The Perioperative Management of Antithrombotic Therapy ... · drugs, we suggest against the routine use of platelet function assays to monitor the antithrom-botic effect of aspirin

DOI 10.1378/chest.08-0675 2008;133;299S-339SChest

Alex C. Spyropoulos, Richard C. Becker and Jack AnsellJames D. Douketis, Peter B. Berger, Andrew S. Dunn, Amir K. Jaffer, Practice Guidelines (8th Edition)

Clinicalof Chest Physicians Evidence-Based : American College*Antithrombotic Therapy

The Perioperative Management of

http://chestjournal.chestpubs.org/content/133/6_suppl/299S.full.html

services can be found online on the World Wide Web at: The online version of this article, along with updated information and

ISSN:0012-3692)http://chestjournal.chestpubs.org/site/misc/reprints.xhtml(

written permission of the copyright holder.this article or PDF may be reproduced or distributed without the priorDundee Road, Northbrook, IL 60062. All rights reserved. No part of Copyright2008by the American College of Chest Physicians, 3300Physicians. It has been published monthly since 1935.

is the official journal of the American College of ChestChest

© 2008 American College of Chest Physicians by guest on December 25, 2011chestjournal.chestpubs.orgDownloaded from


http://chestjournal.chestpubs.org/site/misc/reprints.xhtml

http://chestjournal.chestpubs.org/

The Perioperative Management ofAntithrombotic Therapy*American College of Chest PhysiciansEvidence-Based Clinical Practice Guidelines(8th Edition)

James D. Douketis, MD, FRCP(C); Peter B. Berger, MD, FACP;Andrew S. Dunn, MD, FACP; Amir K. Jaffer, MD;Alex C. Spyropoulos, MD, FACP, FCCP; Richard C. Becker, MD, FACP, FCCP;and Jack Ansell, MD, FACP, FCCP

This article discusses the perioperative management of antithrombotic therapy and is part of the AmericanCollege of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). The primary objectives ofthis article are the following: (1) to address the perioperative management of patients who are receiving vitaminK antagonists (VKAs) or antiplatelet drugs, such as aspirin and clopidogrel, and require an elective surgical orother invasive procedures; and (2) to address the perioperative use of bridging anticoagulation, typically withlow-molecular-weight heparin (LMWH) or unfractionated heparin (UFH). A secondary objective is to address theperioperative management of such patients who require urgent surgery. The recommendations in this articleincorporate the grading system that is discussed in this supplement (Guyatt G et al, CHEST 2008; 133:123S–131S).Briefly, Grade 1 recommendations are considered strong and indicate that the benefits do (or do not) outweighrisks, burden, and costs, whereas Grade 2 recommendations are referred to as suggestions and imply thatindividual patient values may lead to different management choices.The key recommendations in this article include the following: in patients with a mechanical heart valve or atrialfibrillation or venous thromboembolism (VTE) at high risk for thromboembolism, we recommend bridginganticoagulation with therapeutic-dose subcutaneous (SC) LMWH or IV UFH over no bridging during temporaryinterruption of VKA therapy (Grade 1C); in patients with a mechanical heart valve or atrial fibrillation or VTE atmoderate risk for thromboembolism, we suggest bridging anticoagulation with therapeutic-dose SC LMWH,therapeutic-dose IV UFH, or low-dose SC LMWH over no bridging during temporary interruption of VKAtherapy (Grade 2C); in patients with a mechanical heart valve or atrial fibrillation or VTE at low risk forthromboembolism, we suggest low-dose SC LMWH or no bridging over bridging with therapeutic-dose SCLMWH or IV UFH (Grade 2C).In patients with a bare metal coronary stent who require surgery within 6 weeks of stent placement, werecommend continuing aspirin and clopidogrel in the perioperative period (Grade 1C); in patients with adrug-eluting coronary stent who require surgery within 12 months of stent placement, we recommend continuingaspirin and clopidogrel in the perioperative period (Grade 1C).In patients who are undergoing minor dental procedures and are receiving VKAs, we recommend continuingVKAs around the time of the procedure and coadministering an oral prohemostatic agent (Grade 1B); in patientswho are undergoing minor dermatologic procedures and are receiving VKAs, we recommend continuing VKAsaround the time of the procedure (Grade 1C); in patients who are undergoing cataract removal and are receivingVKAs, we recommend continuing VKAs around the time of the procedure (Grade 1C).

(CHEST 2008; 133:299–339S)

Key words: arterial thromboembolism; aspirin; bleeding; bridging anticoagulation; clopidogrel; low-molecular-weightheparin; oral anticoagulant; perioperative; stroke; surgery; unfractionated heparin; venous thromboembolism; vitamin Kantagonist

Abbreviations: APTT � activated partial thromboplastin time; CABG � coronary artery bypass graft; CHADS2 � CongestiveHeart Failure-Hypertension-Age-Diabetes-Stroke; CI � confidence interval; DDAVP � 1-deamino-8-D-arginine vasopressin;INR � international normalized ratio; LMWH � low-molecular-weight heparin; NSAID � nonsteroidal antiinflammatory drug;OR � odds ratio; PCI � percutaneous coronary intervention; SC � subcutaneous; UFH � unfractionated heparin; VKA � vitaminK antagonist; VTE � venous thromboembolism

SupplementANTITHROMBOTIC AND THROMBOLYTIC THERAPY 8TH ED: ACCP GUIDELINES

www.chestjournal.org CHEST / 133 / 6 / JUNE, 2008 SUPPLEMENT 299S



Summary of Recommendations

2.0 Perioperative Management of Patients Who AreReceiving VKAs

2.1. In patients who require temporary inter-ruption of a VKA before surgery or a procedureand require normalization of the INR for thesurgery or procedure, we recommend stoppingVKAs approximately 5 days before surgery overstopping VKAs within a shorter time intervalbefore surgery to allow adequate time for theINR to normalize (Grade 1B).2.2. In patients who have had temporary inter-ruption of a VKA before surgery or a proce-dure, we recommend resuming VKAs approxi-mately 12 to 24 h (the evening of or the nextmorning) after surgery and when there is ade-quate hemostasis over resumption of VKAscloser to surgery (Grade 1C).2.3. In patients who require temporary inter-ruption of a VKA before surgery or a procedureand whose INR is still elevated (ie, > 1.5) 1 to 2days before surgery, we suggest administeringlow-dose (ie, 1 to 2 mg) oral vitamin K tonormalize the INR instead of not administeringvitamin K (Grade 2C).2.4. In patients with a mechanical heart valve oratrial fibrillation or VTE at high risk for throm-boembolism, we recommend bridging anticoag-ulation with therapeutic-dose SC LMWH or IVUFH over no bridging during temporary inter-ruption of VKA therapy (Grade 1C); we suggesttherapeutic-dose SC LMWH over IV UFH(Grade 2C). In patients with a mechanical heartvalve or atrial fibrillation or VTE at moderaterisk for thromboembolism, we suggest bridginganticoagulation with therapeutic-dose SC LMWH,therapeutic-dose IV UFH, or low-dose SCLMWH over no bridging during temporaryinterruption of VKA therapy (Grade 2C); wesuggest therapeutic-dose SC LMWH over other

management options (Grade 2C). In patientswith a mechanical heart valve or atrial fibrilla-tion or VTE at low risk for thromboembolism,we suggest low-dose SC LMWH or no bridgingover bridging with therapeutic-dose SC LMWHor IV UFH (Grade 2C).Values and preferences: In patients at high ormoderate risk for thromboembolism, the recom-mendations reflect a relatively high value on pre-venting thromboembolism and a relatively lowvalue is on preventing bleeding; in patients at lowrisk for thromboembolism, the recommendationsreflect a relatively high value on preventing bleed-ing and a relatively low value on preventing throm-boembolism.

3.0 Perioperative Management of Patients Who AreReceiving Bridging Anticoagulation

3.1. In patients who require temporary inter-ruption of VKAs and are to receive bridginganticoagulation, from a cost-containment per-spective we recommend the use of SC LMWHadministered in an outpatient setting wherefeasible instead of inpatient administration ofIV UFH (Grade 1C).Values and preferences: This recommendation re-flects a consideration not only of the trade-off be-tween the advantages and disadvantages of SCLMWH and IV UFH as reflected in their effects onclinical outcomes (LMWH at least as good, possiblybetter), but also the implications in terms of resourceuse (costs) in a representative group of countries(substantially less resource use with LMWH).3.2. In patients who are receiving bridginganticoagulation with therapeutic-dose SCLMWH, we recommend administering the lastdose of LMWH 24 h before surgery or a proce-dure over administering LMWH closer to sur-gery (Grade 1C); for the last preoperative doseof LMWH, we recommend administering ap-proximately half the total daily dose instead of100% of the total daily dose (Grade 1C). Inpatients who are receiving bridging anticoagu-lation with therapeutic-dose IV UFH, we rec-ommend stopping UFH approximately 4 h be-fore surgery over stopping UFH closer tosurgery (Grade 1C).3.3. In patients undergoing a minor surgicalor other invasive procedure and who arereceiving bridging anticoagulation with ther-apeutic-dose LMWH, we recommend resum-ing this regimen approximately 24 h after (eg,the day after) the procedure when there isadequate hemostasis over a shorter (eg, < 12h) time interval (Grade 1C). In patients under-going major surgery or a high bleeding risk

*From McMaster University (Dr. Douketis), Hamilton, ON,Canada; Geisinger Clinic (Dr. Berger), Danville, PA; MountSinai School of Medicine (Dr. Dunn), New York, NY; Universityof Miami (Dr. Jaffer), Leonard M. Miller, School of Medicine,Miami, FL; Clinical Thrombosis Center (Dr. Spyropoulos),Lovelace Medical Center, Albuquerque, NM; Duke University(Dr. Becker), Durham, NC; and Boston University (Dr. Ansell),Boston, MA.Manuscript accepted December 20, 2007.Reproduction of this article is prohibited without written permissionfrom the American College of Chest Physicians (www.chestjournal.org/misc/reprints.shtml).Correspondence to: Jack E. Ansell, MD, FACP, FCCP, Depart-ment of Medicine, Boston University Medical Center, 88 EastNewton St, Boston, MA 02118; e-mail: [email protected]: 10.1378/chest.08-0675

300S Antithrombotic and Thrombolytic Therapy 8th Ed: ACCP Guidelines



surgery/procedure and for whom postopera-tive therapeutic-dose LMWH/UFH is planned, werecommend either delaying the initiation oftherapeutic-dose LMWH/UFH for 48 to 72 h aftersurgery when hemostasis is secured, administer-ing low-dose LMWH/UFH after surgery whenhemostasis is secured, or completely avoidingLMWH or UFH after surgery over the adminis-tration of therapeutic-dose LMWH/UFH in closeproximity to surgery (Grade 1C). We recommendconsidering the anticipated bleeding risk and ad-equacy of postoperative hemostasis in individualpatients to determine the timing of LMWH orUFH resumption after surgery instead of resum-ing LMWH or UFH at a fixed time after surgeryin all patients (Grade 1C).3.4. In patients who are receiving bridging anti-coagulation with LMWH, we suggest against theroutine use of anti-factor Xa levels to monitor theanticoagulant effect of LMWHs (Grade 2C).

4.0 Perioperative Management of Patients Who AreReceiving Antiplatelet Therapy

4.2. In patients who require temporary inter-ruption of aspirin- or clopidogrel-containingdrugs before surgery or a procedure, we sug-gest stopping this treatment 7 to 10 days beforethe procedure over stopping this treatmentcloser to surgery (Grade 2C).4.3. In patients who have had temporary inter-ruption of aspirin therapy because of surgery ora procedure, we suggest resuming aspirin ap-proximately 24 h (or the next morning) aftersurgery when there is adequate hemostasis in-stead of resuming aspirin closer to surgery(Grade 2C). In patients who have had temporaryinterruption of clopidogrel because of surgeryor a procedure, we suggest resuming clopi-dogrel approximately 24 h (or the next morn-ing) after surgery when there is adequate he-mostasis instead of resuming clopidogrel closerto surgery (Grade 2C).4.4. In patients who are receiving antiplateletdrugs, we suggest against the routine use ofplatelet function assays to monitor the antithrom-botic effect of aspirin or clopidogrel (Grade 2C).4.5. For patients who are not at high risk forcardiac events, we recommend interruption ofantiplatelet drugs (Grade 1C). For patients athigh risk of cardiac events (exclusive of coro-nary stents) scheduled for noncardiac surgery,we suggest continuing aspirin up to and beyondthe time of surgery (Grade 2C); if patients arereceiving clopidogrel, we suggest interruptingclopidogrel at least 5 days and, preferably,within 10 days prior to surgery (Grade 2C). In

patients scheduled for CABG, we recommendcontinuing aspirin up to and beyond the time ofCABG (Grade 1C); if aspirin is interrupted, werecommend it be reinitiated between 6 h and48 h after CABG (Grade 1C). In patients sched-uled for CABG, we recommend interruptingclopidogrel at least 5 days and, preferably, 10days prior to surgery (Grade 1C). In patientsscheduled for PCI, we suggest continuing aspi-rin up to and beyond the time of the procedure;if clopidogrel is interrupted prior to PCI, wesuggest resuming clopidogrel after PCI with aloading dose of 300 to 600 mg (Grade 2C).4.6. In patients with a bare metal coronary stentwho require surgery within 6 weeks of stentplacement, we recommend continuing aspirinand clopidogrel in the perioperative period(Grade 1C). In patients with a drug-eluting coro-nary stent who require surgery within 12 monthsof stent placement, we recommend continuingaspirin and clopidogrel in the perioperative pe-riod (Grade 1C). In patients with a coronary stentwho have interruption of antiplatelet therapy be-fore surgery, we suggest against the routine use ofbridging therapy with UFH, LMWH, directthrombin inhibitors, or glycoprotein IIb/IIIa in-hibitors (Grade 2C).Values and preferences: These recommendations re-flect a relatively high value placed on preventing stent-related coronary thrombosis, a consideration of com-plexity and costs of administering bridging therapy inthe absence of efficacy and safety data in this clinicalsetting, and a relatively low value on avoiding theunknown but potentially large increase in bleeding riskassociated with the concomitant administration of aspi-rin and clopidogrel during surgery.

5.0 Perioperative Management of AntithromboticTherapy in Patients Who Require Dental, Dermato-logic, or Ophthalmologic Procedures

5.1. In patients who are undergoing minordental procedures and are receiving VKAs, werecommend continuing VKAs around the timeof the procedure and coadministering an oralprohemostatic agent (Grade 1B). In patientswho are undergoing minor dental procedures andare receiving aspirin, we recommend continuingaspirin around the time of the procedure (Grade1C). In patients who are undergoing minor dentalprocedures and are receiving clopidogrel, pleaserefer to the recommendations outlined in Section4.5 and Section 4.6.5.2. In patients who are undergoing minor der-matologic procedures and are receiving VKAs, werecommend continuing VKAs around the time ofthe procedure (Grade 1C). In patients who are




undergoing minor dermatologic procedures andare receiving aspirin, we recommend continuingaspirin around the time of the procedure (Grade1C). In patients who are undergoing minor der-matologic procedures and are receiving clopi-dogrel, please refer to the recommendations out-lined in Section 4.5 and Section 4.6.5.3. In patients who are undergoing cataract re-moval and are receiving VKAs, we recommendcontinuing VKAs around the time of the proce-dure (Grade 1C). In patients who are undergoingcataract removal and are receiving aspirin, werecommend continuing aspirin around the timeof the procedure (Grade 1C). In patients who areundergoing cataract removal and are receivingclopidogrel, please refer to the recommendationsoutlined in Section 4.5 and Section 4.6.

6.0 Perioperative Management of AntithromboticTherapy Patients Who Require Urgent Surgical orOther Invasive Procedures

6.1. In patients who are receiving VKAs andrequire reversal of the anticoagulant effect for anurgent surgical or other invasive procedure, werecommend treatment with low-dose (2.5 to 5.0mg) IV or oral vitamin K (Grade 1C). For moreimmediate reversal of the anticoagulant effect,we suggest treatment with fresh-frozen plasma oranother prothrombin concentrate in addition tolow-dose IV or oral vitamin K (Grade 2C).6.2. For patients receiving aspirin, clopi-dogrel, or both, are undergoing surgery, andhave excessive or life-threatening periopera-tive bleeding, we suggest transfusion of plate-lets or administration of other prohemostaticagents (Grade 2C).

T he perioperative management of patients whorequire temporary interruption of vitamin K

antagonists (VKAs) or antiplatelet drugs because of asurgical or other noninvasive procedure is a commonand challenging clinical problem.1 Approximately250,000 such patients are being assessed annually inNorth America alone, which is based on an estimateof the prevalence of patients who are receivinglong-term treatment with a VKA, principally due toatrial fibrillation or a mechanical heart valve (�2.5million),2,3 and an estimate of the annual proportionof patients who are receiving a VKA and requiresurgery or an invasive procedure (�10%).4 Themanagement of these patients is challenging becausethe risk of a thromboemblic event during interrup-tion of VKA or antiplatelet therapy needs to bebalanced against the risk for bleeding when anti-thrombotic therapy is administered in close proxim-

ity to surgery or an invasive procedure. In assessingpatients who are receiving antithrombotic therapyand are undergoing a surgical or other procedure,two principal issues should be addressed:

Is interruption of antithrombotic therapy in theperioperative period needed? In patients who areundergoing a major surgical or invasive procedure,interruption of antithrombotic therapy is typicallyrequired to minimize the risk for perioperativebleeding. Continuation of VKA or aspirin therapy inthe perioperative period confers an increased risk forbleeding.5–9 On the other hand, in patients who areundergoing minor surgical or invasive procedures,such as dental, skin, or eye procedures, interruptionof antithrombotic therapy may not be required.

If antithrombotic therapy is interrupted, is bridg-ing anticoagulation needed? In the context of peri-operative anticoagulation, bridging anticoagulationmay be defined as the administration of a short-acting anticoagulant, such as subcutaneous (SC)low-molecular-weight heparin (LMWH) or IV un-fractionated heparin (UFH), administered typicallyas a therapeutic-dose regimen for approximately 8 to10 days during interruption of VKA therapy whenthe international normalized ratio (INR) is notwithin a therapeutic range. In patients who are re-ceiving antiplatelet drugs alone, bridging anticoagula-tion with LMWH or UFH is, typically, not adminis-tered. In general, the need for bridging anticoagulationis determined by patient risk for thromboembolismduring interruption of antithrombotic therapy. Thetherapeutic aim of bridging anticoagulation is tominimize the time patients are not receiving antico-agulant therapy, thereby minimizing the risk forthromboembolism, and to do this in a manner thatminimizes patients’ risk for perioperative bleeding.

The objective of this article is to provide treatmentrecommendations for patients who are receiving aVKA or antiplatelet drug and may require temporaryinterruption of treatment because of a surgical orother invasive procedure. Following a discussion ofgeneral management principles (Section 1), thisreview addresses the following patient groups as-sessed in clinical practice: patients who are receivingVKAs (Section 2); patients who are receiving bridg-ing anticoagulation after interruption of VKAs (Sec-tion 3); patients who are receiving antiplatelet drugs(Section 4); patients who are receiving VKAs orantiplatelet drugs and are undergoing minor surgicalor invasive procedures (Section 5); and patients whorequire urgent interruption of antithrombotic ther-apy (Section 6). The summary recommendationsfollow the format in Table 1, which frames thequestions this article addresses.

At this juncture, some qualifying and explanatoryremarks are warranted. First, research in periopera-




Table 1—Perioperative Antithrombotic Therapy: Question Definition and Eligibility Criteria

Section PopulationIntervention or

Exposure/Comparison* OutcomesAvailable

MethodologyExclusionCriteria

2.0 Perioperative management of patients who are receiving VKAs2.1 Any patient receiving VKA and

having elective surgeryTiming of interruption of VKA before

surgeryHemostasis at time of

surgery (INR)Observational

studiesNone

2.2 Timing of resumption of VKA orLMWH after surgery

Hemostasis at time ofsurgery (bleed time)

Observationalstudies

None

2.3 INR testing to monitor anticoagulanteffect of VKAs before and aftersurgery vs no testing

Hemostasis at time ofsurgery (APTT,anti-factor Xa)


None

2.4.1 Patients with a mechanical heartvalve having elective surgery

Temporary interruption of VKA andbridging anticoagulation withLMWH/UFH vs no bridging

Stroke, other systemicembolism, majorhemorrhage


Not receivingVKA

2.4.2 Patients with chronic atrialfibrillation having electivesurgery




Not receivingVKA

2.4.3 Patients with VTE havingelective surgery




Not receivingVKA

3.0 Perioperative management of patients who are receiving bridging anticoagulation3.1 Costs of bridging anticoagulation Outpatient SC LMWH vs inpatient

IV UFHHealth-care system

costsObservational

studiesNone

3.2 Any patient receiving UFH orLMWH as bridginganticoagulation and havingelective surgery

Timing of interruption of LMWH orUFH before surgery

Major postoperativebleeding


None

3.3 Timing of resumption of LMWH orUFH after surgery



None

3.4 APTT, and anti-factor Xa testing tomonitor anticoagulant effect ofLMWH and UFH before and aftersurgery



None

4.0 Perioperative management of patients who are receiving antiplatelet therapy4.2 Any patient receiving an

antiplatelet drug and havingelective surgery

Timing of interruption of antiplateletdrugs before surgery

INR before and aftersurgery


None

4.3 Timing of resumption of antiplateletdrugs after surgery

Platelet function assaytesting before andafter surgery


None

4.4 Platelet function assay testing tomonitor antiplatelet effect ofantiplatelet drugs before and aftersurgery vs no testing

APTT and anti-factorXa before and aftersurgery


None

4.5 Any patient receiving antiplateletdrug and having noncardiacsurgery, cardiac surgery, orPCI

Temporary interruption vscontinuation of antiplatelet drugs

Myocardial ischemia,major hemorrhage

Randomizedcontrolled trials,observationalstudies

ReceivingVKA

4.6 Any patient with a coronary stentreceiving antiplatelet drug andhaving noncardiac surgery,cardiac surgery, or PCI

Temporary interruption vscontinuation of antiplatelet drugs



ReceivingVKA

5.0 Perioperative management of antithrombotic therapy in patients who require minor procedures5.1 Any patient receiving VKA or

antiplatelet drug and having aminor dental procedure

Temporary interruption vscontinuation of VKA or antiplatelettherapy

Arterial or VTE, majorhemorrhage


None

5.2 Any patient receiving VKA orantiplatelet drug and having aminor skin procedure




None

5.3 Any patient receiving VKA orantiplatelet drug and having aminor eye procedure




None




tive antithrombotic therapy is an emerging field,with a relative paucity of randomized trials and apreponderance of observational studies assessingperioperative management strategies. Consequently,comparisons of certain management strategies (eg,bridging vs no bridging) are lacking whereas com-parisons of other strategies (eg, continuation vsinterruption of VKAs for minor procedures) arebetter developed. Second, as many of the pertinentobservational studies are based on small patientsamples, they may be underpowered to determine ifa management approach is efficacious (ie, associatedwith a low risk for thromboembolism) or safe (ie,associated with a low risk for bleeding). Such studiesshould, therefore, be interpreted with caution bothbecause of the observational study design and thesmall sample size. Third, it should be acknowledgedthat there is no standardized definition of “bridginganticoagulation.” Although it may be defined as theadministration of a therapeutic-dose regimen of SCLMWH or IV UFH during interruption of a VKA,bridging anticoagulation may also include a regimen oflow-dose SC LMWH. Ultimately, the perioperativeanticoagulant regimen used will depend, to a largeextent, on patient clinical characteristics and the type ofsurgery or procedure they are undergoing.

