davgarcia{at}
  • Accepted 12 March 2010

When choosing a management strategy for a patient who is being treated with a vitamin K antagonist and presents with an INR outside the therapeutic range, consider the risk of both bleeding and thrombosis

Vitamin K antagonists are used by millions of patients to prevent thromboembolism. Primary care providers and specialists face clinical situations where the short term risk of bleeding or the risk of imminent thromboembolism in these patients must be assessed rapidly. In this article we provide several approaches to clinical questions that arise when patients who take vitamin K antagonists are bleeding or have an international normalised ratio (INR) above the therapeutic range. Our goal is to supplement available guidelines published by the American College of Chest Physicians and the British Committee for Standards in Haematology, which, although valuable resources, either do not directly address certain clinical scenarios or provide a range of options, suggesting that no one strategy is preferred, usually because high quality evidence is lacking.


We accessed previous publications from our group, including recent systematic reviews of the literature and original papers, and we integrated their reference lists by performing a new comprehensive literature review of Medline and EMBASE databases up to June 2009 to search for additional studies. We made specific suggestions on the basis of these reviews in concert with our clinical experience. We selected clinical studies to support our suggestions on the basis of the quality of their design (randomised controlled trials first, then prospective cohort studies and retrospective cohort studies). Overall, the available evidence supporting our recommendations is of moderate to high quality.

How should I manage a patient presenting with an INR >4.5 and no signs of bleeding?

What’s the risk of bleeding?

Data from a large registry of patients treated with warfarin show that the 30 day bleeding risk for someone with a single INR measurement between 5 and 9 is less than 1%.1 Clinicians should balance the risk …

View Full Text

Log in

Log in through your institution


* For online subscription