A 68-year-old woman with a history of mitral valve replacement with a mechanical valve was admitted with abdominal pain. Because of the mechanical valve, she was chronically on warfarin (a blood thinner). At the time of admission, her international normalized ratio (INR) was 1.3, indicating that she was under-anticoagulated—her blood was not appropriately "thinned" to prevent possible clots from forming on her mechanical valve. She was treated with unfractionated heparin, a continuously delivered intravenous medication, which quickly thinned her blood appropriately.

Based on her abdominal complaints, a gastroenterology consultant wished to pursue esophagogastroduodenoscopy (EGD), a procedure in which a camera is inserted through the mouth and into the stomach. The gastroenterologist wanted the patient to be off all blood thinners to prevent bleeding complications and in case he needed to do biopsies.

The admitting team caring for the patient wrote an order to stop the heparin at 5:00 pm on the day before the EGD as well as an order to restart it 48 hours afterward. Unfortunately, the order to restart the heparin was missed, and the heparin was not restarted after the procedure. Four days after the EGD, the team noticed that the patient's partial thromboplastin time (PTT), a different marker of blood thinning (one that goes up in response to heparin), was normal and realized that the patient was not being anticoagulated appropriately by the heparin. The heparin was quickly restarted, and the patient was eventually discharged. She did not suffer any negative consequences, but her prolonged period off anticoagulants put her at high risk for acute stroke from blood clots on her mechanical valve.

by Andrew S. Dunn, MD

A number of different errors occurred in this case that placed the patient at risk for a bad outcome. But more broadly, the case highlights an inherently high-risk and common clinical challenge: the management of patients on oral anticoagulation (OAC) who need to undergo surgery or invasive procedures. Patients require oral anticoagulant therapy for many different reasons, including prevention of stroke in the setting of mechanical heart valves (as in this case), prevention of stroke in atrial fibrillation (a heart arrhythmia), and treatment of deep venous thrombosis or pulmonary embolism. Estimates suggest that approximately 250,000 patients each year in North America on oral anticoagulant therapy undergo procedures that require temporary interruption of OAC.(1) The appropriate management requires balancing competing risks, as withholding anticoagulation increases the risk of thromboembolism (TE), while administering parenteral anticoagulation ("bridging" therapy) with intravenous unfractionated heparin (UFH) or subcutaneous low-molecular-weight heparin (LMWH) increases the risk of bleeding. Here "bridging" refers to an anticoagulation bridge for the patient around the time of surgery—providing full-dose anticoagulation with intravenous or subcutaneous anticoagulants as close as possible to the time of surgery and resuming full-dose anticoagulation as soon as possible after surgery. One clear source of the clinical dilemma is the lack of high-quality data to help guide clinicians. There have been no large-scale, randomized trials and very few controlled studies comparing various strategies. Thus, clinicians and guideline panels must use the limited evidence and extrapolate data from other settings to choose strategies and formulate recommendations, respectively.

While a full discussion of all issues in this complex clinical scenario is beyond the scope of this commentary, several key factors must be considered in determining optimal management. Specifically, the clinician must decide if a given patient will benefit from ongoing parenteral anticoagulation ("bridging") during the periprocedural period if their long-term oral anticoagulation must be interrupted. There are several primary considerations (Table 1). First, the risk of TE while anticoagulants are withheld must be estimated. The risk of embolic events is often calculated based on the annual rate determined in other, nonprocedural settings, which is then prorated over the interval patients are expected to be off of their anticoagulation. For example, an 80-year-old patient with atrial fibrillation (an irregular heart rhythm that can lead to stroke) and a CHADS score (a calculated estimate of stroke risk) of 3 has an annual stroke risk of 6% without anticoagulation, which translates into a risk of 0.13% over an 8-day interval (6% ÷ 8/365 = 0.13%).(2) However, the majority of large-cohort studies performed specifically in the setting of procedures or surgery have revealed four- to tenfold higher TE rates in the periprocedural period, when compared to estimated embolic rates in nonbridging settings. Some authors have postulated that this increased risk is due to a perioperative hypercoagulable state.(3-6) Thus, a more evidence-based estimate of the incidence of TE in this 80-year-old would be approximately 0.4%-1.0%. Clinicians generally should try to estimate the patient's risk of TE while off of anticoagulation. Though there is no precise risk of TE that mandates administration of bridging anticoagulation, most guidelines agree that a patient with an annual risk of stroke of approximately 10% or greater should receive bridging anticoagulation. This group includes patients with a mechanical mitral valve and most patients with atrial fibrillation who had a prior stroke. The American College of Chest Physicians (ACCP) Consensus Guidelines provide a more detailed breakdown of thromboembolic risk.(7)