1.0 Perioperative Management ofAntithrombotic Therapy: General

Principles

1.1 Assessment of Thromboembolic Risk AfterInterruption of Antithrombotic Therapy

Interruption of antithrombotic therapy exposespatients to an increased risk for thromboembolicevents, such as stroke or mechanical valve thrombo-sis, with the risk varying depending on the indicationfor antithrombotic therapy and the presence of

comorbid conditions.10 These events can have dev-astating clinical consequences: embolic stroke, whichcan result in major disability or death in 70% ofpatients11,12; thrombosis of a mechanical heart valve,which is fatal in 15% of patients13; and perioperativemyocardial ischemia, which is associated with a two-fold- to fourfold-increased risk for death.14,15 Similarly,interruption of antiplatelet drugs in patients with asirolimus or paclitaxel drug-eluting coronary stent, es-pecially within 6 months of stent placement, signifi-cantly increases the risk for intracoronary stent throm-bosis and myocardial infarction.16,17

Stratifying patients according to their risk forperioperative thromboembolism is based on patients’clinical indication for antithrombotic therapy and thepresence of comorbidities. Although there is novalidated risk stratification of such patients, theapproach we have used in these guidelines is toseparate patients into a high-risk, moderate-risk, orlow-risk group according to their indication for anti-thrombotic therapy (Table 2).

1.2 Assessment of Bleeding Risk Associated WithSurgery or Other Invasive Procedures

The administration of antithrombotic therapy inthe perioperative period should be done in a way thatconsiders the risk for bleeding associated with thesurgery or procedure. Although bleeding is a treat-able perioperative complication, there is emergingevidence that the clinical impact of bleeding isconsiderable and, perhaps, greater than previouslyappreciated.18–20 Furthermore, postoperative bleed-ing delays the resumption of antithrombotic therapy,with the potential to further expose patients to anincreased risk for thromboembolism.21,22

Stratifying patients according to their risk forperioperative bleeding can be based on the risk forbleeding associated with the surgery or procedure

Table 1—Continued

Section Population Intervention or Exposure/Comparison OutcomesAvailable

MethodologyExclusionCriteria

6.0 Perioperative management of antithrombotic therapy patients who require urgent surgical or other invasive procedures6.1 Any patient receiving VKA and

having urgent surgeryVitamin K via different routes (IV vs

oral/SC) to reverse anticoagulanteffect of VKAs; blood products(FFP, PC) to reverse anticoagulanteffect of VKAs vs no bloodproducts

(1) Hemostasis at timeof surgery (INR)


None

(2) Major bleedingand surrogate

6.2 Any patient receiving anantiplatelet drug and havingurgent surgery

DDAVP/prohemostatic drugs andplatelet transfusion to reverseantiplatelet effect of antiplateletdrugs vs no DDAVP/prohemostaticdrugs and platelet transfusions



None

*FFP � fresh frozen plasma; PC � prothrombin concentrate.




and should be coupled with an assessment of post-operative hemostasis.21–24 Although there is no vali-dated method that quantifies perioperative bleedingrisk, special attention is warranted for certain surgi-cal or other invasive procedures associated with ahigh risk for bleeding. In such patients, postoperativeantithrombotic therapy should be administered withcaution, especially therapeutic-dose LMWH or UFHwhen used as bridging anticoagulation. Surgical andother invasive procedures associated with a high bleed-ing risk include: coronary artery bypass or heart valvereplacement surgery25,26; intracranial or spinal sur-gery27; aortic aneurysm repair, peripheral artery bypass,and other major vascular surgery; major orthopedicsurgery, such as hip or knee replacement28; reconstruc-tive plastic surgery29; major cancer surgery; and pros-tate and bladder surgery.30,31

In addition, clinicians should note procedures that,on the surface, may appear to be associated with alow risk for bleeding but in which perioperativeanticoagulation should be undertaken with caution.Such procedures include: resection of colonic polyps,especially sessile polyps � 2 cm in diameter, inwhich bleeding may occur at the transected stalk32;biopsy of the prostate or kidney, in which thepresence of highly vascular tissue and endogenousurokinase may promote post-biopsy bleeding33; andcardiac pacemaker or defibrillator implantation, inwhich separation of the infraclavicular fascial layersand lack of cautery or suturing of unopposed tissueswithin the pacemaker or defibrillator pocket maypredispose to pocket hematoma development.34

1.3 Balancing Thromboembolic Risk and BleedingRisk

Inherent in perioperative antithrombotic manage-ment is the need for individualized patient manage-ment that balances individual risk for thromboem-bolism and bleeding. In patients classified as “highrisk for stroke or thromboembolism,” the need toprevent a thromboembolic event such as embolicstroke or intracoronary stent thrombosis will domi-nate perioperative antithrombotic management, ir-respective of bleeding risk. In such patients, thepotential clinical consequences of such events, whichmay be fatal or may cause permanent disability, will,in most patients, outweigh the potential clinicalconsequences of bleeding and will justify the needfor bridging anticoagulation or perioperative contin-uation of antithrombotic therapy. This approach, ifadopted, should nonetheless consider judicious useof postoperative bridging anticoagulation (ie, as out-lined in Section 3.0) and optimizing intraoperativehemostasis (ie, cautery and other local measures),with the intent of simultaneously minimizing thepotential for major surgical bleeding that wouldhave the undesired effect of delaying the resump-tion of or necessitating interruption of antithrom-botic therapy.

In patients classified as “moderate risk for throm-boembolism,” a single perioperative antithromboticstrategy will not be dominant and management willdepend more on an individual patient risk assess-ment. Thus, in patients at moderate risk for thrombo-

Table 2—Suggested Patient Risk Stratification for Perioperative Arterial or Venous Thromboembolism

Risk Stratum

Indication for VKA Therapy

Mechanical Heart Valve Atrial Fibrillation VTE

High Any mitral valve prosthesis CHADS2 score of 5 or 6 Recent (within 3 mo) VTEOlder (caged-ball or tilting disc)

aortic valve prosthesisRecent (within 3 mo) stroke or

transient ischemic attack,Severe thrombophilia (eg, deficiency

of protein C, protein S orantithrombin, antiphospholipidantibodies, or multipleabnormalities)

Recent (within 6 mo) stroke ortransient ischemic attack

Rheumatic valvular heartdisease

Moderate Bileaflet aortic valve prosthesis andone of the following: atrialfibrillation, prior stroke ortransient ischemic attack,hypertension, diabetes,congestive heart failure, age� 75 yr

CHADS2 score of 3 or 4 VTE within the past 3 to 12 moNonsevere thrombophilic conditions

(eg, heterozygous factor V Leidenmutation, heterozygous factor IImutation)

Recurrent VTEActive cancer (treated within 6 mo

or palliative)

Low Bileaflet aortic valve prosthesiswithout atrial fibrillation and noother risk factors for stroke

CHADS2 score of 0 to 2 (andno prior stroke or transientischemic attack)

Single VTE occurred � 12 mo agoand no other risk factors

*CHADS2 � Congestive heart failure-Hypertension-Age-Diabetes-Stroke.




embolism, the need to prevent thromboembolism willhave less dominance than in “high-risk” patients andbridging anticoagulation may incorporate a modified,less aggressive, approach postoperatively in patientsundergoing surgery or a procedure associated with ahigh bleeding risk. In patients classified as “low risk forthromboembolism,” the need to prevent thromboem-bolism will have even less dominance and cliniciansmay avoid bridging anticoagulation altogether; if given,bridging should be curtailed postoperatively in suchpatients with a high bleeding risk.

1.4 Perioperative Antithrombotic Management:Practical Considerations

In managing antithrombotic therapy before andafter surgery or a procedure, clinicians should notethe following practical management considerations:

• For patients undergoing a major surgical or inva-sive procedure, if the intent is to eliminate anyeffect of antithrombotic therapy, it should bestopped at a time before the procedure (eg, approx-imately 5 days in patients receiving a VKA and 7 to10 days in patients receiving an antiplatelet drug)that leaves minimal or no residual antithromboticeffect at time of the procedure; doing so will mini-mize the risk for intraprocedural bleeding.

• The administration of a rapidly acting anticoagu-lant, such as LMWH or UFH, after surgery oranother invasive procedure increases the risk forbleeding. This risk is dependent on the dose ofanticoagulant (eg, therapeutic-dose more thanlow-dose) and the proximity to surgery that it isadministered (higher bleeding risk when adminis-tered closer to surgery). Delaying resumption of atherapeutic-dose LMWH or UFH regimen (for 48to 72 h after surgery), decreasing the dose ofLMWH or UFH (to a low-dose regimen), oravoiding its use altogether in the postoperativeperiod can mitigate the risk for bleeding.

• For perioperative anticoagulant dosing, althoughthere is evidence that low-dose (prophylactic-dose) LMWH or UFH (eg, dalteparin 5,000 IU qdor UFH 5,000 IU bid) is effective in preventingvenous thromboembolism (VTE), evidence is lack-ing that such low-dose treatment is effective inpreventing arterial thromboembolism.

• In resuming antithrombotic therapy after a surgi-cal or invasive procedure, it takes 2 to 3 days for ananticoagulant effect to begin after the start ofwarfarin,35 it takes 3 to 5 h for a peak anticoagulanteffect to be reached after the start of LMWH,36

whereas it takes minutes for an antiplatelet effectto begin after the start of aspirin37 and 3 to 7 days

for peak inhibition of platelet aggregation to bereached after the start of a (75 mg) maintenancedose of clopidogrel.38

• The majority of surgical or other invasive pro-cedures are being done without hospitalizationor with a short hospital stay; consequently,potential thromboembolic- or bleeding-relatedcomplications are likely to occur while the pa-tient is at home, especially during the initial 2weeks after a procedure.21,22,39,40 Close follow-up ofpatients during the early period after a procedure is,therefore, warranted to allow early detection andexpedited treatment of potential complications.

2.0 Perioperative Management of PatientsWho Are Receiving VKAs

Long-term VKA therapy is widely used for theprimary and secondary prevention of arterial thrombo-embolism and VTE for a wide spectrum of clinicalindications that include atrial fibrillation, mechanicalheart valve placement, VTE, coronary and peripheralartery disease, dilated cardiomyopathy, and primarypulmonary hypertension. In Section 2.0, we will focuson the perioperative anticoagulant management ofpatients with the most common clinical indications forlong-term VKA therapy: mechanical heart valve,chronic atrial fibrillation, and VTE.

2.1 Interruption of VKAs Before Surgery

In patients undergoing major surgery, interruptionof VKAs is generally required to minimize the riskfor perioperative bleeding,5–7 whereas in patientsundergoing certain minor surgical or other proce-dures, some of which are discussed in Section 5.0,interruption of VKAs may not be required. To ourknowledge, there are no randomized trials compar-ing interruption of VKAs vs no interruption or partialinterruption of VKAs before major surgery. Threeobservational studies have assessed continuation8,41

or partial interruption42 of VKAs in patients under-going surgery, with suggestive but not definitivefindings that continuation of VKAs increases the riskfor perioperative bleeding. In one retrospective co-hort study involving 603 VKA-treated patients whounderwent surgery, the majority of whom did nothave interruption of VKA therapy prior to surgery,the incidence of perioperative major bleeding was9.5% (95% confidence interval [CI]: 7.1–12.1),which is high; moreover, compared to patients withan INR � 2.0, patients with an INR � 3.0 appearedto be at higher risk for bleeding complications (oddsratio [OR], 1.6; 95% CI: 0.4–4.0).8 Another retro-spective study assessed 100 patients who underwentsurgery (58 had major surgery) and who had partial




interruption of VKA, with a mean INR of 1.8 (range:1.2 to 4.9) on the day of surgery; in this study, onlytwo (2%) patients had major bleeding although 34(34%) patients required a blood transfusion.42

For patients who are receiving VKA therapy withwarfarin, which has a half-life of 36 to 42 h, treat-ment should be interrupted approximately 5 daysbefore surgery (corresponding approximately to 5half-lives of warfarin) to ensure there is no (orminimal) residual anticoagulant effect remaining bythe time of surgery.43,44 Previous prospective cohortstudies assessing standardized perioperative antico-agulation regimens interrupted warfarin 5 to 6 daysbefore surgery.21,22,39 In one of these studies,22 inwhich warfarin was stopped 5 days before surgeryand the INR was routinely measured on the daybefore surgery, only 15 of 224 (7%) patients had anelevated INR (� 1.5) on the day before surgery,which was corrected with low-dose (1 mg) oralvitamin K. A longer (eg, � 5 days) duration of VKAinterruption to attain a normalized INR by the timeof surgery may be required in patients with amechanical heart valve, who have a higher targetedINR range, which is typically 2.5 to 3.5.45 In addi-tion, advanced age may be associated with a prolon-gation in the decay of the anticoagulant effect ofwarfarin after its interruption. Thus, in a retrospec-tive cohort study of 633 patients who had excessiveanticoagulation (INR � 6.0), increasing age, in 10-year increments, conferred an increased likelihoodfor delayed normalization of the INR after warfarinwas stopped (hazard ratio, 1.18; CI: 1.01–1.38).46

For a minority of patients who are receiving VKAtherapy with phenprocoumon, in whom the VKA-related recommendations in this article would notapply, treatment should be interrupted approxi-mately 10 days before surgery based on the half-lifeof phenprocoumon of 96 to 140 h.47 Some investi-gators have suggested that, in selected patients,warfarin can be interrupted 2 to 3 days beforesurgery to aim for an INR of 1.5 to 1.9 at the time ofsurgery; however, the feasibility and safety of thisapproach remain uncertain.8,42,48

Recommendation

2.1. In patients who require temporary inter-ruption of a VKA before surgery or a proce-dure and require normalization of the INRfor the surgery or procedure, we recommendstopping VKAs approximately 5 days beforesurgery over a shorter time interval to allowadequate time for the INR to normalize(Grade 1B).

2.2 Resumption of VKAs After Surgery

When resuming VKAs after surgery, approxi-mately 48 h is required to attain a partial anticoag-ulant effect, with an INR � 1.5.43 Consequently, thepotential effect of VKAs to promote postoperativebleeding is likely to be mitigated by the delayedonset of their anticoagulant activity. It is reasonable,therefore, to resume VKA therapy on the evening ofthe day of surgery or the next day, with an antici-pated partial anticoagulant effect to occur 48 h later.In one study of 650 patients who resumed VKA afterbridging anticoagulation, with a dose correspondingto patients’ usual dose, the mean duration to achievea therapeutic INR was 5.1 days (SD: 1.1).21 Inanother study of 224 patients who resumed VKAafter bridging anticoagulation, with doubling of pa-tients’ usual dose for the initial 2 days after VKAresumption, the mean duration to achieve a thera-peutic INR was 4.6 days (range: 0 to 10).22 Oneretrospective cohort study involving 100 patientswho received bridging with LMWH found a longerthan expected time to attain a therapeutic INR aftersurgery, which was a median of 7.5 days (inter-quartile range: 4.3 to 13.0) after warfarin wasresumed postoperatively.49 The delay in attainingtherapeutic levels of anticoagulation in this studywas possibly related to suboptimal INR monitoringafter surgery.

Recommendation

2.2. In patients who have had temporary inter-ruption of a VKA before surgery or a proce-dure, we recommend resuming VKAs approxi-mately 12 to 24 h (the evening of or the nextmorning) after surgery and when there is ade-quate hemostasis over resumption of VKAscloser to surgery (Grade 1C).

2.3 Laboratory Monitoring of VKA Therapy

Perioperative monitoring of the INR in patientswho are receiving VKA therapy is predicated onseveral factors, which include the feasibility of INRmonitoring prior to surgery and the time periodbetween interruption of VKAs and surgery. In thepreoperative period, it is reasonable to have INRtesting done at least once before surgery (preferably1 to 2 days before surgery) to confirm a normal ornear-normal INR and, in patients with an elevatedINR (eg, INR � 1.5), to administer low-dose vitaminK. Administration of vitamin K at this time will avoidthe need to administer plasma or other blood prod-ucts by ensuring the INR has normalized by the dayof surgery. In one retrospective cohort study involv-ing 43 patients who required temporary interruption




of VKA and had an INR between 1.5 and 1.9 (mean:1.6) on the day before surgery, administering 1 mgoral vitamin K resulted in 91% of patients having anormal or near normal INR (ie, � 1.4) on the day ofsurgery.50 This study also suggested that preopera-tive administration of low-dose vitamin K does notappear to confer resistance to re-anticoagulationwhen a VKA is resumed after surgery. In the post-operative period, INR testing can be done to approx-imate when therapeutic anticoagulation is (or willbe) attained and, therefore, when LMWH or UFHcan be stopped for patients who have been receivingbridging anticoagulation.

Recommendation

2.3. In patients who require temporary inter-ruption of a VKA before surgery or a procedureand whose INR is still elevated (ie, > 1.5) 1 to 2days before surgery, we suggest administeringlow-dose (ie, 1 to 2 mg) oral vitamin K tonormalize the INR instead of not administeringvitamin K (Grade 2C).

2.4 Patient Risk Stratification and Assessing Needfor Bridging Anticoagulation

In assessing patients who are receiving VKAs forthe three principal indications, a mechanical heartvalve, chronic atrial fibrillation, or VTE, periopera-tive anticoagulant management (and need for bridg-ing) will be driven to a large extent by patients’ riskfor developing thromboembolism, either arterial orvenous, in the perioperative period. In this section,we have attempted to provide a reasonable, thoughlargely empiric, stratification of patients according totheir risk for thromboembolism (high, moderate,low) while acknowledging that comparative data onrisks for perioperative thromboembolism for theproposed risk strata are limited. This risk stratifica-tion scheme may be combined with individual pa-tient factors, which may include prior thromboem-bolism during VKA interruption or a prior embolicstroke, to determine the overall risk for thromboem-bolism (and need for bridging).

2.4.1 Patients With a Mechanical Prosthetic HeartValve

Risk Stratification: Patients with mechanical heartvalves are at increased risk for arterial thromboem-bolism, which includes stroke, systemic embolism,and valvular or intracardiac thrombosis. Risk strati-fication for patients with a mechanical heart valve isbased on studies that have assessed the risk forarterial thromboembolism during anticoagulant ther-apy and on older studies that assessed thromboem-

bolic risk while patients were receiving either noantithrombotic therapy or treatment that is currentlyconsidered suboptimal.51–53 What is lacking, how-ever, are estimates of the risk for thromboembolismin patients who have modern (bileaflet) prosthesesand have not received antithrombotic therapy overan extended time period. As trials that include suchpatients are lacking and are unlikely to be per-formed, clinicians can use the risk classificationproposed here as a general guide for patient man-agement.

Patients at high risk for arterial thromboembolism(� 10%/yr) may include those with one or more ofthe following: (1) a mitral valve prosthesis; (2) anolder-generation (caged-ball or tilting disk) aorticvalve prosthesis; and (3) a recent (within 6 months)stroke or transient ischemic attack. Patients at mod-erate risk for thromboembolism (4 to 10%/yr) mayinclude those with a bileaflet aortic valve prosthesisand one of the following: (1) atrial fibrillation; (2)prior stroke or transient ischemic attack; and (3)other stroke risk factors (hypertension, diabetes,congestive heart failure, age � 75 years). Patients atlow risk for thromboembolism (� 4%/yr) may in-clude those with a bileaflet aortic valve prosthesiswithout atrial fibrillation and who do not have otherrisk factors for stroke.

Assessing Need for Bridging Anticoagulation: In14 prospective cohort studies, bridging anticoagula-tion was assessed in approximately 1,300 patientswith a mechanical heart valve.7,21,22,40,54–64 As shownin Table 3, investigators studied predominantly ther-apeutic-dose LMWH regimens, although two studiesinvolving a total of 118 patients also assessed low-dose LMWH regimens.54,64 The issue of whether alow-dose anticoagulant regimen is efficacious for theprevention of arterial thromboembolism in patientswith a mechanical heart valve, as it is for preventingVTE,65 cannot be definitively addressed based on thelimited available evidence. It is probable that a moreintense, therapeutic-dose, anticoagulant regimenwould be required to prevent valve thrombosis andvalve-related systemic embolism during VKA inter-ruption and, until further evidence to the contrarybecomes available, is the preferred regimen. Theoverall crude risk for perioperative arterial thrombo-embolism was 0.83% (95% CI: 0.43–1.5). There wereno reported episodes of mechanical valve thrombosis.The interpretation of this finding is limited becausethere are no studies, to our knowledge, assessing therisk for arterial thromboembolism in (a comparatorgroup of) patients with a mechanical heart valve whohave VKA interruption for surgery but do not receivebridging anticoagulation.

Mathematical modeling can be used to estimate




Tab

le3—

Non

rand

omiz

edP

rosp

ecti

veC

ohor

tSt

udi

esA

sses

sing

Bri

dgin

gA

ntic

oagu

lati

onA

fter

Inte

rru

ptio

nof

VK

AT

hera

py:

Cli

nica

lD

escr

ipti

onan

dR

esu

lts

(Sec

tion

2.4)

Stud

y/yr

Patie

nts

No.

and

Typ

eof

Proc

edur

e

Bri

dgin

gA

ntic

oagu

latio

nR

egim

enF

ollo

w-u

pA

fter

Proc

edur

e

Clin

ical

Out

com

es,%

No.

Indi

catio

nfo

rV

KA

The

rapy

Art

eria

lThr

ombo

embo

lism

inPa

tient

sW

ithM

echa

nica

lHea

rtV

alve

/A

rter

ialT

hrom

boem

bolis

m

Rec

urre

ntV

TE

inPa

tient

sW

ithV

TE

Dea

thM

ajor

Ble

ed

Kat

holi

etal

7 /197

823

5M

echa

nica

lhea

rtva

lve

(typ

eno

tsp

ecifi

ed)

25su

rgic

alU

FH

:int

erm

itten

tor

cont

inuo

usin

fusi

onN

otsp

ecifi

ed0

Not

appl

icab

le0

221

nons

urgi

cal

Span

dorf

eret

al60

/19

9920

12m

echa

nica

lhea

rtva

lve

(10

aort

ic,

2m

itral

)10

surg

ical

Eno

xapa

rin:

1m

g/kg

bid

1m

o0 0 0

Not

appl

icab

le0

1

4at

rial

fibri

llatio

n4

VT

EG

alla

and

Fuh

s54/

2000

8888

mec

hani

calh

eart

valv

e(2

7ao

rtic

,50

mitr

al,9

aort

ic�

mitr

al,2

tric

uspi

d)

Eno

xapa

rin:

30m

gbi

d1

mo

0N

otap

plic

able

03

Nut

escu

etal

237 /

2001

*23

21is

chem

icst

roke

�hy

perc

oagu

labl

est

ate

18su

rgic

al7

nons

urgi

cal

Dal

tepa

rin:

100

IU/k

gbi

d3

mo

0N

otap

plic

able

00

Tin

mou

thet

al61

/20

0124

12m

echa

nica

lhea

rtva

lve

(7ao

rtic

,5

mitr

al);

6at

rial

fibri

llatio

n;6

VT

E

9su

rgic

al17

nons

urgi

cal

Dal

tepa

rin:

200

IU/k

gqd

1m

o1 0 0

01 0 0

0

Wils

onet

al62

/200

147

7m

echa

nica

lhea

rtva

lve

(typ

eno

tsp

ecifi

ed)

11at

rial

fibri

llatio

n26

VT

E3

othe

r(c

ardi

omyo

path

y)

15su

rgic

al32

nons

urgi

cal

Dal

tepa

rin:

200

IU/k

gqd

or12

0IU

/kg

bid

(5,0

00IU

qdin

9pa

tient

s)

Not

spec

ified

0 0 1 0

00

0

Fer

reir

aet

al56

/200

382

82m

echa

nica

lhea

rtva

lve

(43

aort

ic,

39m

itral

)53

surg

ical

29no

nsur

gica

lE

noxa

pari

n:1

mg/

kgbi

d(d

ose

adju

sted

for

rena

lim

pair

men

t)

3m

o0

Not

appl

icab

leN

otav

aila

ble

1

Bau

doet

al72

/200

741

134

4m

echa

nica

lhea

rtva

lve

orat

rial

fibri

llatio

n67

VT

E

77su

rgic

al33

4no

nsur

gica

lV

ario

usth

erap

eutic

-do

se(2

2%)

orpr

ophy

lact

ic-d

ose

(78%

)re

gim

ens

Not

spec

ified

2(1

/1)

00

1 07

Dou

ketis

etal

21/

2004

†65

013

4m

echa

nica

lhea

rtva

lve

(52

aort

ic,5

2m

itral

,30

aort

ic�

mitr

al)

416

atri

alfib

rilla

tion

251

surg

ical

399

nons

urgi

cal

Dal

tepa

rin:

100

IU/k

gbi

d0.

5m

o1 3

Not

appl

icab

le4

6

Ham

mer

stin

glet

al58

/200

711

6M

echa

nica

lhea

rtva

lve

(76

aort

icva

lve,

31m

itral

valv

e,9

aort

ican

dm

itral

valv

es)

46su

rgic

al70

nons

urgi

cal

Eno

xapa

rin:

1m

g/kg

qdor

bid

1m

o0

00

1




Tab

le3—

Con

tinu

ed

Stud

y/yr

Patie

nts

No.

and

Typ

eof

Proc

edur

e

Bri

dgin

gA

ntic

oagu

latio

nR

egim

enF

ollo

w-u

pA

fter

Proc

edur

e

Clin

ical

Out

com

es,%

No.

Indi

catio

nfo

rV

KA

The

rapy

Art

eria

lThr

ombo

embo

lism

inPa

tient

sW

ithM

echa

nica

lH

eart

Val

ve/A

rter

ial

Thr

ombo

embo

lism

Rec

urre

ntV

TE

inPa

tient

sW

ithV

TE

Dea

thM

ajor

Ble

ed

Kov

acs

etal

22/2

004‡

224

112

mec

hani

calh

eart

valv

e(t

ype

not

spec

ified

)11

2at

rial

fibri

llatio

n

67su

rgic

al15

7no

nsur

gica

lD

alte

pari

n:20

0IU

/kg

qd(5

,000

IUpo

stop

erat

ive

in35

patie

nts

athi

ghri

skfo

rbl

eedi

ng)

3m

o1 1

Not

appl

icab

le0

15

Con

stan

set

al64

/20

0798

30m

echa

nica

lhea

rtva

lve

(14

aort

ic,

16m

itral

)56

atri

alfib

rilla

tion

orar

teri

aldi

seas

e12

VT

E

98no

nsur

gica

lB

emip

arin

:3,5

00IU

qd3

mo

0 0 0

00

0

Tur

pie

and

Dou

ketis

40/2

004§

220

220

mec

hani

calh

eart

valv

e(1

65ao

rtic

,51

mitr

al,5

aort

ic�

mitr

al)

Not

spec

ified

Eno

xapa

rin:

1m

g/kg

bid

3m

o0

Not

appl

icab

le3

8

Jaff

eret

al57

/200

669

20m

echa

nica

lhea

rtva

lve

27at

rial

fibri

llatio

n18

VT

E4

othe

rar

teri

alin

dica

tions

18su

rgic

al47

nons

urgi

cal

Eno

xapa

rin:

1m

g/kg

bid

(30

mg

bid

post

oper

ativ

eaf

ter

surg

ical

proc

edur

es)

1m

o0 0

00

2

Spyr

opou

los

etal

55/

2006

901

246

mec

hani

calh

eart

valv

e34

9at

rial

fibri

llatio

n23

0V

TE

76ot

her

arte

rial

indi

catio

ns

394

surg

ical

507

nons

urgi

cal

The

rape

utic

-dos

e(7

5%)

and

prop

hyla

ctic

-dos

e(2

5%)

UF

Hor

LM

WH

regi

men

s

1m

o8�

2�6

31

Dun

net

al39

/200

626

017

6at

rial

fibri

llatio

n81

VT

E10

5su

rgic

al14

5no

nsur

gica

lE

noxa

pari

n:1.

5m

g/kg

qd1

mo

41

28

Mal

ato

etal

63/2

006

228

53m

echa

nica

lhea

rtva

lve

139

atri

alfib

rilla

tion

26V

TE

10ot

her

arte

rial

indi

catio

ns

101

surg

ical

127

nons

urgi

cal

The

rape

utic

-dos

e(4

0%)

orpr

ophy

lact

ic-d

ose

(60%

)L

MW

Hre

gim

ens

0 3 0

10

6




the perioperative risk of arterial thromboembolism ifbridging anticoagulation is not given based on theprorated fraction of the annual risk of this out-come.66 Thus, the risk of thromboembolism in apatient with mechanical (mitral or aortic) heart valvewho is not treated with a VKA is estimated at 0.046%/d(ie, 17% annual risk67 � 365) or �0.4% for 8 days whenpatients are not therapeutically anticoagulated duringVKA interruption. The finding of a higher rate ofperioperative thromboembolism in studies of bridginganticoagulation compared to that derived from mathe-matical modeling suggests that the risk for thrombo-embolism is higher than expected. What remainsunclear is whether the administration of bridginganticoagulation decreases this rate further thanthat which would be observed if bridging had notbeen administered or whether bridging therapyhas no effect on the perioperative risk for arterialthromboembolism. This issue can only be ad-dressed through randomized trials comparing abridging vs no bridging strategy in patients with amechanical heart valve, which poses challenges interms of feasibility. In the meantime, cliniciansshould consider bridging anticoagulation in pa-tients with a mechanical prosthetic heart valvewho are at high or moderate risk for arterialthromboembolism (ie, stroke or valve thrombosis).

2.4.2 Patients With Chronic Atrial Fibrillation

Risk Stratification: Risk stratification in patientswith chronic atrial fibrillation is based on placebo-controlled randomized trials that assessed differentantithrombotic strategies in patients with nonvalvu-lar atrial fibrillation.68 Patients with rheumatic val-vular heart disease were not included in these trialsbut are considered to be at high risk for stroke.Bridging anticoagulation should be considered inselected patients with chronic atrial fibrillation whoare at high or moderate risk for arterial thromboem-bolism (ie, stroke or systemic embolism).3,69–71

Clinical prediction rules, such as the Congestive HeartFailure-Hypertension-Age-Diabetes-Stroke (CHADS2)score, may help to stratify patients with nonvalvularatrial fibrillation according to their risk for stroke.70

An administrative linked database study4 suggestedthat the CHADS2 score could be applied to theperioperative setting to estimate stroke risk in pa-tients with chronic atrial fibrillation who were un-dergoing surgery. The score ranges from 0 to 6 andis based on the whether any of five risk factors arepresent: congestive heart failure, hypertension, dia-betes, age � 75 years (1 point each); and prior strokeor transient ischemic attack (2 points). Patients athigh risk for arterial thromboembolism (ie, � 10%risk per year) may include those with one or more of

Tab

le3—

Con

tinu

ed

Stud

y/yr

Patie

nts

No.

and

Typ

eof

Proc

edur

e

Bri

dgin

gA

ntic

oagu

latio

nR

egim

enF

ollo

w-u

pA

fter

Proc

edur

e

Clin

ical

Out

com

es,%

No.

Indi

catio

nfo

rV

KA

The

rapy

Art

eria

lThr

ombo

embo

lism

inPa

tient

sW

ithM

echa

nica

lHea

rtV

alve

/A

rter

ialT

hrom

boem

bolis

m

Rec

urre

ntV

TE

inPa

tient

sW

ithV

TE

Dea

thM

ajor

Ble

ed

Hal

britt

eret

al81

/20

07¶

311

55m

echa

nica

lhea

rtva

lve

(29

aort

ic,

26m

itral

)12

4at

rial

fibri

llatio

n50

LV

dysf

unct

ion

59V

TE

23ot

her

(not

spec

ified

)

65su

rgic

al24

6no

nsur

gica

lT

hera

peut

ic-d

ose

LM

WH

orU

FH

in62

%of

mec

hani

cal

hear

tva

lve,

47%

ofat

rial

fibri

llatio

n,an

d55

%of

left

vent

ricu

lar

dysf

unct

ion

13

47

*Bri

dgin

gep

isod

esin

21pa

tient

s.†1

10pa

tient

sco

nsid

ered

high

risk

for

post

oper

ativ

ebl

eedi

ngdi

dno

tre

ceiv

epo

stop

erat

ive

LM

WH

.‡F

ive

patie

nts

had

intr

aope

rativ

eor

post

oper

ativ

em

yoca

rdia

linf

arct

ion

butw

ere

noti

nclu

ded

inT

able

3to

faci

litat

eac

ross

-stu

dyco

mpa

riso

ns,a

sot

her

stud

ies

did

notd

ocum

entp

erio

pera

tive

myo

card

ial

isch

emic

outc

omes

.§F

orty

patie

nts

cons

ider

edhi

ghri

skfo

rpo

stop

erat

ive

blee

ding

did

not

rece

ive

post

oper

ativ

eL

MW

H.

�Pat

ient

grou

p(a

ccor

ding

toin

dica

tion

for

VK

A)

that

outc

ome

even

tsoc

curr

edin

not

spec

ified

.¶3

11br

idgi

ngep

isod

esin

268

patie

nts.




the following: (1) CHADS2 score of 5 or 6; (2) arecent (within 3 months) stroke or transient ischemicattack; or (3) rheumatic valvular heart disease. Pa-tients at moderate risk for thromboembolism (ie, 5 to10% risk per year) include those with a CHADS2score of 3 or 4, whereas patients at low risk forthromboembolism (ie, � 5% risk per year) includethose with a CHADS2 score of 0 to 2 who have nothad a prior stroke or transient ischemic attack.

Assessing Need for Bridging Anticoagulation:In 10 prospective cohort studies, bridging anti-coagulation has been assessed in approximately1,400 patients with chronic atrial fibrilla-tion.21,22,39,55,57,58,60–64 As outlined in Table 3, inves-tigators studied predominantly therapeutic-doseLMWH regimens, although low-dose LMWH regi-mens were also assessed in 4 studies involving ap-proximately 300 patients (exact number not discern-able from published data).56,63,64,72 As in patientswith a mechanical heart valve, the issue of whetherlow-dose anticoagulant regimens are efficacious toprevent arterial thromboembolism is also pertinentto patients with atrial fibrillation. Similar to patientswith a mechanical heart valve, there is inadequatedata to formulate definitive conclusions. It is proba-ble that a more intense, therapeutic-dose, anticoag-ulation regimen is more efficacious to prevent em-bolic stroke and systemic embolism than a low-doseregimen and, until evidence to the contrary is avail-able, it is the preferred management. The overallcrude risk for perioperative arterial thromboembo-lism in patients who received bridging anticoagula-tion was 0.57% (95% CI: 0.26–1.1). In studies thatdescribed the clinical characteristics of such patients,most patients had at least one additional risk factorfor stroke (ie, prior stroke, ventricular dysfunction,hypertension, diabetes, age � 75 years).

There are emerging data assessing the risk forarterial thromboembolism in patients with atrialfibrillation who do not receive bridging anticoag-ulation. In a community-based prospective cohortstudy (Anticoagulation Consortium to ImproveOutcomes Nationally [ACTION]) involving pa-tients who were receiving a VKA, 726 patients withatrial fibrillation had temporary interruption of aVKA and did not receive bridging.73 Four (0.6%)patients developed arterial thromboembolism (2strokes, 1 transient ischemic attack, 1 systemicembolism) during a 1-month follow-up periodafter surgery. In a retrospective cohort studyassessing 690 patients (�90% with atrial fibrilla-tion) who required temporary interruption of VKAtherapy prior to GI endoscopy, there were 11(1.6%) patients who developed a stroke within 1month of the procedure (A. Jaffer, submitted for

publication). Another study examined a linkedadministrative database of patients dischargedfrom hospital after surgery or an invasive proce-dure between 1996 and 2001, during a time periodwhen bridging was not routinely given.4 In thisstudy, the 30-day incidence of postoperative strokein patients with atrial fibrillation was 1.3%, whichwas more than fourfold higher than in patientswithout atrial fibrillation (OR, 4.6; 95% CI: 4.2–5.0). Taken together, these studies suggest that inpatients with atrial fibrillation who undergo sur-gery and do not receive bridging anticoagulation,the risk for perioperative arterial thromboembo-lism, consisting of stroke and transient ischemicattack, is approximately 1%. Furthermore, basedon an average annual risk of stroke of 5%(0.013%/d or �0.1% during 8 days of VKA inter-ruption), these studies suggest that the risk forarterial thromboembolism in the perioperativeperiod without bridging is higher than that pre-dicted based on mathematical modeling.74

2.4.3 Patients With Prior VTE

Risk Stratification: Compared to patients withatrial fibrillation or a mechanical heart valve, thereare several distinctions in the assessment of VKAinterruption and need for bridging anticoagulationin patients with prior VTE (ie, deep vein throm-bosis, pulmonary embolism). Whereas the first twogroups are at risk for stroke and other arterialthromboembolism, patients with VTE are at riskfor recurrent deep vein thrombosis or pulmonaryembolism. The consequences of these outcomesdiffer markedly. Embolic stroke is fatal or associ-ated with significant neurologic deficit in 70% ofpatients.11,12 On the other hand, recurrent VTE isfatal in approximately 4 to 9% of patients and isassociated with less morbidity.75,76 In addition,though low-dose anticoagulation has not beenproven to decrease the risk of arterial thrombo-embolic events, it has been shown in nonbridgingtrials to decrease the risk of postoperative VTE.65

Thus, although there is a lesser role for low-doseLMWH or UFH for patients with atrial fibrillationor mechanical heart valves, there is a strongerrationale for using these medications as bridgingtherapy for patients with prior VTE.

Risk stratification in patients with VTE is basedon an assessment of the risk for recurrent VTEafter the start of treatment,77 and risk factors forrecurrent disease after anticoagulant therapy hasbeen stopped.78,79 Patients at high risk for recur-rent disease may include those with: (1) recent(within 3 months) VTE; or (2) severe thrombo-philic conditions (deficiency of protein C, protein




S or antithrombin, antiphospholipid antibodies, ormultiple thromobophilic abnormalities). Patientswho have had prior VTE after surgery might beconsidered high risk depending on the type ofsurgery they are undergoing (and the associatedthromboembolic risk) and perioperative anti-thrombotic management should be individualized.Risk stratification is less precise in other patientsand would be predicated on individualized factors.Patients at moderate risk for recurrent diseasemay include those with prior VTE within thepast 3 to 12 months, nonsevere thrombophilicconditions (heterozygous carrier of factor VLeiden mutation or factor II mutation), recurrentVTE, or active cancer (treated within 6 months orpalliative). Patients at low risk may includethose in whom VTE occurred � 12 months agoand do not have any of the above-mentioned riskfactors.

Assessing Need for Bridging Anticoagulation: Pro-spective cohort studies have evaluated bridging an-ticoagulation using therapeutic and low-dose regi-mens of various LMWHs in approximately 500patients with prior VTE39,55,57,59–64 (Table 3). Theoverall crude risk for recurrent symptomatic VTEwas 0.60% (95% CI: 0.13–1.7). However, the efficacyof low-dose LMWH or UFH as perioperative antico-agulation is unknown as data are lacking in regard tothe risk of recurrent VTE without bridging antico-agulation.

Recommendation

2.4. In patients with a mechanical heart valveor atrial fibrillation or VTE at high risk (Table2) for thromboembolism, we recommend bridg-ing anticoagulation with therapeutic-dose SCLMWH or IV UFH over no bridging during tem-porary interruption of VKA therapy (Grade 1C);we suggest therapeutic-dose SC LMWH over IVUFH (Grade 2C). In patients with a mechanicalheart valve or atrial fibrillation or VTE at moder-ate risk (Table 2) for thromboembolism, we sug-gest bridging anticoagulation with therapeutic-dose SC LMWH, therapeutic-dose IV UFH, orlow-dose SC LMWH over no bridging duringtemporary interruption of VKA therapy (Grade2C); we suggest therapeutic-dose SC LMWH overother management options (Grade 2C). In patientswith a mechanical heart valve or atrial fibrillationor VTE at low risk (Table 2) for thromboembo-lism, we suggest low-dose SC LMWH or no bridg-ing over bridging with therapeutic-dose SCLMWH or IV UFH (Grade 2C).

Values and preferences: In patients at high or moderaterisk for thromboembolism, the recommendations re-flect a relatively high value on preventing thromboem-bolism and a relatively low value is on preventingbleeding; in patients at low risk for thromboembolism,the recommendations reflect a relatively high value onpreventing bleeding and a relatively low value onpreventing thromboembolism.

3.0 Perioperative Management of PatientsWho Are Receiving Bridging

Anticoagulation

In patients who require temporary interruptionof VKAs and are to receive bridging anticoagula-tion, several treatment regimens have been as-sessed21,22,39,40,56,57,80,81 and are summarized in Table3. In total, � 4,000 patients who had temporary inter-ruption of a VKA and received bridging anticoagulationhave been studied to date, of whom approximately72% received therapeutic-dose LMWH, approximately20% received low-dose LMWH, and approximately 8%received therapeutic-dose UFH.

3.1 Perioperative Anticoagulation TreatmentRegimens

3.1.1 Therapeutic-Dose UFH

Therapeutic-dose IV UFH had been the most com-monly used bridging regimen7,82,83 but its use hasdeclined in recent years,84,85 likely because of theincreased inconvenience of IV drug administration andthe increase in the number of surgical procedures thatare being done without hospitalization.86 In studies thatassessed bridging anticoagulation with therapeutic-dose UFH, a dose-adjusted IV infusion was used,administered to achieve a target activated partialthromboplastin time (APTT) of 1.5 to 2.0 times thecontrol APTT value, with the infusion stopped approx-imately 4 h before surgery and resumed during theinitial 24 h after surgery.7,83 An emerging alternative toIV UFH is SC UFH, which is administered as a fixed,weight-based dose regimen (250 IU/kg bid) withoutAPTT monitoring and has shown to be efficacious andsafe for the treatment of acute VTE.87 The use offixed-dose SC UFH may provide a practical alternativeto IV UFH for bridging anticoagulation, though it hasonly been assessed in a small number of patients whorequired temporary interruption of warfarin for surgery.55

3.1.2 Therapeutic-Dose LMWH

Clinicians have, in recent years, increasinglyturned to therapeutic-dose SC LMWH in lieu ofUFH as bridging anticoagulation,84,85 likely because




it can be easily administered outside of hospital andwithout laboratory monitoring.36 There is no stan-dardized bridging regimen with LMWH, and severaltherapeutic-dose regimens have been studied: dalte-parin 200 IU/kg qd; enoxaparin 1.5 mg/kg qd; tinza-parin 175 IU/kg qd; dalteparin 100 IU/kg bid; andenoxaparin 1 mg/kg bid.21,22,39,80

In the postoperative period, the use of therapeutic-dose LMWH administered can vary. The 3 principalmanagement approaches that have been studied are:(1) to administer therapeutic-dose LMWH within afixed time period after a procedure (within initial24 h); (2) to administer therapeutic-dose LMWHwithin a varied time period after a procedure (24 to72 h), with the initiation depending on the procedure-related bleeding risk and the adequacy of postopera-tive hemostasis; and (3) to replace therapeutic-doseLMWH with low-dose LMWH in select patientswho are undergoing a procedure associated with ahigh bleeding risk.

3.1.3 Low-Dose LMWH or UFH

Low-dose UFH (eg, UFH, 5,000 IU bid) orlow-dose LMWH (eg, enoxaparin, 30 mg bid, dalte-parin, 5,000 IU qd), which is typically used for theprevention of deep vein thrombosis in at-risk surgicaland medical patients, provides another bridging an-ticoagulation treatment option.65 This approach forperioperative anticoagulation has not been as widelystudied as therapeutic-dose regimens and has beenassessed mainly in lower-risk patients with atrialfibrillation or those with prior VTE. Though thisanticoagulant regimen has been used with the pre-sumed intent of providing some antithrombotic effi-cacy but at a lower risk for perioperative bleeding,data are very limited in regard to the efficacy oflow-dose UFH or LMWH to prevent arterial throm-boembolism in patients with a mechanical heartvalve or chronic atrial fibrillation.54,63,64,72 Indirectevidence from nonsurgical clinical settings involvingpatients with atrial fibrillation who received VKAsindicates that, compared to patients who were ther-apeutically anticoagulated, patients who had sub-therapeutic anticoagulation (INR � 2.0) were morelikely to develop a stroke and such strokes were moresevere and associated with greater mortality.88–90

Although, to our knowledge, there have been nostudies that have assessed the efficacy of low-doseLWMH or UFH to prevent arterial thromboembo-lism, a recent trial involving patients with chronicatrial fibrillation found that treatment with idrapa-rinux, a synthetic anti-Xa inhibitor, when adminis-tered in a therapeutic-dose regimen was as effica-cious as VKA therapy (INR range 2.0 to 3.0) for theprevention of stroke and systemic embolism.91 This

finding supports the premise that administration of atherapeutic-dose regimen of a non-VKA anticoagu-lant with properties similar to LMWHs is efficaciousfor the prevention of arterial thromboembolism.

Low-dose LMWH or UFH may be incorporatedinto a perioperative anticoagulation regimen in twopossible clinical scenarios. The first is as a “stand-alone” regimen in patients with prior VTE who arereceiving VKA therapy and are at moderate or lowrisk for recurrent disease in whom a therapeutic-dose anticoagulation regimen may not be consid-ered. In such patients, a low-dose LMWH or UFHregimen could be used during interruption of a VKAwith the intent of preventing recurrent venous (butnot arterial) thromboembolism. The rationale forthis approach is based on the established efficacy oflow-dose LMWH or UFH to prevent postoperativeVTE.65 The second scenario is in patients with anyclinical indication for VKA therapy who are under-going surgery that is associated with a high risk forbleeding (eg, cardiac, neurosurgical, urologic, majororthopedic). In such patients, administration oftherapeutic-dose LMWH or UFH during the initial48 to 72 h after surgery (or for the entire postoper-ative period) may confer an unacceptably high riskfor bleeding complications and a less intense antico-agulant regimen consisting of low-dose SC LMWHor UFH is likely to confer a lower risk for postoper-ative bleeding. Thus, in a registry involving 1,077patients who received bridging anticoagulation, post-operative low-dose LMWH or UFH was associatedwith a lower risk for minor bleeding compared tobridging anticoagulation with therapeutic-dose LMWHor UFH (OR � 0.46; CI: 0.20 –1.01).80 However,this registry was underpowered to detect potentialdifferences in major bleeding with low-dose ortherapeutic-dose anticoagulation.

To our knowledge, no prospective trials havecompared low-dose and therapeutic-dose LMWH orUFH as bridging anticoagulation to assess bothefficacy, in terms of preventing arterial thromboem-bolism, and safety, in terms of associated bleedingrisk. Although it is plausible that low-dose LMWH orUFH will confer a lower risk for bleeding complica-tions, one cannot exclude the possibility that suchtreatment will be less effective in preventing arterialthromboembolism than a therapeutic-dose regimen.This issue can only be resolved through well-designed randomized trials assessing different bridg-ing anticoagulation strategies.