Second, assuming that the blood thinners are completely stopped for the procedure itself, the risk of postprocedural major bleeding if full-dose anticoagulation is administered soon after surgery needs to be determined. Some studies suggest that the rate is low for invasive procedures and minor surgery (e.g., endoscopy) but may be substantially higher after major surgery.(4,5,8)

Third, the consequences of TE and bleeding (should they occur) need to be considered. TE in the setting of an artificial heart valve is typically catastrophic and leads to death or major disability in 70% of patients.(9) The long-term consequences of postoperative major bleeding are usually less severe, though it can still be fatal in up to 3% of patients.(10)

Fourth, the efficacy of anticoagulation needs to be assessed. Full-dose anticoagulation has been shown to be extremely efficacious for venous TE (approximately 90% reduction) and moderately effective for prevention of arterial TE for patients with atrial fibrillation or mechanical heart valves (66%-75%).(11-14)

As described above, there are only scattered data to help address these four key considerations. Given the limitations of the available data, it is not surprising that there is no consensus on management.(15,16) It should be noted, though, that there is some agreement at the extremes. Patients at the highest TE risk (e.g., mechanical mitral valve or VTE within the prior 3 months) typically receive bridging anticoagulation, while patients at the lowest risk (e.g., single VTE event >12 months prior or atrial fibrillation with 0-1 stroke risk factor) are usually managed simply by withholding warfarin during the perioperative/periprocedure period (i.e., they don't need bridging). For most patients and most procedures, there is no consensus on optimal management. Guideline panels have recognized the limited evidence and offer clinicians a range of clinical options for patients in the intermediate-risk group.(7) Using the limited evidence, clinicians need to develop a rational approach that can maximize benefit for patients at highest TE risk while minimizing risk. A management strategy based on these principles is presented in Table 2.

In the case above, because of the mechanical valve, the patient was believed to be at extremely high risk for TE. Thus, the clinicians chose to bridge the patient up to the time of the procedure, stop the anticoagulation during the procedure itself, and restart the parenteral anticoagulation shortly afterward. When a bridging strategy is chosen, implementation can be cumbersome and logistically difficult, which can lead to transition errors (as in this case) and harm to patients (which, luckily, did not happen in this case). The possibility of transition errors during bridging therapy was investigated by Bell and colleagues, who reviewed 45,220 elderly patients on OAC to determine, over a 5-year period, how frequently warfarin was unintentionally discontinued after elective surgery.(17) They found that warfarin prescriptions were not refilled after 7.5% of ambulatory procedures and 11.4% of overnight hospitalizations, compared to 4.8% for a control group who did not undergo a procedure. This result confirms that perioperative transition errors are common and suggests that the transition from hospital to home carries greater risk for this error than the transition after ambulatory procedures.

In recognition of the high-risk nature of anticoagulants, The Joint Commission has designated reducing harm from anticoagulant therapy as one of their National Patient Safety Goals (NPSGs).(18) The NPSG includes eight specific elements of performance, including using approved protocols for the management of anticoagulant therapy and evaluating the safety of anticoagulation practices. Application of these requirements and patient safety principles to bridging therapy is essential but difficult. Most importantly, hospitals and health care systems should aim to consolidate the process into areas that have expertise and a systematic approach to management. This may consist of a formal anticoagulation clinic or other group of providers with experience and skill in anticoagulation management. Written protocols for selecting patients who may benefit from bridging therapy and for administering bridging anticoagulation when indicated are important to foster uniformity of practice and to systematize complex aspects of management. The use of electronic systems to prompt clinicians when key aspects of management are due, such as assessing the INR prior to surgery and soon after warfarin is resumed, is also an increasingly important method of avoiding transition errors and improving patient safety.

Take-Home Points

• Perioperative management of OAC is a high-risk scenario due to the lack of high-quality data and the potential for errors in stopping and starting oral anticoagulants.
• Most patients at low TE risk do not require bridging, while bridging is likely beneficial for patients at the highest risk (see Table 2).
• The decision regarding bridging anticoagulation is most difficult for intermediate-risk patients (see Table 2). The risks may outweigh the benefits for those undergoing major surgery. For patients on OAC due to VTE, use of prophylaxis-dose anticoagulation may be an efficacious and safer option.
• Specific protocols need to be in place to ensure that the timing of bridging doses and restarting of the patients' OAC occur correctly.