3.1.4 Costs of Bridging Anticoagulation TreatmentRegimens

Recent studies have compared the costs of bridg-ing anticoagulation before and after surgery with




in-hospital administered IV UFH and out-of-hospitalbridging anticoagulation SC LMWH.92–95 In a pro-spective cohort study assessing perioperative antico-agulation with patient-administered SC LMWH,nurse-administered SC LWMH, and in-hospital ad-ministered IV UFH, the anticoagulant-related costsfor patients undergoing an overnight surgical proce-dure were estimated at $672, $933, and $3,916 (allUSD), respectively.93 Another cohort study compar-ing costs in 26 patients who received in-hospital IVUFH and 40 patients who received out-of-hospitalSC LMWH and underwent elective surgery found asignificantly lower mean total health-care cost (by$13,114 USD) in patients who received periopera-tive LMWH.94 In a decision analysis study involvingpatients who required VKA interruption for GIendoscopy, similar findings were found in terms oflower costs associated with out-of-hospital use ofLMWH as bridging anticoagulation.96 Taken to-gether, these findings indicate that, compared toin-patient administration of IV UFH, there is con-siderable cost savings with the use of SC LMWHs,which can be administered in an outpatient setting,typically by the patient or by another health-careprovider.92–95 Additional studies are needed to assessthe feasibility and costs of unmonitored SC UFHas bridging anticoagulation for patients in whomLMWH may be contraindicated, such as those withsevere renal insufficiency or in whom LMWHs maybe unavailable or too costly.87

Recommendation

3.1. In patients who require temporary inter-ruption of VKAs and are to receive bridginganticoagulation, from a cost containment per-spective we recommend the use of SC LMWHadministered in an outpatient setting wherefeasible instead of inpatient administration ofIV UFH (Grade 1C).Values and preferences: This recommendation re-flects a consideration not only of the trade-off be-tween the advantages and disadvantages of SCLMWH and IV UFH as reflected in their effects onclinical outcomes (LMWH at least as good, possiblybetter), but also the implications in terms of resourceuse (costs) in a representative group of countries(substantially less resource use with LMWH).

3.2 Interruption of Bridging AnticoagulationBefore Surgery

Bridging anticoagulation with IV UFH, which hasa half-life of approximately 45 min, can be inter-rupted 4 h before planned surgery, a time intervalthat approximates 5 elimination half-lives of UFH,36

and is in accordance with the practice used inbridging anticoagulation studies.7,55 In patients whoare receiving bridging anticoagulation with SCLMWHs, which have elimination half-lives of 4 to5 h,36 the last dose should be administered 20 to 25 hbefore surgery (or on the morning of the day beforesurgery), a time interval that approximates 5 elimi-nation half-lives of LMWHs.36 There is evidencesuggesting that there will be a residual anticoagulanteffect if therapeutic-dose LMWH is given too closeto the time of the procedure. Thus, in a prospectivecohort study involving 73 patients who receivedtherapeutic-dose or low-dose LMWH as bridginganticoagulation, 30% (11 of 37) of patients whoreceived therapeutic-dose LMWH (qd or bid doseregimens) had a residual anticoagulant effect (de-fined as an anti-factor Xa � 0.10 IU/mL) at the timeof surgery whereas � 1% (1 of 36) of patients whoreceived low-dose LMWH had a residual anticoag-ulant effect at surgery.97 In another prospectivecohort study of 98 patients who received bridginganticoagulation with enoxaparin 1 mg/kg bid, withthe last dose given on the evening before surgery, adetectable residual anticoagulant effect (anti-factorXa � 0.10 IU/mL) was found in 100% (98 of 98) ofpatients.98 Furthermore, 34% of patients had ananticoagulant effect at the time of surgery that isconsidered within the therapeutic range (anti-factorXa � 0.50 IU/mL). Although there were no majorbleeds in the patients from both of these studies,most patients underwent low bleeding risk proce-dures and the potential for bleeding in patientshaving major surgical or other higher-risk invasiveprocedures cannot be excluded. Taken together,these findings suggest that the last preoperativedose of therapeutic-dose LMWH before surgeryshould be reduced to minimize the risk for aresidual anticoagulant effect at the time of sur-gery. Until prospective trials address this issuefurther, one management option, especially inpatients who are undergoing major surgery or arereceiving spinal/epidural anesthesia, is to adminis-ter only the morning dose of LMWH in patientsreceiving a twice-daily therapeutic-dose regimenand to reduce by 50% the total dose of LMWHgiven in patients who are receiving a once-dailytherapeutic-dose regimen.

Recommendation

3.2. In patients who are receiving bridginganticoagulation with therapeutic-dose SC LMWH,we recommend administering the last dose ofLMWH 24 h before surgery or a procedureover administering LMWH closer to surgery(Grade 1C); for the last preoperative dose of




LMWH, we recommend administering ap-proximately half the total daily dose instead of100% of the total daily dose (Grade 1C). Inpatients who are receiving bridging anticoag-ulation with therapeutic-dose IV UFH, werecommend stopping UFH approximately 4 hbefore surgery over stopping UFH closer tosurgery (Grade 1C).

3.3 Resumption of Bridging Anticoagulation AfterSurgery

Following parenteral administration, LMWHs in-duce a rapid anticoagulant effect, with the potentialfor a detectable anticoagulant effect to occur within1 h and a peak anticoagulant effect to occur within 3to 5 h after administration.36 With UFH, though thetime to a peak anticoagulant effect varies, there is thepotential for this to also occur within 3 to 5 hfollowing an IV bolus and infusion.36 Consequently,clinicians should exercise caution when administer-ing these drugs in patients who have recently hadsurgery or other invasive procedures because of thepotential for bleeding at the surgical site, especiallywhen hemostasis is not secured. Three factors ap-pear to affect the risk for surgery-related bleeding:(1) the proximity to surgery that the anticoagulant isadministered; (2) the dose of anticoagulant adminis-tered; and (3) the type of surgery and its associatedbleeding risk. A superimposed consideration is thatbleeding can occur after any surgery or procedure,irrespective of the anticipated surgery-related riskfor bleeding and postoperative anticoagulant man-agement. Consequently, postoperative administra-tion of UFH and LMWHs should consider both theanticipated risk for bleeding, which is determinedpreoperatively, and the adequacy of surgical hemo-stasis, which is determined postoperatively.

3.3.1 Proximity to Surgery That Anticoagulants AreAdministered

In a pooled analysis of studies involving patientswho had major orthopedic surgery and receivedlow-dose fondaparinux (2.5 mg/d) 4 to 8 h postoper-atively or low-dose enoxaparin (40 to 60 mg/d) 12 to24 h postoperatively, the risk for major bleeding wassignificantly higher in fondaparinux-treated patients(2.7% vs 1.7%; p � 0.01).99 In another pooled anal-ysis that compared bleeding in patients who receivedlow-dose LMWH either within 6 h or 12 to 24 h aftermajor orthopedic surgery, the risk for bleeding washigher in patients who received LMWH closer tosurgery (6.3% [95% CI: 5–7] vs 2.5% [95% CI: 1–3]).100

In patients who received bridging anticoagulation,three prospective cohort studies suggest that delay-ing the postoperative initiation of therapeutic-dose

LMWH until hemostasis is secured and deferring (oravoiding altogether) postoperative therapeutic-doseLMWH are associated with a low risk for bleeding.In one study assessing 650 patients who underwent abroad spectrum of surgical and nonsurgical proce-dures and received therapeutic-dose bridging anti-coagulation, postoperative management, which in-cluded resumption of therapeutic-dose bridging withLMWH, depended on the anticipated bleeding riskand adequacy of postoperative hemostasis.21 Thus,patients who had procedures associated with a lowrisk for bleeding (eg, GI endoscopy, cardiac cathe-terization) resumed LMWH approximately 24 hafter the procedure (ie, day after procedure); pa-tients who had major surgery (eg, open abdominalsurgery) or in whom there was inadequate postoper-ative hemostasis resumed LMWH 48 to 72 h aftersurgery; and patients who had major surgery associ-ated with a high risk for bleeding (eg, cardiac,neurosurgical, urologic, major orthopedic) did notreceive any postoperative LMWH. With this ap-proach, the incidence of major bleeding was 1.0%during the first week after surgery, with no fatalbleeds. In another study involving 220 patients witha mechanical heart valve that used the same postop-erative anticoagulant management approach, the in-cidence of major bleeding was 2.3% during the firstweek after surgery, with no fatal bleeds.40 Anotherprospective cohort study involved 224 patients, inwhom once-daily therapeutic-dose LMWH or low-dose LMWH (in patients having surgery associatedwith a high bleeding risk) started on the day aftersurgery and administered only if postoperative he-mostasis was secured.22 The risk for bleeding duringthe first week after surgery in patients who receivedtherapeutic-dose LMWH was 2.9%, with no fatalbleeds. The risk for arterial thromboembolism withthe perioperative management approach used inthese studies is low (� 1%) and is discussed furtherin Section 2.4.

3.3.2 Dose of Anticoagulant Administered

A prospective multicenter registry evaluated 493patients who required interruption of a VKA andreceived bridging with LMWH or UFH or no bridg-ing (Jaffer A, submitted for publication). After ad-justment for surgical and patient-specific bleedingrisk factors, the administration of therapeutic-doseLMWH or UFH after the surgery or procedureconferred a greater than fourfold greater risk formajor bleeding (OR, 4.4; 95% CI: 1.5–14.7) com-pared to the postoperative administration of either alow-dose LMWH or UFH regimen or no bridging.




3.3.3 Type of Surgery and Associated Bleeding Risk

A prospective bridging study39 of 260 patients inwhich all patients received therapeutic-dose once-daily LMWH perioperatively, with the first postop-erative dose administered 12 to 24 h after surgery,categorized surgeries or procedures as major (ex-pected duration � 1 h) or minor (expected duration� 1 h). This study39 found that major bleedingoccurred in 0.7% (1 of 148) of patients who had aninvasive procedure, in 0% (0 of 72) of patients whohad minor surgery, and in 20.0% (8 of 40) of patientswho had major surgery.

Taken together, these findings suggest that inpatients who receive bridging anticoagulation withtherapeutic-dose LMWH, this regimen should beadministered carefully in the postoperative period. Itappears that therapeutic-dose LMWH can be safelyresumed on the day after surgery in patients whohave had a minor surgical or other invasive proce-dure and in whom there is adequate hemostasis. Onthe other hand, administering therapeutic-doseLMWH within 24 h after major surgery appears toconfer an unacceptably high risk for bleeding com-plications.

In patients undergoing major surgery or a proce-dure (surgical or nonsurgical) associated with a highbleeding risk, management options that are prefera-ble over administering therapeutic-dose SC LMWHor IV UFH in close proximity to surgery (ie, within24 h) include: (1) delaying the resumption oftherapeutic-dose LMWH or UFH for 48 to 72 hafter the surgery/procedure; (2) administering onlylow-dose LMWH after the surgery/procedure; and(3) avoiding the use of LMWH altogether in thepostoperative period. The management option cho-sen is individualized and will depend on both thebleeding risk associated with the surgical or otherinvasive procedure and the adequacy of postopera-tive hemostasis. For example, in patients undergoingmajor surgery (eg, bowel resection), it may be rea-sonable to delay the resumption of therapeutic-doseLMWH or UFH. In patients undergoing a surgery(eg, radical prostatectomy) or procedure (eg, kidneybiopsy) associated with a high risk for bleeding, itmay be reasonable to not administer any LMWH orUFH after surgery and to simply resume VKAs.

Recommendation

3.3. In patients undergoing a minor surgical orother invasive procedure and who are receivingbridging anticoagulation with therapeutic-doseLMWH, we recommend resuming this regi-men approximately 24 h after (eg, the dayafter) the procedure when there is adequate

hemostasis over a shorter (eg, < 12 h) timeinterval (Grade 1C). In patients undergoingmajor surgery or a high bleeding risk surgery/procedure and for whom postoperativetherapeutic-dose LMWH/UFH is planned, werecommend either delaying the initiation oftherapeutic-dose LMWH/UFH for 48 to 72 h aftersurgery when hemostasis is secured, administer-ing low-dose LMWH/UFH after surgery whenhemostasis is secured, or completely avoidingLMWH or UFH after surgery over the adminis-tration of therapeutic-dose LMWH/UFH in closeproximity to surgery (Grade 1C). We recommendconsidering the anticipated bleeding risk and ad-equacy of postoperative hemostasis in individualpatients to determine the timing of LMWH orUFH resumption after surgery instead of resum-ing LMWH or UFH at a fixed time after surgeryin all patients (Grade 1C).

3.4 Laboratory Monitoring of BridgingAnticoagulation

In patients who are receiving IV UFH as bridginganticoagulation, clinicians can use the APTT to guidepreoperative and postoperative anticoagulation. How-ever, use of an UFH dosing nomogram, which wasnot designed for use in the perioperative setting, maybe misleading.39,101 For example, a dosing nomo-gram for IV UFH may result in excessive anticoag-ulation (ie, APTT � 150 s) for up to a 24-h periodwhile dose adjustments are being made. Althoughshort periods of excessive anticoagulation with UFHmay not increase the risk for bleeding in nonopera-tive clinical settings,102,103 such short periods ofover-anticoagulation might increase the risk for bleed-ing in a postoperative setting.

In patients who are receiving bridging anticoagu-lation with therapeutic-dose LMWH, there is noestablished role for routine perioperative monitoringof anti-factor Xa levels, as in certain nonoperativeclinical settings.36 In selected patient groups, such asthose with severe renal insufficiency (ie, calculatedcreatinine clearance � 30 mL/min or serum creati-nine � 178 mmol/L or � 2 g/dL) or at extremes ofbody weight, anti-factor Xa monitoring may be con-sidered to guide treatment as in nonoperative clinicalsettings.36

Recommendation

3.4. In patients who are receiving bridginganticoagulation with LMWH, we suggest againstthe routine use of anti-factor Xa levels to monitorthe anticoagulant effect of LMWHs (Grade 2C).




4.0 Perioperative Management of PatientsWho Are Receiving Antiplatelet Therapy

An increasing number of patients are receiving anti-platelet drugs for the primary and secondary preven-tion of myocardial infarction or stroke and for theprevention of coronary stent thrombosis after place-ment of a bare metal or drug eluting stent.104,105 Theperioperative management of these patients is increas-ing in complexity because the spectrum of risk for acardiovascular event varies widely. In addition,these patients may be receiving treatment withone of several antiplatelet drug regimens, whichinclude: (1) aspirin alone; (2) clopidogrel alone; (3)aspirin combined with clopidogrel; (4) aspirincombined with dipyridamole; or (5) cilostozol,either alone or combined with either aspirin orclopidogrel. This section will focus on patientswho are receiving aspirin and/or clopidogrel.

4.1 Risk Stratification

Patients who are receiving antiplatelet therapyencompass a spectrum of risk for cardiovascularevents that depends, to a large extent, on the clinicalindication for antiplatelet therapy and whether pa-tients are receiving such treatment for the primary orsecondary prevention of cardiovascular disease. Cli-nicians should incorporate risk stratification in deci-sions concerning temporary interruption or continu-ation of antiplatelet therapy in the perioperativeperiod. There are no risk classification schemes, toour knowledge, that encompass the spectrum ofbenefit from antiplatelet agents. Nonetheless, pa-tients at low risk for perioperative cardiovascularevents in whom temporary interruption of antiplate-let drugs would not be expected to confer a substan-tial increased risk for cardiovascular events includethose who are receiving antiplatelet therapy (typi-cally aspirin) for the primary prevention of myocar-dial infarction or stroke.106 On the other hand,patients at high risk for cardiovascular events inwhom it may be preferable to continue antitplatelettherapy perioperatively include those who have hadrecent (within 3 to 6 months) placement of a baremetal or drug-eluting coronary stent, and to a lesserextent, who have suffered a myocardial infarctionwithin the past 3 months.107 The risk for cardiovas-cular events in these high-risk groups should beweighed against the risk and clinical impact ofbleeding with the operation planned when antiplate-let drugs are continued in the perioperative period.

4.2 Interruption of Antiplatelet Therapy BeforeSurgery

For patients who are receiving aspirin, whichirreversibly inhibits platelet function through cyclo-

oxygenase-1 inhibition, clinicians intending no anti-platelet effect at the time of surgery should interrupttherapy 7 to 10 days before surgery.108 Althoughaspirin has a half-life of 15 to 20 min, it irreversiblyinhibits platelet cyclooxygenase-1 and, therefore, itseffect persists for 7 to 10 days, which approximatesthe platelet lifespan.109,110 Consequently, 4 to 5 daysafter stopping aspirin will result in approximately50% of platelets having normal function, whereas 7to 10 days after stopping aspirin will result in � 90%of platelets having normal function. In patients whoare receiving clopidogrel, a thienopyridine derivativethat irreversibly inhibits adenosine diphosphate re-ceptor-mediated platelet activation and aggregationand has a half-life of 8 h, treatment should beinterrupted 7 to 10 days before surgery since it takesabout that many days to replace the platelet pool.111

This approach also applies to ticlopidine, anotherthienopyridine derivative which is used less fre-quently compared to clopidogrel, in part because ofan increased risk for drug-induced adverse effectssuch as neutropenia.112,113

Another antiplatelet agent that is used in combi-nation with aspirin is dipyridamole, a pyrimidopyri-midine derivative with antiplatelet and vasodilatorproperties that is indicated for secondary strokeprevention in patients with cerebrovascular dis-ease.114 Dipyridamole has reversible effects on plate-let function and has an elimination half-life of ap-proximately 10 h.115 However, since dipyridamole iscombined with aspirin (200 mg dipyridamole � 25mg aspirin), this drug would need to be interrupted7 to 10 days before elective surgery to allow elimi-nation of the antiplatelet effects of both drugs.

Cilostazol is a phosphodiesterase inhibitor withantiplatelet and vasodilatory properties that revers-ibly affects platelet function through cyclic adeno-sine monophosphate (cAMP) mediated inhibition ofplatelet activation and aggregation. Cilostazol may beused in patients with coronary artery disease, typi-cally if they have a coronary stent116 or peripheralarterial disease.117 The pharmacokinetics of cilosta-zol are dose-dependent, with an elimination half-lifeof approximately 10 h.118 Consequently, this drugwould need to be interrupted 2 to 3 days (corre-sponding to 5 elimination half-lives of cilostazol)before surgery to ensure elimination of its antiplate-let effect at the time of surgery.

For patients who are receiving a nonselectivenonsteroidal antiinflammatory drug (NSAID) or acyclyooxygenase-2 selective NSAID (ie, celecoxib),there is reversible inhibition of platelet-mediatedcyclooxygenase activity. To ensure there is no resid-ual antiplatelet effect at the time of surgery, theNSAID should be stopped at a time that correspondsto 5 elimination half-lives for that drug.119,120 For




NSAIDs with a short, 2 to 6 h, half-life (eg, ibupro-fen, diclofenac, ketoprofen, indomethacin), thesedrugs should be stopped on the day before surgery.For NSAIDs with an intermediate, 7 to 15 h, half-life(eg, naproxen, sulindac, diflunisal, celecoxib), suchtreatment should be stopped 2 to 3 days beforesurgery. Finally, for NSAIDs with a long, � 20 hhalf-life (eg, meloxicam, nabumetone, piroxicam),these drugs should be stopped 10 days beforesurgery.

Recommendation

4.2. In patients who require temporary inter-ruption of aspirin- or clopidogrel-containingdrugs before surgery or a procedure, we sug-gest stopping this treatment 7 to 10 days beforethe procedure over stopping this treatmentcloser to surgery (Grade 2C).

4.3 Resumption of Antiplatelet Therapy AfterSurgery

In patients who have temporary interruption ofantiplatelet drugs before surgery, these agentsshould, in general, be resumed as soon as there isadequate postoperative hemostasis after surgery.Four studies assessed bridging anticoagulation inpatients with a mechanical heart valve, some of whomwere receiving both VKAs and aspirin.21,22,40,55 In thesestudies, aspirin was resumed on the same day as VKAs,starting with the usual maintenance dose of 81 mgdaily. Data are lacking in regard to the resumption ofclopidogrel or other antiplatelet drugs after surgery.One issue that warrants consideration is whetherresumption of treatment should be with a mainte-nance dose of clopidogrel (75 mg/d), which achievesmaximal platelet function inhibition 5 to 10 daysafter its administration,121–123 or with a loading dose(300 to 600 mg/d), which achieves maximal plateletfunction inhibition within 2 to 15 h after administra-tion.124–126 The dose of clopidogrel resumption willdepend largely on whether a patient has a coronarystent, the type of stent implanted, and how recentlythe stent was implanted.

Recommendation

4.3. In patients who have had temporary interrup-tion of aspirin therapy because of surgery or aprocedure, we suggest resuming aspirin approxi-mately 24 h (or the next morning) after surgerywhen there is adequate hemostasis instead ofresuming aspirin closer to surgery (Grade 2C). Inpatients who have had temporary interruption ofclopidogrel because of surgery or a procedure,

we suggest resuming clopidogrel approximately24 h (or the next morning) after surgery whenthere is adequate hemostasis instead of resumingclopidogrel closer to surgery (Grade 2C).

4.4 Laboratory Monitoring of Antiplatelet Therapy

Platelet function assays are available to measurethe antiplatelet activity of aspirin, clopidogrel and,potentially, other antiplatelet drugs, prior to sur-gery.127 However, these methods are not well stud-ied outside of a cardiac surgery or percutaneouscoronary intervention (PCI) setting.128 Furthermore,the clinical significance of the assay results is uncertainas they have not been shown to identify patients atincreased risk for perioperative bleeding.127 Additionalresearch is needed to identify the potential clinicalutility of platelet function assays.

Recommendation

4.4. In patients who are receiving antiplateletdrugs, we suggest against the routine use ofplatelet function assays to monitor the anti-thrombotic effect of aspirin or clopidogrel(Grade 2C).

4.5 Surgery in Patients Receiving AntiplateletTherapy

4.5.1 Noncardiac Surgery

In patients who are receiving antiplatelet drugtherapy and are undergoing noncardiac surgery,there are no randomized trials or prospective cohortstudies that compare the clinical benefits and risks ofcontinuing antiplatelet drugs with their temporaryinterruption. A randomized placebo-controlled trialinvolving patients undergoing hip fracture repair orjoint replacement surgery assessed de novo use ofaspirin compared to no aspirin use in the perioper-ative period and reported higher rates of majorbleeding in aspirin treated patients (2.9% vs 2.4%,p � 0.04).9

Studies in patients undergoing abdominal or pelvicsurgery are limited. A retrospective cohort study in52 patients found that perioperative continuation ofaspirin increases the risk for bleeding after prosta-tectomy.31 A retrospective cohort study involving 200patients who underwent intraabdominal surgeryfound that 12 of 55 (22%) patients with aspirin-associated abnormal platelet function had excessiveperioperative bleeding, whereas 7 of 97 (7%) withnormal platelet function had excessive bleeding.129

However, another study involving 52 patients whohad surgery found that perioperative aspirin use didnot confer an increased risk for bleeding.130




Retrospective cohort studies have suggested in-creased rates of bleeding with perioperative contin-uation of clopidogrel.131,132 In addition, one prospec-tive cohort study in patients who underwentbronchoscopy found significantly higher incidencesof moderate or severe bleeding after biopsy inpatients who received clopidogrel (61%) or clopi-dogrel and aspirin (100%) compared to no antiplate-let drug (2%).133

4.5.2 CABG

Elective coronary artery bypass graft (CABG)surgery is frequently done in patients who are re-ceiving antiplatelet therapy with aspirin and/or clo-pidogrel.134 In addition, 10 to 15% of patients hos-pitalized with an acute coronary syndrome willrequire urgent CABG surgery during their hospital-ization and such patients are typically receivingantiplatelet therapy (aspirin alone or aspirin andclopidogrel) and anticoagulant therapy, the later withLMWH or UFH.135 Minimizing the risk for periop-erative bleeding in patients undergoing CABG sur-gery is important because of an increased risk fordeath and other adverse outcomes in patients whorequire a blood transfusion in the perioperativeperiod.136 In one study involving 11,963 patients whounderwent CABG, of whom 49% received transfu-sion of RBCs, transfusion was associated with signif-icant increases in mortality (OR, 1.77; CI: 1.67–1.87), renal failure (OR, 2.06; CI: 1.87–2.27), andneurologic events (OR, 1.37; CI: 1.30–1.44).137 Theperioperative management of anticoagulant therapyis comparable to that of noncardiac surgery, withinterruption of LMWH or UFH at a time prior toCABG surgery that will eliminate the anticoagulanteffect by the time of surgery. On the other hand, theperioperative management of antiplatelet therapy ismore problematic since 7 to 10 days after stoppingtreatment is required to eliminate an antiplateleteffect and urgent CABG is often required withoutadvance notice of 7 to 10 days.

In patients who are receiving aspirin and requireCABG surgery, observational studies show that con-tinuing aspirin in the perioperative period appears toconfer an increased risk for mediastinal bleeding,blood transfusion, and reoperation,138,139 althoughthis finding was not found in all studies.140 Againstthese risks, aspirin use within 5 days prior to CABGwas shown in a large cohort study to confer a lowerrisk of postoperative mortality and without a con-comitant increase in reoperation for bleeding orneed for blood transfusion.26 Based on this benefit ofcontinued perioperative aspirin use in patients un-dergoing CABG, if aspirin therapy has been inter-rupted before surgery, it should be administered

early after surgery, always within 48 h after CABGand, preferably, within 6 h after surgery.141

In patients who are receiving clopidogrel andrequire CABG, there are no data to suggest benefitfrom the administration of clopidogrel in the periop-erative period whereas there are preliminary datasuggesting harm with this approach.142 In the CanRapid risk stratification of Unstable Angina PatientsSupress Adverse Outcomes With Early Implemen-tation of the American College of Cardiology/Amer-ican Heart Association guidelines (CRUSADE)study that included 2,855 patients with a non-STelevation myocardial infarction, 87% of patients un-derwent CABG within 5 days of prior clopidogrelexposure.143 Such patients had a 70% higher likeli-hood for a transfusion requirement of 4 U or more ofRBCs. The most compelling data about the risk ofbleeding among patients undergoing CABG who arereceiving clopidogrel come from the Clopidogrelin Unstable angina to prevent Recurrent Events(CURE) trial.144 In a postrandomization subgroupanalysis of this trial, exposure to clopidogrel within 5days prior to CABG was associated with an approx-imately 50% increase in major bleeding. Other ret-rospective studies have confirmed the increased riskfor bleeding with prior clopidogrel exposure in pa-tients undergoing CABG surgery.145–148 An in-creased risk of bleeding appears to occur even whenCABG is performed in an off-pump manner.149

Mitigating the risk for perioperative bleeding andtransfusion with antifibrinolytic drugs, such as apro-tinin, or platelet transfusion is problematic becauseboth treatments are associated with adverse effects.Although two randomized trials suggested a reduc-tion in the need for transfusion among patientstreated with aprotinin undergoing surgery while onclopidogrel without raising concerns of excessive riskof thrombosis, aprotinin appears to be associatedwith an increased risk for thrombotic and otheradverse effects,150,151 and is no longer available forclinical use in the United States. In one observationalstudy involving 4,374 patients undergoing CABGsurgery for ST-segment elevation myocardial infarc-tion, aprotinin use was associated with a 55% in-creased risk for myocardial infarction or congestiveheart failure and a 181% increased risk for stroke orencephalopathy.152 Similarly, pre-CABG platelettransfusion is associated with longer surgery and, par-adoxically, more bleeding and reoperation for bleedingcomplications.153 Alternative antifibrinolytic agents thatcan be used in lieu of aprotonin to reduce perioperativebleeding in patients undergoing cardiac surgery includeepsilon-aminocaproic acid and tranexamic acid, whichhave been shown to reduce transfusion require-ments.154,155 The efficacy of 1-deamino-8-D-arginine




vasopressin (DDAVP) in patients undergoing cardiacsurgery who have been exposed to aspirin is lessclear.156

4.5.3 PCIs

Randomized trials have compared a 325-mg doseof aspirin to placebo among patients undergoing aballoon angioplasty type of PCI. The reduction inrisk associated with aspirin administration in thesestudies has led to the recommendation that aspirinbe administered in all patients prior to any PCIprocedure.17

Clopidogrel has been compared to placebo amongpatients with acute coronary syndromes and found toreduce the risk of procedure-related events.157,158 Inpostrandomization subgroup analyses of patients un-dergoing PCI in these trials, clopidogrel was partic-ularly beneficial among patients undergoing PCI andthe benefit seems to apply not only to patients whoreceived stents but to those undergoing balloonangioplasty and other types of PCI procedures aswell. Pretreatment with clopidogrel is recommendedbefore any type of PCI procedure whenever it can beaccomplished and such treatment should be contin-ued during the periprocedural period. Among pa-tients on long-term clopidogrel therapy, one studyshowed that periprocedural administration of a 600-mgloading dose of clopidogrel to such patients resulted ina greater inhibition of platelet aggregation than notreceiving a loading dose.159 Other studies have sug-gested that clinical outcomes are improved if PCI isperformed after a 600-mg loading dose of clopidogrelhas been administered; few patients in those studieswere receiving long-term clopidogrel and it is notknown whether chronically administered clopidogrelachieves the same effect.160–162

Recommendation

4.5. For patients who are not at high risk forcardiac events, we recommend interruption ofantiplatelet drugs (Grade 1C). For patients athigh risk of cardiac events (exclusive of coro-nary stents) scheduled for noncardiac surgery,we suggest continuing aspirin up to and beyondthe time of surgery (Grade 2C); if patients arereceiving clopidogrel, we suggest interruptingclopidogrel at least 5 days and, preferably,within 10 days prior to surgery (Grade 2C). Inpatients scheduled for CABG, we recommendcontinuing aspirin up to and beyond the time ofCABG (Grade 1C); if aspirin is interrupted, werecommend it be reinitiated between 6 h and48 h after CABG (Grade 1C). In patients sched-uled for CABG, we recommend interrupting

clopidogrel at least 5 days and, preferably, 10days prior to surgery (Grade 1C). In patientsscheduled for PCI, we suggest continuing aspi-rin up to and beyond the time of the procedure;if clopidogrel is interrupted prior to PCI, wesuggest resuming clopidogrel after PCI with aloading dose of 300 to 600 mg (Grade 2C).

4.6 Surgery in Patients With Coronary Stents

Patients who are receiving antiplatelet therapybecause of a bare metal or drug-eluting stent in thecoronary arteries deserve special consideration be-cause of the high thrombotic risk if antiplatelet drugtherapy is interrupted. In such patients who areundergoing noncardiac surgery, there is a markedlyincreased risk of coronary stent thrombosis in thepostoperative period, especially if surgery is under-taken in close proximity to stent placement.163–172

Furthermore, the clinical impact of stent thrombosisin this clinical setting is considerable, as it will befatal or associated with a large myocardial infarctionin � 50% of affected patients.107,163,173–175 A retro-spective cohort study assessed 40 consecutive pa-tients who had elective noncardiac surgery � 6weeks after coronary artery stenting.163 In this study,eight patients (20%) died postoperatively, in whomall but one had perioperative interruption of clopi-dogrel or aspirin.

To mitigate the risk for perioperative stentthrombosis, elective noncardiac surgery should beavoided during the period after stent placementwhen stent endothelialization is ongoing as this isthe time when coronary stents are most suscepti-ble to thrombosis. In patients with a bare metalstent, the aforementioned study suggested that therisk of thrombosis with surgery was higher inpatients who had surgery within 2 weeks of stent-ing compared to more than 2 weeks after stenting(p � 0.15).163 A larger study suggested that therisk of bare metal stent thrombosis and otheradverse events is increased if noncardiac surgery isperformed within 6 weeks of stent placement,164

which is consistent with the approximate timerequired for endothelialization around the baremetal stents.176,177

A relevant, but poorly studied, issue in the man-agement of patients with coronary stents who requiresurgery is whether bridging therapy is warranted forpatients in whom interruption of antiplatelet therapyis required because of a high bleeding risk associatedwith the planned surgery. In such patients, bridgingtherapy might consist of administering LMWH orUFH in a manner similar to that in patients whorequire temporary interruption of VKAs, though thisapproach has not been formally studied to assess




efficacy and should be weighed against a potentialincreased risk for postoperative bleeding. An alter-native approach might be the use of bridging therapywith short-acting antiplatelet drugs such as the gly-coprotein IIb/IIIa antagonists tirofiban or eptifi-batide, which have been studied in patients under-going PCI.17 These agents have elimination half-livesof approximately 2 h and their interruption 10 hbefore surgery would allow restoration of plateletfunction by the time of surgery.108 Emerging short-acting antiplatelet drugs that target platelet P2 re-ceptors, such as cangrelor, may have clinical utility inthe perioperative setting because of rapid reversal ofantiplatelet activity after treatment is stopped.178

Studies are needed to assess the efficacy and safetyof bridging therapy in patients who are receivingantiplatelet drugs and, until relevant data are avail-able, clinical judgment and caution are urged inregard to the use of short-acting antithromboticagents in patients who require temporary interrup-tion of aspirin and/or clopidogrel.

Less is known about the timing of noncardiacsurgery in patients with a sirolimus- or paclitaxel-eluting coronary stent, in whom a longer time isrequired for coronary reendothelialization than pa-tients with a bare metal stent. There have beenseveral reports of thrombosis of such drug-elutingstents during the intraoperative and postoperativeperiod, in some cases even when surgery is per-formed years after stent placement.165–170 Thoughaspirin is recommended indefinitely after placementof a drug-eluting stent and clopidogrel is recom-mended for at least 3 months after placement of asirolimus-eluting stent and 6 months after placementof a paclitaxel-eluting stent, most patients are receiv-ing combined aspirin-clopidogrel therapy for at least12 months after placement of a drug-elutingstent.179,180 Consequently, elective surgery should bedelayed for 12 months after placement of a drug-eluting stent whenever possible.

Recommendation

4.6. In patients with a bare metal coronary stentwho require surgery within 6 weeks of stentplacement, we recommend continuing aspirinand clopidogrel in the perioperative period(Grade 1C). In patients with a drug-eluting cor-onary stent who require surgery within 12months of stent placement, we recommendcontinuing aspirin and clopidogrel in theperioperative period (Grade 1C). In patientswith a coronary stent who have interruptionof antiplatelet therapy before surgery, wesuggest against the routine use of bridgingtherapy with UFH, LMWH, direct thrombin

inhibitors, or glycoprotein IIb/IIIa inhibitors,(Grade 2C).Values and preferences: These recommendationsreflect a relatively high value placed on preventingstent-related coronary thrombosis, and a consider-ation of complexity and costs of administering bridgingtherapy in the absence of efficacy and safety data in thisclinical setting, and a relatively low value on avoidingthe unknown but potentially large increase in bleedingrisk associated with the concomitant administrationof aspirin and clopidogrel during surgery.

5.0 Perioperative Management ofAntithrombotic Therapy in Patients Who

Require Dental, Dermatologic, orOphthalmologic Procedures

Minor dental, dermatologic, and ophthalmologicprocedures can comprise up to 20% of all surgicaland nonsurgical procedures performed in patientswho are receiving antithrombotic therapy.21,55 Asthese procedures are typically associated with rela-tively little blood loss, a key question relates to thesafety of continuing antithrombotic therapy aroundthe time of the procedure and whether continuingtreatment confers an increased risk of clinicallyimportant bleeding. A practical issue that also relatesto the management of such patients is that most, ifnot all, minor procedures are undertaken in a clinicor other out-of-hospital setting. Consequently, bleed-ing that may occur after the procedure will occurwhile the patient is home and may generate concernand anxiety for the patient. Patients should, there-fore, be given instructions to deal with potentialbleeding, which usually requires prolonged localpressure over the site of a dental or dermatologicprocedure. In addition, patients should be advisedwhen bleeding is excessive and warrants medicalattention.

In our review of randomized and nonrandomizedprospective studies that assessed the risk of bleedingin patients who continue VKAs or antiplatelet drugsduring minor procedures, which are summarized inTables 4–8, we have focused on postproceduralbleeding. To distinguish bleeding that is clinicallyimportant and requires medical attention, frombleeding that does not require medical attention andis, typically, self-limiting, we classified bleeding intothree categories: (1) major bleeding, which refers tobleeding that requires transfusion of � 2 U packedRBCs181; (2) clinically relevant nonmajor bleeding,which refers to bleeding that is not major butrequires medical attention (eg, application of wounddressing or additional sutures); and (3) minor bleed-ing, which refers to bleeding that is self-limiting,




Tab

le4

—R

ando

miz

edT

rial

sA

sses

sing

Con

tinu

atio

nof

Ant

ithr

omb

otic

The

rapy

inP

atie

nts

Und

ergo

ing

Den

tal

Pro

cedu

res:

Cli

nica

lD

escr

ipti

onan

dR

esu

lts

(Sec

tion

5.1)

*

Stud

y/yr

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up

Clin

ical

Out

com

es,N

o./T

otal

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

TT

hera

pyN

o.G

roup

Thr

ombo

embo

lism

Eve

nts

Min

orB

leed

ing

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Bor

ea etal

182 /

1993

30V

KA

MH

V,n

�30

Den

tale

xtra

ctio

ns15

Tre

atm

ent:

cont

inue

VK

A�

irrig

atio

n�

tran

exam

icac

id

10d

0T

reat

men

t:1/

15(6

.67%

)C

ontr

ol:0

/15

Tre

atm

ent:

1/15

(6.7

%)

Con

trol

:2/1

5(1

3.3%

)

Tre

atm

ent:

0/15

Con

trol

:0/1

5

15C

ontr

ol:s

top

VK

AE

vans et

al18

3 /20

02

109

VK

AN

otre

port

edD

enta

lext

ract

ions

57T

reat

men

t:co

ntin

ueV

KA

Not

spec

ified

0T

reat

men

t:13

/57

(22.

8%)

Tre

atm

ent:

2/57

(3.5

%)

Tre

atm

ent:

0/57

52C

ontr

ol:s

top

VK

Ada

y�

2C

ontr

ol:7

/52

(13.

5%)

Con

trol

:0/5

2C

ontr

ol:0

/52

Sacc

oet

al18

4 /20

06

131

VK

AV

TE

�16

,AF

�39

,M

HV

�59

,va

lvul

arhe

art

dise

ase

�13

Ora

lsur

geri

es(d

enta

lex

trac

tions

,fix

ture

inse

rtio

ns,e

xcis

ion

ofcy

stic

neof

orm

atio

ns)

65T

reat

men

t:V

KA

(ful

l-dos

e)�

tran

exam

icac

id�

spon

ge

10d

0T

reat

men

t:no

tre

port

edC

ontr

ol:n

otre

port

ed

Tre

atm

ent:

6/65

(9.2

%)

Con

trol

:10/

66(1

5.2%

)

Tre

atm

ent:

0/65

Con

trol

:0/6

6

66C

ontr

ol:V

KA

(red

uced

-dos

e)da

y�

3A

rdek

ian

etal

209 /

2000

39A

spir

inV

TE

�6,

AF

�1,

coro

nary

arte

rydi

seas

e�

20,

stro

ke�

10

Ora

lsur

gery

(sim

ple

and

com

poun

dde

ntal

extr

actio

ns,

and

mor

eco

mpl

expr

oced

ures

)

19T

reat

men

t:co

ntin

ueas

piri

nN

otsp

ecifi

ed0

Tre

atm

ent:

0/19

Tre

atm

ent:

4/19

(21.

1%)

Tre

atm

ent:

0/19

20C

ontr

ol:s

top

aspi

rin

day

�7

Con

trol

:0/2

0C

ontr

ol:2

/20

(10%

)C

ontr

ol:0

/20

*MH

V�

mec

hani

calh

eart

valv

e;A

T�

antit

hrom

botic

;AF

�at

rial

fibri

llatio

n.




Tab

le5—

Ran

dom

ized

Tri

als

Ass

essi

ngP

rohe

mos

tati

cIn

terv

enti

ons

inP

atie

nts

Und

ergo

ing

Den

tal

Pro

cedu

res:

Cli

nica

lD

escr

ipti

onan

dR

esu

lts

(Sec

tion

5.1)

Stud

y/yr

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up

Clin

ical

Out

com

es,N

o./T

otal

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

TT

hera

pyN

o.G

roup

Thr

ombo

embo

lism

Eve

nts

Min

orB

leed

ing

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Al-B

elas

yan

dA

mer

185 /

2003

30V

KA

VT

E�

11,

AF

�7,

Den

tale

xtra

ctio

ns15

Tre

atm

ent:

surg

ical

glue

At

leas

t10

d0

Tre

atm

ent:

0/15

Tre

atm

ent:

0/15

Tre

atm

ent:

0/15

Stro

ke�

4,M

HV

�3,

VH

D�

5

15C

ontr

ol:g

elat

insp

onge

Con

trol

:0/1

5C

ontr

ol:5

/15

(33.

3%)

Con

trol

:0/1

5

Car

ter

and

Gos

s186 /

2003

85V

KA

VT

E�

20,

AF

�14

,C

AD

�10

,V

HD

�33

,st

roke

�1

Den

tale

xtra

ctio

ns43

Tre

atm

ent:

day

�2

tran

exam

icac

idN

otsp

ecifi

ed0

Tre

atm

ent:

0/43

Tre

atm

ent:

2/43

(4.6

%)

Tre

atm

ent:

0/43

52C

ontr

ol:d

ay�

5tr

anex

amic

acid

Con

trol

:0/5

2C

ontr

ol:1

/52

(1.9

%)

Con

trol

:0/5

2

Car

ter

etal

187 /

2003

49V

KA

VT

E�

8,A

F�

12,

CA

D�

5,st

roke

�5,

VH

D�

19

Den

tale

xtra

ctio

ns23

Tre

atm

ent:

surg

ical

glue

10d

0T

reat

men

t:0/

23T

reat

men

t:2/

23(8

.7%

)T

reat

men

t:0/

23

26C

ontr

ol:d

ay�

7tr

anex

amic

acid

Con

trol

:0/2

6C

ontr

ol:0

/26

Con

trol

:0/2

6

Hal

fpen

nyet

al18

8 /20

01

46V

KA

VT

E�

16,

AF

�12

,C

AD

�3,

stro

ke�

2,M

HV

�9,

VH

D�

3

Den

tale

xtra

ctio

ns20

Tre

atm

ent:

surg

ical

glue

(Ber

ipla

st)

Not

spec

ified

0T

reat

men

t:0/

20T

reat

men

t:2/

20(1

0%)

Tre

atm

ent:

0/20

26C

ontr

ol:s

urgi

cal

glue

(Sur

gice

l)C

ontr

ol:0

/26

Con

trol

:1/2

6(3

.8%

)C

ontr

ol:0

/26

Ram

strö

met

al18

9 /19

93

89V

KA

VT

E�

29,

AF

�9,

CA

D�

6,st

roke

�15

,M

HV

�29

,V

HD

�6

Ora

lsur

geri

es(d

enta

lex

trac

tions

,en

dodo

ntic

surg

ery)

44T

reat

men

t:tr

anex

amic

acid

Not

spec

ified

0T

reat

men

t:2/

44(4

.5%

)T

reat

men

t:0/

44T

reat

men

t:0/

44

45C

ontr

ol:s

alin

em

outh

was

hC

ontr

ol:2

/45

(4.4

%)

Con

trol

:10/

45(2

2.2%

)C

ontr

ol:0

/45

*VH

D�

valv

ular

hear

tdi

seas

e;C

AD

�co

rona

ryar

tery

dise

ase;

see

Tab

le4

for

expa

nsio

nof

abbr

evia

tions

.




Tab

le6

—C

ohor

tSt

udi

esA

sses

sing

Con

tinu

atio

nof

Ant

ithr

omb

otic

The

rapy

inP

atie

nts

Und

ergo

ing

Den

tal

Pro

cedu

res:

Cli

nica

lD

escr

ipti

onan

dR

esu

lts

(Sec

tion

5.1)

*

Stud

y/yr

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up

Clin

ical

Out

com

es,N

o./T

otal

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

TT

hera

pyN

o.G

roup

Thr

ombo

embo

lism

Eve

nts

Min

orB

leed

ing

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Blin

der

etal

197/

2001

249

VK

AV

TE

�23

,AF

�53

,D

enta

lext

ract

ions

190

Tre

atm

ent:

INR

�2

Not

spec

ified

0T

reat

men

t:0/

190

Tre

atm

ent:

27/

190

(14.