Andrew S. Dunn, MD
Professor of Medicine

Chief, Division of Hospital Medicine
Mount Sinai School of Medicine

New York, NY

1. Spyropoulos AC, Turpie AGG. Perioperative bridging interruption with heparin for the patient receiving long-term anticoagulation. Curr Opin Pulm Med. 2005;11:373-379. [go to PubMed]

2. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA. 2001;285:2864-2870. [go to PubMed]

3. Pengo V, Cucchini U, Denas G, et al; Italian Federation of Centers for the Diagnosis of Thrombosis and Management of Antithrombotic Therapies (FCSA). Standardized low-molecular-weight heparin bridging regimen in outpatients on oral anticoagulants undergoing invasive procedure or surgery: an inception cohort management study. Circulation. 2009;119:2920-2927. [go to PubMed]

4. Kovacs MJ, Kearon C, Rodger M, et al. Single-arm study of bridging therapy with low-molecular-weight heparin for patients at risk of arterial embolism who require temporary interruption of warfarin. Circulation. 2004;110:1658-1663. [go to PubMed]

5. Douketis JD, Johnson JA, Turpie AG. Low-molecular-weight heparin as bridging anticoagulation during interruption of warfarin: assessment of a standardized periprocedural anticoagulation regimen. Arch Intern Med. 2004;164:1319-1326. [go to PubMed]

6. Dunn AS, Turpie AGG. Perioperative management of patients receiving oral anticoagulants: a systematic review. Arch Intern Med. 2003;163:901-908. [go to PubMed]

7. Douketis JD, Berger PB, Dunn AS, Jaffer AK, Spyropoulos AC, Becker RC, Ansell J; American College of Chest Physicians. The perioperative management of antithrombotic therapy: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th edition). Chest. 2008;133:299S-339S. [go to PubMed]

8. Dunn AS, Spyropoulos AC, Turpie AGG. Bridging therapy in patients on long-term oral anticoagulants who require surgery: the Prospective Peri-operative Enoxaparin Cohort Trial (PROSPECT). J Thromb Haemost. 2007;5:2211-2218. [go to PubMed]

9. Petty GW, Brown Jr RD, Whisnant JP, Sicks JD, O'Fallon WM, Wiebers DO. Ischemic stroke subtypes: a population-based study of functional outcome, survival, and recurrence. Stroke. 2000;31:1062-1068. [go to PubMed]

10. Kearon C, Hirsh J. Management of anticoagulation before and after elective surgery. N Engl J Med. 1997;336:1506-1511. [go to PubMed]

11. Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med. 1999;131:492-501. [go to PubMed]

12. Cannegieter SC, Rosendaal FR, Briët E. Thromboembolic and bleeding complications in patients with mechanical heart valve prostheses. Circulation. 1994;89:635-641. [go to PubMed]

13. Agnelli G, Prandoni P, Becattini C, et al; Warfarin Optimal Duration Italian Trial Investigators. Extended oral anticoagulant therapy after a first episode of pulmonary embolism. Ann Intern Med. 2003;139:19-25. [go to PubMed]

14. Kearon C, Ginsberg JS, Kovacs MJ, et al; Extended Low-Intensity Anticoagulation for Thrombo-Embolism Investigators. Comparison of low-intensity warfarin therapy with conventional-intensity warfarin therapy for long-term prevention of recurrent venous thromboembolism. N Engl J Med. 2003;349:631-639. [go to PubMed]

15. Douketis JD, Crowther MA, Cherian SS, Kearon CB. Physician preferences for perioperative anticoagulation in patients with a mechanical heart valve who are undergoing elective noncardiac surgery. Chest. 1999;116:1240-1246. [go to PubMed]

16. Douketis JD, Crowther MA, Cherian SS. Peri-operative anticoagulation in patients with chronic atrial fibrillation who are undergoing surgery: results of a physician survey. Can J Cardiol. 2000;16:326-330. [go to PubMed]

17. Bell CM, Bajcar J, Bierman AS, Li P, Mamdani MM, Urbach DR. Potentially unintended discontinuation of long-term medication use after elective surgical procedures. Arch Intern Med. 2006;166:2525-2531. [go to PubMed]

18. The Joint Commission. Accreditation Program: Hospital. 2009 National Patient Safety Goals. [Available at]

Table 1. Key Factors to Consider in Managing Periprocedural Anticoagulation.

Table 2. Approach to the Perioperative Management of Patients on Oral Anticoagulation.