2%)

Tre

atm

ent:

0/19

0C

ontr

ol:0

/59

CA

D�

27,

VH

D�

112

59C

ontr

ol:I

NR

,1.5

–2.0

Con

trol

:0/5

9C

ontr

ol:3

/59

(5.1

%)

Cam

pbel

let

al190/2

000

25V

KA

Not

repo

rted

Ora

lsur

geri

es(d

enta

lex

trac

tions

,sof

ttis

sue)

12T

reat

men

t:co

ntin

ueV

KA

1d

0T

reat

men

t:1/

12(8

.33%

)T

reat

men

t:0/

12T

reat

men

t:0/

12

13C

ontr

ol:s

top

VK

Ada

y�

3to

�4

Con

trol

:2/1

3(1

5.4%

)C

ontr

ol:0

/13

Con

trol

:0/1

3

Can

non

and

Dha

rmar

191/

2003

70V

KA

VT

E�

15,A

F�

10,

CA

D�

12,

stro

ke�

16,

MH

V�

8,V

HD

�9

Den

tale

xtra

ctio

ns35

Tre

atm

ent:

cont

inue

VK

A5

d0

Tre

atm

ent:

2/35

(5.7

)%T

reat

men

t:0/

35T

reat

men

t:0/

35

35C

ontr

ol:s

top

VK

Ada

y�

2;su

ture

s�

glue

Con

trol

:3/3

5(8

.6%

)C

ontr

ol:0

/35

Con

trol

:0/3

5

Del

laV

alle

etal

198/2

003

40V

KA

MH

V�

40D

enta

lext

ract

ions

40T

reat

men

t:st

opV

KA

;no

cont

rol

Not

spec

ified

0T

reat

men

t:16

/40

(40%

)T

reat

men

t:2/

40(5

.0%

)T

reat

men

t:0/

40

Dev

anie

tal

192/

1998

65V

KA

VT

E�

19,A

F�

12,

CA

D�

12,

stro

ke�

8,M

HV

�1,

VH

D�

16

Den

tale

xtra

ctio

ns33

Tre

atm

ent:

cont

inue

VK

A5

d0

Tre

atm

ent:

1/33

(3.0

%)

Tre

atm

ent:

0/33

Tre

atm

ent:

0/33

32C

ontr

ol:s

top

VK

Ada

y�

2C

ontr

ol:1

/32

(3.1

%)

Con

trol

:0/3

2C

ontr

ol:0

/32

Gas

par

etal

193/

1997

47V

KA

VT

E�

13,

Den

tale

xtra

ctio

ns32

Tre

atm

ent:

cont

inue

VK

A10

d0

Tre

atm

ent:

0/32

Tre

atm

ent:

2/32

(6.2

%)

Tre

atm

ent:

0/32

AF

�2,

MH

V�

23,

stro

ke�

4,V

HD

�2,

CA

D�

3

15C

ontr

ol:s

top

VK

Ada

y�

3C

ontr

ol:0

/15

Con

trol

:1/1

5(6

.7%

)C

ontr

ol:0

/15

Mad

anet

al210/

2005

51A

spir

inV

TE

�18

,CA

D�

6,st

roke

�4,

arte

rial

thro

mbo

embo

lism

�35

,PA

D�

5

Ora

lsur

geri

es(n

otsp

ecifi

ed)

51T

reat

men

t:co

ntin

ueas

piri

n;no

cont

rol

14d

0T

reat

men

t:1/

51(2

.0%

)T

reat

men

t:0/

51T

reat

men

t:0/

51

Mar

tinow

itzet

al199/1

990

40V

KA

VT

E�

7,M

HV

�18

,V

HD

�18

,C

AD

�6,

stro

ke�

4

Den

tale

xtra

ctio

ns40

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

lN

otsp

ecifi

ed0

Tre

atm

ent:

1/40

(2.5

%)

Tre

atm

ent:

0/40

Tre

atm

ent:

0/40




Tab

le6

—C

onti

nued

Stud

y/yr

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up

Clin

ical

Out

com

es,N

o./T

otal

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

TT

hera

pyN

o.G

roup

Thr

ombo

embo

lism

Eve

nts

Min

orB

leed

ing

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Ram

lian

dA

bdul

Rah

man

200/

2005

30V

KA

VT

E�

1,A

F�

9,V

HD

�20

Den

tale

xtra

ctio

ns30

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

l10

d0

Tre

atm

ent:

3/30

(10%

)T

reat

men

t:1/

30(3

.3%

)T

reat

men

t:0/

30

Rus

soet

al201/

2000

104

VK

AM

HV

�10

4,A

F�

42D

enta

lpro

cedu

res

104

Tre

atm

ent:

stop

VK

Ada

y�

2;no

cont

rol

3m

o0

Tre

atm

ent:

0/10

4T

reat

men

t:2/

104

(1.9

%)

Tre

atm

ent:

0/10

4

Zano

net

al196/

2003

500

VK

AV

TE

�33

,AF

�59

,st

roke

�50

,M

HV

�39

,V

HD

�12

,C

AD

�78

,ot

her

�4

Den

tale

xtra

ctio

ns25

0T

reat

men

t:co

ntin

ueV

KA

Not

spec

ified

0T

reat

men

t:0/

250

Tre

atm

ent:

4/25

0(1

.6%

)T

reat

men

t:0/

250

250

Con

trol

:sto

pV

KA

Con

trol

:0/2

50C

ontr

ol:3

/250

(1.2

%)

Con

trol

:0/2

50

Zusm

anet

al202/

1992

23V

KA

AF

�5,

MH

V�

2,V

HD

�11

,M

I�

5

Den

tale

xtra

ctio

ns23

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

lN

otsp

ecifi

ed0

Tre

atm

ent:

1/23

(4.3

%)

Tre

atm

ent:

3/23

(13%

)T

reat

men

t:0/

23

Bar

rero

etal

203/

2002

125

VK

AV

TE

�15

,AF

�57

,st

roke

�18

,M

HV

�10

,PA

D�

2,V

HD

�17

,C

AD

�2,

othe

r�

4

Den

tale

xtra

ctio

ns,

root

leve

rage

,os

teot

omy

125

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

ltr

anex

amic

acid

,pr

essu

re,i

rrig

atio

n,su

rgic

al

Not

spec

ified

0T

reat

men

t:1/

229

(7.9

%)

Tre

atm

ent:

0/22

9T

reat

men

t:1/

229

(0.4

%)

Cie

slik

-Bie

leck

aet

al204/2

005

40V

KA

,asp

irin

AF

�3,

MH

V�

11,

CA

D�

6,V

HD

�3,

CA

D�

15,

othe

r�

5

Den

tale

xtra

ctio

ns,

othe

ror

alsu

rger

y40

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

l;su

ture

s,sp

onge

Not

spec

ified

0T

reat

men

t:0/

40T

reat

men

t:2/

40(5

.0%

)T

reat

men

t:0/

40

Kei

aniM

otla

ghet

al205/2

003

40V

KA

AF

�6,

MH

V�

22,

VT

E�

12D

enta

lext

ract

ions

,ot

her

oral

surg

ery

40T

reat

men

t:co

ntin

ueV

KA

;no

cont

rol

tran

exam

icac

id†

14d

0T

reat

men

t:0/

40T

reat

men

t:0/

40T

reat

men

t:0/

40

Gar

cia-

Dar

enne

set

al206/2

003

96V

KA

MH

V�

56,4

0(V

TE

,C

AD

,str

oke)

Sing

lean

dm

ultip

lede

ntal

extr

actio

ns96

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

ltr

anex

amic

acid

†,su

ture

s,gl

ue

Not

spec

ified

0T

reat

men

t:0/

96T

reat

men

t:3/

96(3

.1%

)T

reat

men

t:0/

96

Saou

ret

al194/

1994

396

VK

AA

F�

39,

MH

V�

396

Den

tale

xtra

ctio

ns15

6T

reat

men

t:co

ntin

ued

VK

AN

otsp

ecifi

ed0

Tre

atm

ent:

0/15

6T

reat

men

t:0/

156

Tre

atm

ent:

0/15

6

240

Con

trol

:dis

cont

inue

VK

Ada

y�

2C

ontr

ol:0

/240

Con

trol

:2/2

40(0

.8%

)C

ontr

ol:0

/240

Stre

etan

dL

eung

195/1

990

14V

KA

Den

tale

xtra

ctio

ns12

Tre

atm

ent:

cont

inue

VK

AN

otsp

ecifi

ed0

Tre

atm

ent:

0/12

Tre

atm

ent:

1/12

(8.3

%)

Tre

atm

ent:

0/12

2C

ontr

ol:s

top

VK

A�

tran

exam

icac

id�

sutu

res

Con

trol

:0/2

Con

trol

:0/2

Con

trol

:0/2

*See

Tab

les

4,5

for

expa

nsio

nof

abbr

evia

tions

.MI

�m

yoca

rdia

linf

arct

ion;

PAD

�pe

riph

eral

arte

rydi

seas

e.†M

outh

was

h.




usually with pressure at the bleeding site, and doesnot require medical attention.

5.1 Dental Procedures

Minor dental procedures assessed consist of singleor multiple tooth extractions and endodontic (rootcanal) procedures. The studies assessing periproce-dural antithrombotic therapy in patients having den-tal procedures are summarized in Tables 4–7.

5.1.1 Patients Who Are Receiving VKAs

Three randomized trials, summarized in Table 4,involving a total of 270 patients compared continuingVKA therapy with interrupting treatment prior to adental procedure.182–184 In these studies, there wereno episodes of thromboembolism or major bleedingwith either perioperative management strategy. Inone trial that compared continuing VKA vs stoppingtreatment 2 days before the procedure, there weremore clinically relevant nonmajor bleeds in patientswho continued VKA therapy (26.3% vs 13.5%).183 Inanother trial that compared continuing VKA therapyin conjunction with coadministered tranexamic acidmouthwash vs stopping VKA therapy 3 days beforethe procedure, there were fewer clinically relevantnonmajor bleeds in patients who continued VKAtherapy (9.2% vs 15.2%).184

Five randomized trials, summarized in Table 5,compared different prohemostatic drugs in a total of299 patients who continued VKA therapy around thetime of a dental procedure.185–189 In these studies,there were no episodes of thromboembolism ormajor bleeding. In one trial which compared tranex-amic acid mouthwash to saline mouthwash in pa-tients who continued VKA therapy, there were fewerclinically relevant nonmajor bleeds in patients whoreceived tranexamic acid (0% vs 22.2%).189 Anothertrial compared treatment with 2 or 5 days of tranex-amic acid mouthwash before the procedure.186 Inthis study, there was a lower incidence of clinicallyrelevant nonmajor bleeding in patients who received5 days of tranexamic acid (1.9% vs 4.6%). In theother trials, continuing VKA therapy while coadmin-istering a prohemostatic agent was associated with noepisodes of major bleeding although the incidence ofclinically relevant nonmajor bleeding varied acrossstudies.185,187,188

Seven prospective cohort studies, summarized inTable 6, assessed bleeding in patients who continuedVKA therapy during dental extraction and in acontrol group of patients who interrupted VKAtherapy before the procedure.190–196 In one studyinvolving 396 patients, of whom 156 continued VKAtherapy and 240 discontinued this treatment, therewere no major bleeds and the incidence of clinically

Tab

le7—

Coh

ort

Stu

dies

Ass

essi

ngP

rohe

mos

tati

cIn

terv

enti

ons

inP

atie

nts

Und

ergo

ing

Den

tal

Pro

cedu

res:

Cli

nica

lD

escr

ipti

onan

dR

esu

lts

(Sec

tion

5.1)

*

Stud

y

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up,

d

Clin

ical

Out

com

es,N

o./T

otal

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

TT

hera

pyN

o.G

roup

Thr

ombo

embo

lism

Eve

nts

Min

orB

leed

ing

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Blin

der

etal

207 /

1999

150

VK

AV

TE

�5,

AF

�26

,C

AD

�29

,V

HD

�70

Den

tale

xtra

ctio

ns50

Tre

atm

ent

1:tr

anex

amic

acid

100

Tre

atm

ent

1:4/

50(8

%)

Tre

atm

ent

1:2/

50(4

%)

Tre

atm

ent

1:0/

50

50T

reat

men

t2:

50co

ntro

l:su

rgic

algl

ue

Tre

atm

ent

2:2/

50(4

%)

Tre

atm

ent

2:2/

50(4

%)

Tre

atm

ent

2:0/

50

-ve

cont

rol0

/50

Con

trol

:3/5

0(6

%)

Con

trol

:0/5

0B

odne

ret

al20

8 /19

9869

VK

AV

TE

�14

,A

F�

23,

MH

V�

32

Den

tale

xtra

ctio

ns69

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

l10

0T

reat

men

t:3/

69(4

%)

Tre

atm

ent:

0/69

Tre

atm

ent:

0/69

*PA

D�

peri

pher

alar

tery

dise

ase;

Tre

atm

ent�

trea

tmen

tgr

oup;

VH

D�

valv

ular

hear

tdi

seas

e;-v

eco

ntro

l�pa

tient

sno

tre

ceiv

ing

antit

hrom

botic

ther

apy.

See

Tab

les

4,5

for

expa

nsio

nof

abbr

evia

tions

..




Tab

le8

—C

ohor

tSt

udi

esA

sses

sing

Con

tinu

atio

nof

Ant

ithr

omb

otic

The

rapy

inP

atie

nts

Und

ergo

ing

Der

mat

olog

icP

roce

dure

s:C

lini

cal

Des

crip

tion

and

Res

ult

s(S

ecti

on5.

2)*

Stud

y

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up

Clin

ical

Out

com

es,N

o./T

otal

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

ntith

rom

botic

The

rapy

No.

Gro

upT

hrom

boem

bolis

mE

vent

s,%

Min

orB

leed

ing

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Alc

alay

212 /

2001

93V

KA

AF

�10

,C

AD

�1,

MH

V�

3,ot

her

�2

Cut

aneo

ussu

rger

ies

16T

reat

men

t:co

ntin

ueV

KA

Not

spec

ified

0T

reat

men

t:0/

16T

reat

men

t:0/

16T

reat

men

t:0/

16

77-v

eco

ntro

lC

ontr

ol:0

/77

Con

trol

:0/7

7C

ontr

ol:0

/77

Bar

tlett

216 /

1999

171

Asp

irin

Not

repo

rted

Cut

aneo

ussu

rger

ies

52T

reat

men

t:co

ntin

ueas

piri

n5–

10d

0T

reat

men

t:1/

52(2

%)

Tre

atm

ent:

2/52

(3.8

%)

Tre

atm

ent:

0/52

119

-ve

cont

rol

Con

trol

:5/1

19(4

.2%

)C

ontr

ol:4

/119

(3.4

%)

Con

trol

:0/1

19

Bill

ings

ley

and

Mal

oney

213 /

1997

306

Not

repo

rted

Cut

aneo

ussu

rger

ies

81T

reat

men

t1:

cont

inue

aspi

rin

1d

0T

reat

men

t1:

17/8

1(2

1%)

Tre

atm

ent

1:1/

81(1

%)

Tre

atm

ent

1:0/

81

VK

A,

aspi

rin

12T

reat

men

t2:

cont

inue

VK

AT

reat

men

t2:

6/12

(50%

)T

reat

men

t2:

1/12

(8%

)T

reat

men

t2:

0/12

213

-ve

cont

rol

Con

trol

:28/

213

(13.

1%)

Con

trol

:1/2

13(0

.5%

)C

ontr

ol:0

/213

Kar

giet

al21

4 /20

0210

2N

otre

port

edC

utan

eous

surg

erie

s37

Tre

atm

ent

1:co

ntin

ueas

piri

n6–

10d

0T

reat

men

t1:

8/37

(21.

6%)

Tre

atm

ent

1:0/

37T

reat

men

t1:

0/37

VK

A,

aspi

rin

21T

reat

men

t2:

cont

inue

VK

AT

reat

men

t2:

7/21

(33.

3%)

Tre

atm

ent

2:5/

21(2

3.8%

)T

reat

men

t2:

0/21

44-v

eco

ntro

lC

ontr

ol:6

/44

(13.

6%)

Con

trol

:0/4

4C

ontr

ol:0

/44

Shal

oman

dW

ong21

7 /20

03

253

Asp

irin

Not

repo

rted

Cut

aneo

ussu

rger

ies

41T

reat

men

t�

cont

inue

das

piri

n3–

6m

o0

Tre

atm

ent-

0/41

Tre

atm

ent:

0/41

Tre

atm

ent:

0/41

212

-ve

cont

rol

Con

trol

:3/2

12(1

.4%

)C

ontr

ol:0

/212

Con

trol

:0/2

12

Syed

etal

215 /

2004

96V

KA

Not

repo

rted

Cut

aneo

ussu

rger

ies

47T

reat

men

t�

cont

inue

dV

KA

Not

spec

ified

0T

reat

men

t:0/

47T

reat

men

t:12

/47

(25.

5%)

Tre

atm

ent:

0/47

49-v

eco

ntro

lC

ontr

ol:0

/49

Con

trol

:3/4

9(6

.1%

)C

ontr

ol:0

/49

*See

Tab

les

4–7

for

expa

nsio

nof

abbr

evia

tions

.




relevant nonmajor or minor bleeding was low inpatients who continued and interrupted VKA ther-apy, at 0% and 0.8%, respectively.194 In anotherstudy that assessed bleeding in 250 patients whounderwent dental extractions during VKA therapyand a control group of 250 patients who had dentalextractions but were not receiving a VKA, there wereno major bleeds reported, and the incidence ofclinically relevant nonmajor bleeding was similar inthe VKA and no VKA groups, of 1.6% and 1.2%,respectively.196 Five additional smaller studies in-volving between 14 and 249 patients provided similarresults, with no major bleeds reported. However,these studies reported incidences of minor bleedingof 0 to 8.3%. In 11 prospective cohort studies,summarized in Table 6, that assessed continuation ofVKA therapy but without a control group, there wereno major bleeds reported.196–206 Two other cohortstudies with a total of 211 patients, summarized inTable 7, assessed different prohemostatic agents inpatients undergoing dental extraction and receiving aVKA.207,208 There were no major bleeds reported,and the incidence of clinically relevant nonmajor andminor bleeding was 0 to 8%.

Taken together, these studies indicate that con-tinuing VKA therapy around the time of a minordental procedure does not confer an increase inclinically important major bleeding. However, thesestudies were not adequately powered to exclude thepossibility that undertaking dental procedures dur-ing VKA therapy confers an increased risk for clini-cally relevant nonmajor or minor bleeding. Untilsuch studies are undertaken, it is reasonable toconsider coadministration of a local prohemostaticagent, which appears to decrease the risk for clini-cally relevant nonmajor and minor bleeding.

5.1.2 Patients Who Are Receiving AntiplateletDrugs

In one small randomized trial of 39 patients thatcompared continuing or interrupting aspirin before adental procedure, there were no major bleeds inboth treatment arms but the incidence of clinicallyrelevant nonmajor bleeding was higher in patientswho continued aspirin therapy (21% vs 10%).209 In acohort study involving 51 patients who continuedaspirin therapy around the time of a dental procedure,there were no major bleeds and clinically relevantnonmajor bleeding occurred in one patient.210

Studies are lacking in regard to patients who arereceiving clopidogrel and require a dental proce-dure, although it is probable that the continuation ofclopidogrel and aspirin in patients undergoing dentalprocedures will increase the risk of bleeding abovethat seen with aspirin alone.211 In high-risk patients

with a coronary stent implanted within the past year,or perhaps at any time in the past in the case of adrug-eluting stent, the risk of stent thrombosis withclopidogrel interruption probably outweighs the riskof procedure-related bleeding associated with con-tinuation of treatment. In lower-risk patients withouta coronary stent, periprocedural management isuncertain, although it is probably reasonable tointerrupt clopidogrel therapy and to continue aspiringiven that combined antiplatelet treatment will in-crease bleeding risk above that of the risk with eitherdrug alone.

Recommendation

5.1. In patients who are undergoing minordental procedures and are receiving VKAs, werecommend continuing VKAs around the timeof the procedure and coadministering an oralprohemostatic agent (Grade 1B). In patientswho are undergoing minor dental proceduresand are receiving aspirin, we recommendcontinuing aspirin around the time of theprocedure (Grade 1C). In patients who areundergoing minor dental procedures and arereceiving clopidogrel, please refer to the rec-ommendations outlined in Section 4.5 andSection 4.6.

5.2 Dermatologic Procedures

Minor dermatologic procedures assessed includeexcisions of basal and squamous cell carcinomas,actinic keratoses, and malignant or premalignantnevi. Studies of periprocedural antithrombotic man-agement are summarized in Table 8.


Four cohort studies assessed continuing VKA ther-apy around the time of a dermatologic procedure ina total of 96 patients.212–215 In one study that as-sessed 47 patients who continued VKA therapy (and49 control patients who were not receiving VKAtherapy at the time of procedure), there were nomajor bleeds but the incidence of clinically rele-vant nonmajor bleeding was higher in patientswho continued VKA therapy (25.5% vs 6.1%).215

There were similar findings in two other cohortstudies,213,214 whereas no bleeds were reported ina third study involving 16 patients who continuedVKA therapy.212


Four cohort studies assessed continuing aspirintherapy around the time of a dermatologic procedure




in a total of 211 patients.213,214,216,217 There were nomajor bleeds associated with continuing aspirin. Intwo studies, minor bleeding was more frequent inpatients who continued aspirin around the time of theprocedure compared to patients who were not receiv-ing aspirin (21% vs 13%, respectively; 21.6% vs 13.6%,respectively).213,214 However, in two other cohort stud-ies the incidence of minor bleeding appeared compa-rable in patients who received or who did not receiveaspirin around the time of the procedure (1.9% vs4.2%, respectively; 0% vs 1.4%, respectively).216,217

Data for patients who are receiving clopidogreland require dermatologic procedures are limited tocase reports of patients who developed thromboem-bolic events during antiplatelet therapy interrup-tion.218 Periprocedural management in such patientscan follow that in patients undergoing dental orother minor procedures.

Recommendation

5.2. In patients who are undergoing minor der-matologic procedures and are receiving VKAs, werecommend continuing VKAs around the time ofthe procedure (Grade 1C). In patients who areundergoing minor dermatologic procedures andare receiving aspirin, we recommend continuingaspirin around the time of the procedure (Grade1C). In patients who are undergoing minor der-matologic procedures and are receiving clopi-dogrel, please refer to the recommendations out-lined in Section 4.5 and Section 4.6.

5.3 Ophthalmologic Procedures

Minor ophthalmologic procedures assessed aredominated by cataract extraction, which was under-taken in � 90% of patients studied. A small minorityof patients studied had vitreoretinal or other oph-thalmologic procedures. Our recommendations,therefore, pertain to patients undergoing cataractextraction. These studies are summarized in Table 9.


Six prospective cohort studies assessed bleeding inpatients who continued VKA therapy during oph-thalmologic surgery and in a control group of pa-tients who were either not receiving VKA therapy orwho interrupted VKA therapy before surgery.219–224

Two other prospective cohort studies assessed bleed-ing in patients who continued VKA therapy duringophthalmologic surgery but did not have a controlgroup.225,226 In one prospective cohort study assess-ing patients who had cataract surgery, there was noapparent increase in arterial thromboembolic events

in 208 patients who discontinued VKAs compared to526 patients who continued VKAs and the incidenceof such events appeared higher in patients whocontinued VKAs (1.14% vs 0.48%).221 In these pa-tients, there were no major or clinically relevantnonmajor bleeds. In another cohort study involving639 patients who continued VKAs and 1,203 controlswho were not taking VKAs around the time ofcataract surgery, there were no arterial thromboem-bolic events.219 There appeared to be a higherincidence of clinically relevant nonmajor bleeding(0.16% vs 0.08%) and minor bleeding (1.41% vs0.67%) in patients who continued VKAs althoughthere were no major bleeds reported. While othersmaller cohort studies demonstrated similar re-sults,225,226 one cohort study involving 125 patientswho had cataract surgery reported a high rate ofmajor bleeding (8.7%).224


A prospective cohort study assessing patients whounderwent cataract surgery found no important in-crease in arterial thromboembolic events in 977patients who interrupted aspirin compared to 3,363patients who continued aspirin (0.20% vs 0.65%).221

In these patients, there were no major or clinicallyrelevant nonmajor bleeds and marginally higherclinically relevant nonmajor bleeds in patients whocontinued aspirin (0.06% vs 0%). Other studiesreported similar results in patients undergoing cata-ract or vitreoretinal surgery.222,223

There are few data in regard to the safety ofcontinuing clopidogrel in patients undergoing oph-thalmologic surgery. One study of patients undergo-ing cataract surgery found that although subconjunc-tival hemorrhage was more common in patients whowere receiving either clopidogrel or warfarin thanaspirin or no antithrombotic drugs, there were nosight-threatening bleeding complications.227 Onestudy described a patient who was receiving aspirinand clopidogrel and underwent an intracapsularextraction and anterior vitrectomy in whom thepostoperative course was complicated by extensivehyphema and vitreous hemorrhage that clearedwithin 3 months.228 As with other minor procedures,perioperative management will be driven by throm-boembolic risk.

Recommendation

5.3. In patients who are undergoing cataractremoval and are receiving VKAs, we recom-mend continuing VKAs around the time of theprocedure (Grade 1C). In patients who are un-




Tab

le9

—C

ohor

tSt

udi

esA

sses

sing

Con

tinu

atio

nof

Ant

ithr

omb

otic

The

rapy

inP

atie

nts

Und

ergo

ing

Oph

thal

mol

ogic

Surg

ery:

Cli

nica

lD

escr

ipti

onan

dR

esu

lts

(Sec

tion

5.3)

*

Stud

y

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up

Clin

ical

Out

com

es,N

o./T

otal

(%)

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

ntith

rom

botic

The

rapy

No.

Gro

upT

hrom

boem

bolis

mE

vent

sM

inor

Ble

edin

g

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Hir

schm

anan

dM

orby

219/2

006

1,84

2V

KA

Not

repo

rted

Cat

arac

tsu

rger

y63

9T

reat

men

t:co

ntin

ueV

KA

10d

0T

reat

men

t-9/

639

�1.

4%T

reat

men

t:1/

63(0

.16%

)T

reat

men

t:0/

639

1,20

3-v

eco

ntro

lC

ontr

ol:8

/120

3(0

.7%

)C

ontr

ol:1

/120

3(0

.08%

)C

ontr

ol:0

/120

3

Kal

lioet

al220/

2000

1,16

7V

KA

,asp

irin

Not

repo

rted

Cat

arac

tsu

rger

y(�

80%

)

482

Tre

atm

ent

1:st

opas

piri

nda

y�

3ltr

eatm

ent

2:st

opV

KA

day

�2

Not

spec

ified

0T

reat

men

t1:

18/

482

(3.7

%);

trea

tmen

t2:

3/76

(3.9

%)

Tre

atm

ent

1:0/

482;

trea

tmen

t2:

0/76

Tre

atm

ent

1:0/

482;

trea

tmen

t2:

0/76

76-v

eco

ntro

lC

ontr

ol:2

5/60

9(4

.1%

)C

ontr

ol:0

/609

Con

trol

:0/6

09

Kat

z etal

221/

2003

37,6

11V

KA

,asp

irin

VT

E�

206,

CA

D�

793,

stro

ke�

738,

VH

D�

248,

CA

D�

2,04

0,ot

her

�16

9

Cat

arac

tsu

rger

y3,

363

Tre

atm

ent

1:co

ntin

ueas

piri

n10

dT

reat

men

t1:

22/

3,36

3(0

.65%

)T

reat

men

t1:

0/3,

363

Tre

atm

ent

1:2/

3,36

3(0

.06%

)T

reat

men

t1:

0/3,

363

977

Tre

atm

ent

2:st

opas

piri

nT

reat

men

t2:

2/97

7(0

.20%

)T

reat

men

t2:

0/97

7T

reat

men

t2:

0/97

7T

reat

men

t2:

0/97

714

,322

-ve

cont

rola

spir

inC

ontr

olas

piri

n:33

/14,

322

(0.2

3%)

Con

trol

aspi

rin:

3/14

,322

(0.0

2%)

Con

trol

aspi

rin:

5/14

,322

(0.0

3%)

Con

trol

aspi

rin:

0/14

,322

526

Tre

atm

ent

3:co

ntin

ueV

KA

Tre

atm

ent

4:6/

526

(1.1

4%)

Tre

atm

ent

4:0/

526

Tre

atm

ent

4:0/

526

Tre

atm

ent

4:0/

526

208

Tre

atm

ent

4:st

opV

KA

Tre

atm

ent

5:1/

208

(0.4

8%)

Tre

atm

ent

5:0/

208

Tre

atm

ent

5:0/

208

Tre

atm

ent

5:0/

208

18,2

15-v

eco

ntro

lVK

AC

ontr

olV

KA

:52/

18,2

15(0

.29%

)C

ontr

olV

KA

:3/

18,2

15(0

.02%

)C

ontr

olV

KA

:7/

18,2

15(0

.04%

)C

ontr

olV

KA

:0/

18,2

15L

umm

ean

dL

aatik

aine

n222/

1994

351

VK

A,a

spir

inN

otre

port

edC

atar

act

surg

ery

54T

reat

men

t1:

cont

inue

aspi

rin

At

leas

t3

d0

Tre

atm

ent

1:6/

54(1

1.1%

)T

reat

men

t1:

0/54

Tre

atm

ent

1-0/

54

17T

reat

men

t2:

stop

VK

AT

reat

men

t2:

0/17

Tre

atm

ent

2:3/

17(1

7.6%

)T

reat

men

t2-

0/17

208

-ve

cont

rol

Con

trol

:21/

208

(10.

1%)

Con

trol

:0/2

08C

ontr

ol-

0/20

8

Nar

endr

anan

dW

illia

mso

n223/

2003

541

VK

A,a

spir

inV

TE

�1,

Vitr

eore

tinal

surg

ery

60T

reat

men

t1:

cont

inue

aspi

rin

Not

spec

ified

0T

reat

men

t1:

0/60

Tre

atm

ent

1:2/

60(3

.3%

)T

reat

men

t1-

0/60

AF

�2,

7T

reat

men

t2:

stop

VK

AT

reat

men

t2:

1/7

(14.

3%)

Tre

atm

ent

2:1/

7(1

4.3%

)T

reat

men

t2-

1/7

�14

.3%

MH

V�

3,O

ther

�1

474

-ve

cont

rol

Con

trol

:10/

474

(2.1

%)

Con

trol

:0/4

74C

ontr

ol:0

/474

Rob

erts

etal

225/

1991

31V

KA

,asp

irin

,di

pyri

dam

ole,

sulfi

n-py

razo

ne

Not

repo

rted

Cat

arac

tsu

rger

y31

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

l2

mo

0T

reat

men

t:0/

31T

reat

men

t:1/

31(3

.2%

)T

reat

men

t-0/

31

Rot

enst

reic

het

al226/2

001

35V

KA

VT

E�

4,A

F�

27,V

HD

�4

Cat

arac

tsu

rger

y35

Tre

atm

ent:

cont

inue

VK

A;n

oco

ntro

l30

d0

Tre

atm

ent:

2/35

(5.7

%)

Tre

atm

ent:

0/35

Tre

atm

ent:

0/35




dergoing cataract removal and are receivingaspirin, we recommend continuing aspirinaround the time of the procedure (Grade 1C). Inpatients who are undergoing cataract removaland are receiving clopidogrel, please refer tothe recommendations outlined in Section 4.5and Section 4.6.

6.0 Perioperative Management ofAntithrombotic Therapy Patients Who

Require Urgent Surgical or OtherInvasive Procedures

6.1 Patients Who Are Receiving VKAs

In the nonbleeding patient who requires rapid(within 12 h) reversal of the anticoagulant effect ofVKAs because of an urgent surgical or other invasiveprocedure, treatment options that have been as-sessed in observational studies include fresh-frozenplasma, prothrombin concentrates, and recombinantfactor VIIa.229 No randomized trials, to date and toour knowledge, have compared these treatments inpatients who require urgent reversal of anticoagula-tion.230 In addition to these treatment options, allpatients should receive vitamin K, at a dose of 2.5 to5.0 mg po or by slow IV infusion.231 Administeringfresh-frozen plasma, prothrombin concentrates, orrecombinant factor VIIa alone will temporarily over-ride but will not eliminate the anticoagulant effect ofVKAs, which persist until VKAs are endogenouslymetabolized or neutralized by vitamin K. For exam-ple, as fresh-frozen plasma has an elimination half-life of 4 to 6 h, not administering vitamin K will leadto reemergence of a VKA-associated anticoagulanteffect within 12 to 24 h. If surgery is urgent but canbe delayed for 18 to 24 h, the anticoagulant effect ofVKAs is likely to be neutralized by IV vitamin K, ata dose of 2.5 to 5.0 mg without the need for bloodproduct or recombinant factor VII administra-tion.230,232

Recommendation

6.1. In patients who are receiving VKAs andrequire reversal of the anticoagulant effect foran urgent surgical or other invasive procedure,we suggest treatment with low-dose (2.5 to 5.0mg) IV or oral vitamin K (Grade 1C). For moreimmediate reversal of the anticoagulant effect,we suggest treatment with fresh-frozen plasmaor another prothrombin concentrate in additionto low-dose IV or oral vitamin K (Grade 2C).

6.2 Patients Who Are Receiving Antiplatelet Drugs

There is no pharmacologic agent that can reversethe antithrombotic effect of aspirin, clopidogrel, or

Tab

le9

—C

onti

nued

Stud

y

Patie

nts

Typ

eof

Proc

edur

e

Peri

proc

edur

alIn

terv

entio

n

Fol

low

-up

Clin

ical

Out

com

es,N

o./T

otal

(%)

No.

Typ

eof

Tre

atm

ent

Indi

catio

nfo

rA

ntith

rom

botic

The

rapy

No.

Gro

upT

hrom

boem

bolis

mE

vent

sM

inor

Ble

edin

g

Clin

ical

lyR

elev

ant

Non

maj

orB

leed

ing

Maj

orB

leed

Wir

bela

uer

etal

224/2

004

128

VK

AV

TE

�9,

AF

�11

,C

atar

act

surg

ery

103

Tre

atm

ent:

cont

inue

VK

AN

otsp

ecifi

ed0

Tre

atm

ent:

0/10

3T

reat

men

t:0/

103

Tre

atm

ent:

9/10

3�

8.7%

MH

V�

14,C

AD

�17

,VH

D�

15,o

ther

�26

19C

ontr

ol:s

top

VK

A16

dpr

ior

Con

trol

:0/1

9C

ontr

ol:0

/19

Con

trol

:0/1

9

*See

Tab

les

4–7

for

expa

nsio

nof

abbr

evia

tions

.




ticlopidine, which irreversibly inhibit platelet func-tion. Consequently, patients who require an urgentsurgical or other invasive procedure that requiresnormalized platelet function may receive transfusedplatelets, which would not be affected by prioradministration of antiplatelet drugs.233 However, theefficacy and safety of platelet transfusion in patientswho are not thrombocytopenic and who require anurgent surgical or other invasive procedure are notknown. One randomized trial in 11 healthy volun-teers who received aspirin (325 mg loading dose, 81mg maintenance dose) and clopidogrel (300-mg or600-mg loading dose, 75-mg maintenance dose)found that subsequent transfusion of 12.5 U plateletsled to normalized platelet function as determined byplatelet function assays.234 However, studies to as-sess the efficacy and safety of a platelet transfusion toneutralize the antiplatelet effects of aspirin or clopi-dogrel in the perioperative setting are lacking. Untilsuch studies are done, it is reasonable to limitplatelet transfusion to those patients who have ex-cessive or life-threatening bleeding in the perioper-ative period.

Potential alternatives to platelet transfusion inpatients who have been exposed to antiplateletdrugs are prohemostatic agents. These includeε-aminocaproic acid and tranexamic acid, which areantifibrinolytic agents, and 1-deamino-8-D-argininevasopressin, which increases plasma levels of vonWillebrand factor and associated coagulation factorVIII. These agents may improve platelet function inpatients who have been exposed to antiplateletdrugs.235 However, outside of the setting of cardiacsurgery, these drugs have not been widely stud-ied113,236 and should be limited to patients who haveexcessive or life-threatening perioperative bleedingbecause of potential prothrombotic effects.

Recommendation

6.2. For patients receiving aspirin, clopidogrel,or both, are undergoing surgery and have ex-cessive or life-threatening perioperative bleed-ing, we suggest transfusion of platelets or ad-ministration of other prohemostatic agents(Grade 2C).

Conflict of Interest Disclosures

Dr. Ansell discloses that he has received consultant fees fromBristol-Myers Squibb, Roche Diagnostics, and InternationalTechnidyne Corporation. He is also on the speakers bureau forRoche Diagnostic Corporation and Sanofi-Aventis, and is the pastpresident of the Anticoagulation Forum.

Dr. Douketis reveals no real or potential conflicts of interestor commitment.

Dr. Dunn discloses that he received grant monies from and ison the speakers bureau for Sanofi-Aventis. He has also served onadvisory committees for Sanofi-Aventis, Eisai, and Roche Diag-nostics.

Dr. Jaffer discloses that he has received consultant fees fromSanofi-Aventis and AstraZeneca, and that he is on the speakersbureau for Sanofi-Aventis.

Dr. Becker reveals no real or potential conflicts of interest orcommitment.

Dr. Spyropoulos discloses that he has received consultantfees from Boehringer Ingelheim, and has served on the speakersbureau for Sanofi-Aventis and Eisai.

Dr. Berger discloses that he has spoken at Council on MedicalEducation-approved scientific symposia supported by Bristol-Myers Squibb, Sanofi-Aventis, the Medicines Company, Astra-Zeneca, Medtronic, Schering-Plough, Lilly, and Daiichi Sankyo.He has served as a consultant for PlaCor, Lilly, Daiichi Sankyo,Molecular Insight Pharmaceuticals, and CV Therapeutics. Dr.Berger also owns equity in Lumen, Inc (a company that isdeveloping an embolic protection device).

References1 Ansell JE. The perioperative management of warfarin ther-

apy. Arch Intern Med 2003; 163:881–8832 American Heart Association. Heart Disease and Stroke

Statistics Update. Dallas, TX: American Heart Association,2001

3 Go AS, Hylek EM, Borowsky LH, et al. Warfarin use amongambulatory patients with nonvalvular atrial fibrillation: theanticoagulation and risk factors in atrial fibrillation (ATRIA)study. Ann Intern Med 1999; 131:927–934

4 White RH, Kaatz S, Douketis J, et al. Comparison of the30-day incidence of ischemic stroke and bleeding aftermajor surgery in patients with or without atrial fibrillation(AF). J Thromb Haemost 2007; 5(suppl 1):O-M-035

5 McKenna R. Abnormal coagulation in the postoperativeperiod contributing to excessive bleeding. Med Clin NorthAm 2001; 85:1277–1310

6 Tinker JH, Tarhan S. Discontinuing anticoagulant therapy insurgical patients with cardiac valve prostheses: observationsin 180 operations. JAMA 1978; 239:738–739

7 Katholi RE, Nolan SP, McGuire LB. The management ofanticoagulation during noncardiac operations in patientswith prosthetic heart valves: a prospective study. Am Heart J1978; 96:163–165

8 Torn M, Rosendaal FR. Oral anticoagulation in surgicalprocedures: risks and recommendations. Br J Haematol2003; 123:676–682

9 Anonymous. Prevention of pulmonary embolism and deepvein thrombosis with low dose aspirin: Pulmonary EmbolismPrevention (PEP) trial. Lancet 2000; 355:1295–1302

10 Dunn AS, Turpie AG. Perioperative management of pa-tients receiving oral anticoagulants: a systematic review.Arch Intern Med 2003; 163:901–908

11 Kaplan RC, Tirschwell DL, Longstreth WT Jr, et al. Vascu-lar events, mortality, and preventive therapy following isch-emic stroke in the elderly. Neurology 2005; 65:835–842

12 Longstreth WT Jr, Bernick C, Fitzpatrick A, et al. Fre-quency and predictors of stroke death in 5,888 participantsin the Cardiovascular Health Study. Neurology 2001; 56:368–375

13 Martinelli J, Jiminez A, Rabago G, et al. Mechanical cardiacvalve thrombosis: is thrombectomy justified? Circulation1991; 84(suppl):70S–75S

14 Kini AS, Lee P, Marmur JD, et al. Correlation of postper-




cutaneous coronary intervention creatine kinase-MB andtroponin I elevation in predicting mid-term mortality. Am JCardiol 2004; 93:18–23

15 Landesberg G, Einav S, Christopherson R, et al. Perioper-ative ischemia and cardiac complications in major vascularsurgery: importance of the preoperative twelve-lead electro-cardiogram. J Vasc Surg 1997; 26:570–578

16 Stone GW, Aronow HD. Long-term care after percutaneouscoronary intervention: focus on the role of antiplatelettherapy. Mayo Clin Proc 2006; 81:641–652

17 Popma JJ, Berger P, Ohman EM, et al. Antithrombotictherapy during percutaneous coronary intervention: theSeventh ACCP Conference on Antithrombotic and Throm-bolytic Therapy. Chest 2004; 126(suppl):576S–599S

18 Yang X, Alexander KP, Chen AY, et al. The implications ofblood transfusions for patients with non-ST-segment eleva-tion acute coronary syndromes: results from the CRUSADENational Quality Improvement Initiative. J Am Coll Cardiol2005; 46:1490–1495

19 Linkins LA, Choi PT, Douketis JD. Clinical impact ofbleeding in patients taking oral anticoagulant therapy forVTE: a meta-analysis. Ann Intern Med 2003; 139:893–900

20 Rao SV, Jollis JG, Harrington RA, et al. Relationship ofblood transfusion and clinical outcomes in patients withacute coronary syndromes. JAMA 2004; 292:1555–1562

21 Douketis JD, Johnson JA, Turpie AG. Low-molecular-weight heparin as bridging anticoagulation during interrup-tion of warfarin: assessment of a standardized periproce-dural anticoagulation regimen. Arch Intern Med 2004;164:1319–1326

22 Kovacs MJ, Kearon C, Rodger M, et al. Single-arm study ofbridging therapy with low-molecular-weight heparin forpatients at risk of arterial embolism who require temporaryinterruption of warfarin. Circulation 2004; 110:1658–1663

23 Dagi TF. The management of postoperative bleeding. SurgClin North Am 2005; 85:1191–1213

24 Litaker D. Preoperative screening. Med Clin North Am1999; 83:1565–1581

25 Jones HU, Muhlestein JB, Jones KW, et al. Preoperative useof enoxaparin compared with unfractionated heparin in-creases the incidence of re-exploration for postoperativebleeding after open-heart surgery in patients who presentwith an acute coronary syndrome: clinical investigation andreports. Circulation 2002; 106:I19–I22

26 Mangano DT. Multicenter Study of Perioperative IschemiaResearch: aspirin and mortality from coronary bypass sur-gery. N Engl J Med 2002; 347:1309–1317

27 Lazio BE, Sumard JM. Anticoagulation in neurosurgicalpatients. Neurosurgery 1999; 45:838–847

28 Patterson BM, Marchand R, Ranawat C. Complications ofheparin therapy after total joint arthroplasty. J Bone JointSurg Am 1989; 71:1130–1134

29 Hoy E, Granick M, Benevenia J, et al. Reconstruction ofmusculoskeletal defects following oncologic resection in 76patients. Ann Plast Surg 2006; 57:190–194

30 Nielson JD, Gram J, Holm-Nielsen A, et al. Postoperativeblood loss after transurethral prostatectomy is dependent onin-situ fibrinolysis. Br J Urol 1997; 889–893

31 Watson CJ, Deane AM, Doyle PT, et al. Identifiable factorsin post-prostatectomy haemorrhage: the role of aspirin. Br JUrol 1990; 66:85–87

32 Sorbi D, Norton I, Conio M, et al. Postpolypectomy lowerGI bleeding: descriptive analysis. Gastrointest Endosc 2000;51:690–696

33 Ihezue CU, Smart J, Dewbury KC, et al. Biopsy of theprostate guided by transrectal ultrasound: relation between

warfarin use and incidence of bleeding complications. ClinRadiol 2005; 60:459–463

34 Wiegand UK, LeJeune D, Boguschewski F, et al. Pockethematoma after pacemaker or implantable cardioverterdefibrillator surgery: influence of patient morbidity, opera-tion strategy, and perioperative antiplatelet/anticoagulationtherapy. Chest 2004; 126:1177–1186

35 Wessler S, Gitel SN. Pharmacology of heparin and warfarin.J Am Coll Cardiol 1986; 8:10B–20B

36 Hirsh J, Raschke R. Heparin and low-molecular-weightheparin: the Seventh ACCP Conference on Antithromboticand Thrombolytic Therapy. Chest 2004; 126(suppl):188S–203S

37 Pedersen AK, FitzGerald GA. Dose-related kinetics ofaspirin: presystemic acetylation of platelet cyclooxygenase.N Engl J Med 1984; 311:1206–1211

38 Savcic M, Hauert J, Bachmann F, et al. Clopidogrel loadingdose regimens: kinetic profile of pharmacodynamic responsein healthy subjects. Semin Thromb Haemost 1999; 25(suppl2):15–19

39 Dunn AS, Spyropoulos A, Turpie AG. Bridging therapy inpatients on long-term oral anticoagulants who require sur-gery: the Prospective Peri-operative Enoxaparin CohortTrial (PROSPECT). J Thromb Haemost 2007; 5:2211–2218

40 Turpie AG, Douketis JD. Enoxaparin is effective and safe asbridging anticoagulation in patients with a mechanical pros-thetic heart valve who reqire temporary interruption ofwarfarin because of surgery or an invasive procedure [ab-stract]. Blood 2004; 104:202A

41 Carrel TP, Klingenmann W, Mohacsi PJ, et al. Perioperativebleeding and thromboembolic risk during non-cardiac sur-gery in patients with mechanical prosthetic heart valves: aninstitutional review. J Heart Valve Dis 1999; 8:392–398

42 Larson BJ, Zumberg MS, Kitchens CS. A feasibility study ofcontinuing dose-reduced warfarin for invasive procedures inpatients with high thromboembolic risk. Chest 2005; 127:922–927

43 White RH, McKittrick T, Hutchinson R, et al. Temporarydiscontinuation of warfarin therapy: changes in the interna-tional normalized ratio. Ann Intern Med 1995; 122:40–42

44 Palareti G, Legnani C. Warfarin withdrawal: pharmacokinetic-pharmacodynamic considerations. Clin Pharmacokinet 1996;30:300–313

45 Salem DN, Stein PD, Al-Ahmad A, et al. Antithrombotictherapy in valvular heart disease–native and prosthetic: theSeventh ACCP Conference on Antithrombotic and Throm-bolytic Therapy. Chest 2004; 126(suppl):457S–482S

46 Hylek EM, Regan S, Go AS, et al. Clinical predictors ofprolonged delay in return of the international normalizedratio to within the therapeutic range after excessive antico-agulation with warfarin. Ann Intern Med 2001; 135:393–400

47 Gadisseur AP, van der Meer FJ, Adriaansen HJ, et al.Therapeutic quality control of oral anticoagulant therapycomparing the short-acting acenocoumarol and the long-acting phenprocoumon. Br J Haematol 2002; 117:940–946

48 Marietta M, Bertesi M, Simoni L, et al. A simple and safenomogram for the management of oral anticoagulation priorto minor surgery. Clin Lab Haematol 2003; 25:127–130

49 Deerhake JP, Merz JC, Cooper JV, et al. The duration ofanticoagulation bridging therapy in clinical practice maysignificantly exceed that observed in clinical trials. J ThrombThrombolysis 2007; 23:107–113

50 Woods K, Douketis JD, Kathirgamanathan K, et al. Low-dose oral vitamin K to normalize the international normal-ized ratio prior to surgery in patients who require temporaryinterruption of warfarin. J Thromb Thrombolysis 2007;24:93–97




51 Cannegieter SC, Rosendaal FR, Briet I. Thromboembolicand bleeding complications in patients with medical me-chanical heart valve prostheses. Circulation 1994; 94:635–641

52 Herin D, Piper C, Bergemann R, et al. Thromboembolicand bleeding complications following St. Jude medical valvereplacement: results of the German Experience with Low-Intensity Anticoagulation Study. Chest 2005; 127:53–59

53 Tominaga R, Kurisu K, Ochiai Y, et al. A 10-year experiencewith the Carbomedics cardiac prosthesis. Ann Thorac Surg2005; 79:784–789

54 Galla JM, Fuhs BE. Outpatient anticoagulation protocol formechanical valve recipients undergoing non-cardiac surgery[abstract]. J Am Coll Cardiol 2000; 35:531A

55 Spyropoulos AC, Turpie AGG, Dunn AS, et al. Clinicaloutcomes with unfractionated heparin or low-molecular-weight heparin as bridging therapy in patients on long-termoral anticoagulants: the REGIMEN registry. J ThrombHaemost 2006; 4:1246–1252

56 Ferreira I, Dos L, Tornos P, et al. Experience with enox-aparin in patients with mechanical heart valves who mustwithhold acenocumarol. Heart 2003; 89:527–530

57 Jaffer AK, Ahmed M, Brotman DJ, et al. Low-molecular-weight-heparins as periprocedural anticoagulation for pa-tients on long-term warfarin therapy: a standardized bridg-ing therapy protocol. J Thromb Thrombolysis 2005; 20:11–16

58 Hammerstingl C, Tripp C, Schmidt H, et al. Periproceduralbridging therapy with low-molecular-weight heparin inchronically anticoagulated patients with prosthetic mechan-ical heart valves: experience in 116 patients from theprospective BRAVE registry. J Heart Valve Dis 2007; 16:285–292

59 Santamaria MA, Mateo J, Pujol N, et al. Management ofanticoagulation in patients who require colonoscopy orgastroscopy: saftey and efficacy of sodic bemiparin [Hibor][abstract]. Blood 2004; 105:4060

60 Spandorfer JM, Lynch S, Weitz HH, et al. Use of enoxaparinfor the chronically anticoagulated patient before and afterprocedures. Am J Cardiol 1999; 84:478–480

61 Tinmouth AH, Morrow BH, Cruickshank MK, et al. Dalte-parin as periprocedure anticoagulation for patients on war-farin and at high risk of thrombosis. Ann Pharmacother2001; 35:669–674

62 Wilson S, Morgan J, Gray L, et al. A model for perioperativeoutpatient management of anticoagulation in high-risk pa-tients: an evaluation of effectiveness and safety. Can J HospPharm 2001; 54:269–277

63 Malato A, Anastasio R, Cigna V, et al. Perioperative bridgingtherapy with low molecular weight heparin in patientsrequiring interruption of long-term oral anticoagulant ther-apy. Haematologica 2006; 91:10

64 Constans M, Santamaria A, Mateo J, et al. Low-molecular-weight heparin as bridging therapy during interruption oforal anticoagulation in patients who require colonoscopy orgastroscopy. Int J Clin Pract 2007; 61:212–217

65 Geerts WH, Pineo GF, Heit JA, et al. Prevention of venousthromboembolism: the Seventh ACCP Conference on An-tithrombotic and Thrombolytic Therapy. Chest 2004;126(suppl):338S–400S

66 Kearon C, Hirsh J. Management of anticoagulation beforeand after elective surgery. N Engl J Med 1997; 336:1506–1511

67 Baudet EM, Oca CC, Roques XF, et al. A 5/12 yearexperience with the St. Jude Medical cardiac valve prosthe-sis: early and late results of 737 valve replacements in 671patients. J Thorac Cardiovasc Surg 1985; 90:137–144

68 Hart RG, Benavente O, McBride R, et al. Antithrombotictherapy to prevent stroke in patients with atrial fibrillation: ameta-analysis. Ann Intern Med 1999; 131:492–501

69 Pearce LA, Hart RG, Halperin JL. Assessment of threeschemes for stratifying stroke risk in patients with nonval-vular atrial fibrillation. Am J Med 2000; 109:45–51

70 Gage BF, Waterman AD, Shannon W, et al. Validation ofclinical classification schemes for predicting stroke: resultsfrom the National Registry of Atrial Fibrillation. JAMA2001; 285:2864–2870

71 Gage BF, Van Walraven C, Pearce L, et al. Selectingpatients with atrial fibrillation for anticoagulation: stroke riskstratification in patients taking aspirin. Circulation 2004;110:2287–2292

72 Baudo F, de Cataldo F, Mostarda G, et al. Management ofpatients on long-term oral anticoagulant therapy undergoingelective surgery: survey of the clinical practice in the Italiananticoagulation clinics. Intern Emerg Med 2007; 2:280–284

73 Garcia DA, Regan S, Henault L, et al. Risk of thromboem-bolism with short-term interruption of warfarin therapyArch Intern Med 2008; 168:63–69

74 Blacker DJ, Wijdicks EF, McClelland RL. Stroke risk inanticoagulated patients with atrial fibrillation undergoingendoscopy. Neurology 2003; 61:964–968

75 Douketis JD, Gu CS, Schulman S, et al. The risk of fatalpulmonary embolism after stopping anticoagulant therapy inpatients with VTE. Ann Intern Med 2007; 147:766–774

76 Douketis JD, Kearon C, Bates S, et al. The risk of fatalpulmonary embolism in patients with treated venous throm-boembolism. JAMA 1998; 279:458–462

77 Douketis JD, Foster GA, Crowther MA, et al. Clinical riskfactors and timing of recurrent venous thromboembolismduring the initial 3 months of anticoagulant therapy. ArchIntern Med 2000; 160:3431–3436

78 Kyrle PA, Eichinger S. Deep vein thrombosis. Lancet 2005;365:1163–1174

79 White R. The epidemiology of venous thromboembolism.Circulation 2003; 107(suppl 1):14–18

80 Spyropoulos AC, Dunn A, Turpie AG, et al. Perioperativebridging therapy with unfractionated heparin or low-molec-ular-weight heparin in patients with mechanical heart valveson long-term oral anticoagulatns: results from the REGI-MEN registry [abstract]. J Am Coll Cardiol 2005;45:352A

81 Halbritter KM, Wawer A, Beyer J, et al. Bridging anticoag-ulation for patients on long-term vitamin-K-antagonists: aprospective 1 year registry of 311 episodes. J ThrombHaemost 2007; 3:2823–2825

82 Eckman MH, Beshansky JR, Durand-Zaleski I, et al. Anti-coagulation for noncardiac procedures in patients withprosthetic heart valves. Does low risk mean high cost? JAMA1990; 263:1513–1521

83 Katholi RE, Nolan SP, McGuire LB. Living with prostheticheart valves: subsequent noncardiac operations and the riskof thromboembolism or hemorrhage. Am Heart J 1976;92:162–167

84 Douketis JD, Crowther MA, Cherian SS. Perioperativeanticoagulation in patients with chronic atrial fibrillationwho are undergoing elective surgery: results of a physiciansurvey. Can J Cardiol 2000; 16:326–330

85 Douketis JD, Crowther MA, Cherian SS, et al. Physicianpreferences for perioperative anticoagulation in patientswith a mechanical heart valve who are undergoing electivenoncardiac surgery. Chest 1999; 116:1240–1246

86 Centers for Disease Control and Prevention. CDC FactBook, 2001–2002. Atlanta, GA: Centers for Disease Controland Prevention, 2003

87 Kearon C, Ginsberg JS, Julian JA, et al. Comparison of




fixed-dose weight-adjusted unfractionated heparin and low-molecular-weight heparin for acute treatment of VTE.JAMA 2006; 296:935–942

88 Hylek EM, Go AS, Chang Y, et al. Effect of intensity of oralanticoagulation on stroke severity and mortality in atrialfibrillation. N Engl J Med 2003; 349:1019–1026

89 Hylek EM, Skates SJ, Sheehan MA, et al. An analysis of thelowest effective intensity of prophylactic anticoagulation forpatients with nonrheumatic atrial fibrillation. N Engl J Med1996; 335:540–546

90 O’Donnell M, Oczkowski W, Fang J, et al. Preadmissionantithrombotic treatment and stroke severity in patients withatrial fibrillation and acute ischaemic stroke: an observa-tional study. Lancet Neurol 2006; 5:749–754

91 Buller HR, Bousser MG, Bouthier J, et al. Comparison ofidraparinux with vitamin K antagonists for prevention ofthromboembolism in patients with atrial fibrillation: a ran-domised, open-label, non-inferiority trial. Lancet 2008; 371:278–280

92 Fanikos J, Tsilimingras K, Kucher N, et al. Comparison ofefficacy, safety, and cost of low-molecular-weight heparinwith continuous-infusion unfractionated heparin for initia-tion of anticoagulation after mechanical prosthetic valveimplantation. Am J Cardiol 2004; 93:247–250

93 Amorosi SL, Tsilimingras K, Thompson D, et al. Costanalysis of ‘bridging therapy’ with low-molecular-weightheparin versus unfractionated heparin during temporaryinterruption of chronic anticoagulation. Am J Cardiol 2004;93:509–511

94 Spyropoulos AC, Frost FJ, Hurley JS, et al. Costs andclinical outcomes associated with low-molecular-weight hep-arin vs unfractionated heparin for perioperative bridging inpatients receiving long-term oral anticoagulant therapy.Chest 2004; 125:1642–1650

95 Spyropoulos AC, Jenkins P, Bornikova L. A disease manage-ment protocol for outpatient perioperative bridge therapywith enoxaparin in patients requiring temporary interruptionof long-term oral anticoagulation. Pharmacotherapy 2004;24:649–658

96 Goldstein JL Larson LR, Yamashita BD, et al. Low molec-ular weight heparin versus unfractionated heparin in thecolonoscopy peri-procedure period: a cost modeling study.Am J Gastroenterol 2001; 96:2360–2366

97 Douketis JD, Woods K, Foster GA, et al. Bridging antico-agulation with low-molecular-weight heparin after interrup-tion of warfarin therapy is associated with a residual antico-agulant effect prior to surgery. Thromb Haemost 2005;94:528–531

98 O’Donnell MJ, Kearon C, Johnson J, et al. Preoperativeanticoagulant activity after bridging low-molecular-weightheparin for temporary interruption of warfarin. Ann InternMed 2007; 146:184–187

99 Turpie AG, Bauer KA, Eriksson BI, et al. Fondaparinux vsenoxaparin for the prevention of VTE in major orthopedicsurgery: a meta-analysis of 4 randomized double-blind stud-ies. Arch Intern Med 2002; 162:1833–1840

100 Strebel N, Prins M, Agnelli G, et al. Preoperative orpostoperative start of prophylaxis for VTE with low-molec-ular-weight heparin in elective hip surgery? Arch InternMed 2002; 162:1451–1456

101 Cruickshank MK, Levine MN, Hirsh J, et al. A standardheparin nomogram for the management of heparin therapy.Arch Intern Med 1991; 151:333–337

102 Anand SS, Yusuf S, Pogue J, et al. Relationship of activatedpartial thromboplastin time to coronary events and bleedingin patients with acute coronary syndromes who receiveheparin. Circulation 2003; 107:2884–2888

103 Dolovich LR, Ginsberg JS, Douketis JD, et al. A meta-analysis comparing low-molecular-weight heparins with un-fractionated heparin in the treatment of VTE: examiningsome unanswered questions regarding location of treatment,product type, and dosing frequency. Arch Intern Med 2000;160:181–188

104 Tricoci P, Roe MT, Mulgund J, et al. Clopidogrel to treatpatients with non-ST-segment elevation acute coronary syn-dromes after hospital discharge. Arch Intern Med 2006;166:806–811

105 Fox KA, Goodman SG, Anderson FA Jr, et al. Fromguidelines to clinical practice: the impact of hospital andgeographical characteristics on temporal trends in the man-agement of acute coronary syndromes; the Global Registryof Acute Coronary Events (GRACE). Eur Heart J 2003;24:1414–1424

106 Berger JS, Roncaglioni MC, Avanzini F, et al. Aspirin for theprimary prevention of cardiovascular events in women andmen: a sex-specific meta-analysis of randomized controlledtrials. JAMA 2006; 295:306–313

107 Iakovou I, Schmidt T, Bonizzoni E, et al. Incidence,predictors, and outcome of thrombosis after successfulimplantation of drug-eluting stents. JAMA 2005; 293:2126 –2130

108 Patrono C, Coller B, FitzGerald GA, et al. Platelet-activedrugs: the relationships among dose, effectiveness, and sideeffects; the Seventh ACCP Conference on Antithromboticand Thrombolytic Therapy. Chest 2004; 126(suppl):234S–264S

109 Roth GJ, Majerus PW. The mechanism of the effect ofaspirin on human platelets: I. Acetylation of a particulatefraction protein. J Clin Invest 1975; 56:624–632

110 Roth GJ, Stanford N, Majerus PW. Acetylation of prosta-glandin synthase by aspirin. Proc Natl Acad Sci U S A 1975;72:3073–3076

111 Taggart DP, Siddiqui A, Wheatley DJ. Low-dose preopera-tive aspirin therapy, postoperative blood loss, and transfu-sion requirements. Ann Thorac Surg 1990; 50:424–428

112 Quinn MJ, Fitzgerald DJ. Ticlopidine and clopidogrel.Circulation 1999; 100:1667–1672

113 Harder S, Klinkhardt U, Alvarez JM. Avoidance of bleedingduring surgery in patients receiving anticoagulant and/orantiplatelet therapy: pharmacokinetic and pharmacody-namic considerations. Clin Pharmacokinet 2004; 43:963–981

114 Caplain H, Cariou R. Long-term activity of clopidogrel: athree-month appraisal in healthy volunteers. Semin ThrombHaemost 1999; 25(suppl 2):21–24

115 Lenz TL, Hilleman DE. Aggrenox: a fixed-dose combinationof aspirin and dipyridamole. Ann Pharmacother 2000; 34:1283–1290

116 Douglas JS Jr, Holmes DR Jr, Kereiakes DJ, et al. Coronarystent restenosis in patients treated with cilostazol. Circula-tion 2005; 112:2826–2832

117 Thompson PD, Zimet R, Forbes WP, et al. Meta-analysis ofresults from eight randomized, placebo-controlled trials onthe effect of cilostazol on patients with intermittent claudi-cation. Am J Cardiol 2002; 90:1314–1319

118. Schror K. The pharmacology of cilostazol. Diabet ObesMetab 2002; 4(suppl 2):S14–S19

119 Munster T, Furst DE. Nonsteroidal anti-inflammatorydrugs, disease-modifying antirheumatic drugs, nonopioidanalgesics, and drugs used in gout. 8th ed. New York, NY:McGraw-Hill, 2001

120 Munster T, Furst DE. Pharmacotherapeutic strategies fordisease-modifying antirheumatic drug (DMARD) combina-tions to treat rheumatoid arthritis (RA). Clin Exp Rheumatol1999; 17(6 suppl 18):S29–S36




121 Quinn MJ, Fitzgerald DJ. Ticlopidine and clopidogrel.Circulation 1999; 100:1667–1672

122 Steinhubl SR, Lauer MS, Mukherjee DP, et al. The durationof pretreatment with ticlopidine prior to stenting is associ-ated with the risk of procedure-related non-Q-wave myocar-dial infarctions. J Am Coll Cardiol 1998; 32:1366–1370

123 Schuhlen H, Kastrati A, Dirschinger J, et al. Intracoronarystenting and risk for major adverse cardiac events during thefirst month. Circulation 1998; 98:104–111

124 Steinhubl SR, Berger PB, Brennan DM, et al. Optimaltiming for the initiation of pre-treatment with 300 mgclopidogrel before percutaneous coronary intervention.J Am Coll Cardiol 2006; 47:939–943

125 Gawaz M, Seyfarth M, Muller I, et al. Comparison of effectsof clopidogrel versus ticlopidine on platelet function inpatients undergoing coronary stent placement. Am J Cardiol2001; 87:332–336

126 Muller I, Seyfarth M, Rudiger S, et al. Effect of a highloading dose of clopidogrel on platelet function in patientsundergoing coronary stent placement. Heart 2001; 85:92–93

127 Geiger J, Teichmann L, Grossmann R, et al. Monitoring ofclopidogrel action: comparison of methods. Clin Chem2005; 51:957–965

128 Steinhubl SR, Talley JD, Braden GA, et al. Point-of-caremeasured platelet inhibition correlates with a reduced riskof an adverse cardiac event after percutaneous coronaryintervention: results of the GOLD (AU-Assessing Ultegra)multicenter study. Circulation 2001; 103:2572–2578

129 Kitchen L, Erichson RB, Sideropoulos H. Effect of drug-induced platelet dysfunction on surgical bleeding. Am J Surg1982; 143:215–217

130 Ferraris VA, Swanson E. Aspirin usage and perioperativeblood loss in patients undergoing unexpected operations.Surg Gynecol Obstet 1983; 156:439–442

131 Sharma AK, Ajani AE, Hamwi SM, et al. Major noncardiacsurgery following coronary stenting: when is it safe tooperate? Catheter Cardiovasc Interv 2004; 63:141–145

132 Reddy PR, Vaitkus PT. Risks of noncardiac surgery aftercoronary stenting. Am J Cardiol 2005; 95:755–757

133 Ernst A, Eberhardt R, Wahidi M, et al. Effect of routineclopidogrel use on bleeding complications after transbron-chial biopsy in humans. Chest 2006; 129:734–737

134 Thom T, Haase N, Rosamond W, et al. Heart disease andstroke statistics–2006 update: a report from the AmericanHeart Association Statistics Committee and Stroke StatisticsSubcommittee. Circulation 2006; 113:e85–e151

135 Yusuf S, Zhao F, Mehta SR, et al. Effects of clopidogrel inaddition to aspirin in patients with acute coronary syn-dromes without ST-segment elevation. N Engl J Med 2001;345:494–502

136 Eikelboom JW, Mehta SR, Anand SS, et al. Adverse impactof bleeding on prognosis in patients with acute coronarysyndromes. Circulation 2006; 114:774–782

137 Koch CG, Li L, Duncan AI, et al. Morbidity and mortalityrisk associated with red blood cell and blood-componenttransfusion in isolated coronary artery bypass grafting. CritCare Med 2006; 34:1608–1616

138 Sethi GK, Copeland JG, Goldman S, et al. Implications ofpreoperative administration of aspirin in patients undergo-ing coronary artery bypass grafting: Department of VeteransAffairs Cooperative Study on Antiplatelet Therapy. J AmColl Cardiol 1990; 15:15–20

139 Merritt JC, Bhatt DL. The efficacy and safety of perioper-ative antiplatelet therapy. J Thromb Thrombolysis 2002;13:97–103

140 Bybee KA, Powell BD, Valeti U, et al. Preoperative aspirintherapy is associated with improved postoperative outcomes

in patients undergoing coronary artery bypass grafting.Circulation 2005; 112:1286–1292

141 Topol E. Aspirin with bypass surgery: from taboo to newstandard of care. N Engl J Med 2002; 347:1359–1360

142 Alexander JH, Berger PB, Hafley G, et al. Impact of earlyclopidogrel use on angiographic and clinical outcomes fol-lowing coronary artery bypass surgery: findings from PRE-VENT IV. Am Coll Cardiol 2006; 47(suppl A):932–257

143 Mehta RH, Roe MT, Mulgund J, et al. Acute clopidogrel useand outcomes in patients with non-ST-segment elevationacute coronary syndromes undergoing coronary artery by-pass surgery. J Am Coll Cardiol 2006; 48:281–286

144 Yusuf S, Zhao F, Mehta SR, et al. Effects of clopidogrel inaddition to aspirin in patients with acute coronary syn-dromes without ST-segment elevation. N Engl J Med 2001;345:494–502

145 Eikelboom JW, Mehta SR, Anand SS, et al. Adverse impactof bleeding on prognosis in patients with acute coronarysyndromes. Circulation 2006; 114:774–782

146 Purkayastha S, Athanasiou T, Malinovski V, et al. Doesclopidogrel affect outcome after coronary artery bypassgrafting? A meta-analysis. Heart 2006; 92:531–532

147 Hongo RH, Ley J, Dick SE, et al. The effect of clopidogrelin combination with aspirin when given before coronaryartery bypass grafting. J Am Coll Cardiol 2002; 40:231–237

148 Chen L, Bracey AW, Radovancevic R, et al. Clopidogrel andbleeding in patients undergoing elective coronary arterybypass surgery by intravenous aprotinin during cardiotho-racic surgery. Circulation 2004; 110:2004

149 Kapetanakis EI, Medlam DA, Petro KR, et al. Effect ofclopidogrel premedication in off-pump cardiac surgery: arewe forfeiting the benefits of reduced hemorrhagic sequelae?Circulation 2006; 113:1667–1674

150 Akowuah E, Shrivastava V, Jamnadas B, et al. Comparison oftwo strategies for the management of antiplatelet therapyduring urgent surgery. Ann Thorac Surg 2005; 1:149–152

151 van der Linden J, Lindvall G, Sartipy U. Aprotinin decreasespostoperative bleeding and number of transfusions in pa-tients on clopidogrel undergoing coronary artery bypassgraft surgery: a double-blind, placebo-controlled, random-ized clinical trial. Circulation 2005; 112(suppl I):276–280

152 Mangano DT, Tudor IC. The risk associated with aprotininin cardiac surgery. N Engl J Med 2006; 254:353–365

153 Spiess BD, Royston D, Levy JH, et al. Platelet transfusionsduring coronary artery bypass graft surgery are associated withserious adverse outcomes. Transfusion 2004; 44:1143–1148

154 Carless PA, Moxey AJ, Stokes BJ, et al. Are antifibrinolyticdrugs equivalent in reducing blood loss and transfusion incardiac surgery? A meta-analysis of randomized head-to-head trials. BMC Cardiovasc Disord 2005; 5:19

155 Brown JR, Birkmeyer NJ, O’Connor GT. Meta-analysiscomparing the effectiveness and adverse outcomes of anti-fibrinolytic agents in cardiac surgery. Circulation 2007;115:2801–2813

156 Pleym H, Stenseth R, Wahba A, et al. Prophylactic treat-ment with desmopressin does not reduce postoperativebleeding after coronary surgery in patients treated withaspirin before surgery. Anesth Analg 2004; 98:578–584

157 Mehta SR, Yusuf S, Peters RJ, et al. Effects of pretreatmentwith clopidogrel and aspirin followed by long-term therapyin patients undergoing percutaneous coronary intervention:the PCI-CURE study. Lancet 2001; 358:527–533

158 Sabatine MS, Cannon CP, Gibson CM, et al. Effect ofclopidogrel pretreatment before percutaneous coronary in-tervention in patients with ST-elevation myocardial infarc-tion: the PCI-CLARITY Study. JAMA 2005; 294:1224–1232

159 Kastrati A, von Beckerath N, Joost A, et al. Loading with 600




mg clopidogrel in patients with coronary artery disease withand without chronic clopidogrel therapy. Circulation 2004;110:1916–1919

160 Patti G, Colonna G, Pasceri V, et al. A randomized trial ofhigh loading dose of clopidogrel for reduction of peri-procedural myocardial infarction in patients undergoingcoronary intervention: results from the ARMYDA-2 (Anti-platelet therapy for Reduction of MYocardial Damage dur-ing Angioplasty) Study. Circulation 2005; 111:2099–2106

161 Kastrati A, Mehilli J, Schulhen H, et al. A clinical trial ofabciximab in elective percutaneous coronary interventionafter pretreatment with clopidogrel. N Engl J Med 2004;350:232–238

162 Hausleiter J, Kastrati A, Mehilli J, et al. A randomized trialcomparing phosphorylcholine-coated stenting with balloonangioplasty as well as abciximab with placebo for restenosisreduction in small coronary arteries. J Intern Med 2004;256:388–397

163 Kaluza GL, Joseph J, Lee JR, et al. Catastrophic outcomes ofnoncardiac surgery soon after coronary stenting. J Am CollCardiol 2000; 35:1288–1294

164 Wilson SH, Fasseas P, Orford JL, et al. Clinical outcome ofpatients undergoing non-cardiac surgery in the two monthsfollowing coronary stenting. J Am Coll Cardiol 2003; 42:234–240

165 Vicenzi MN, Meislitzer T, Heitzinger B, et al. Coronaryartery stenting and non-cardiac surgery: a prospective out-come study. Br J Anaesth 2006; 96:686–693

166 McFadden EP, Stabile E, Regar E, et al. Late thrombosis indrug-eluting coronary stents after discontinuation of anti-platelet therapy. Lancet 2004; 364:1519–1521

167 Nasser M, Kapeliovich M, Markiewicz W. Late thrombosisof sirolimus-eluting stents following noncardiac surgery.Catheter Cardiovasc Interv 2005; 65:516–519

168 Compton PA, Zankar AA, Adesanya AO, et al. Risk ofnoncardiac surgery after coronary drug-eluting stent implan-tation. Am J Cardiol 2006; 98:1212–1213

169 Bakhru M, Saber W, Brotman D, et al. Is discontinuation ofantiplatelet therapy after 6 months safe in patients withdrug-eluting stents undergoing noncardiac surgery? CleveClin J Med 2006; 73:S23

170 Fleron MH, Dupuis M, Mottet P, et al. Non cardiac surgeryin a patient with coronary stenting: think sirolimus now! AnnFr Anesth Reanim 2003; 22:733–735

171 Collet JP, Montalescot G, Blanchet B, et al. Impact of prioruse of recent withdrawal of oral antiplatelet agents on acutecoronary syndromes. Circulation 2004; 110:2361–2367

172 Ferrari E, Benhamou M, Cerboni P, et al. Coronary syn-dromes following aspiring withdrawal: a special risk for latestent thrombosis. J Am Coll Cardiol 2005; 45:456–459

173 Wilson S, Rihal CS, Bell MR, et al. Timing of coronary stentthrombosis in patients treated with ticlopidine and aspirin.Am J Cardiol 1999; 83:1006–1011

174 Cutlip DE, Baim DS, Ho KK, et al. Stent thrombosis in themodern era: a pooled analysis of multicenter coronary stentclinical trials. Circulation 2001; 103:1967–1971

175 Schouten O, Bax JJ, Poldermans D. Management of patientswith cardiac stents undergoing noncardiac surgery. CurrOpin Anaesthesiol 2007; 20:274–278

176 Bergerson P, Rudondy P, Poyen V, et al. Long-term periph-eral stent evaluation using angioscopy. Int Angiol 1991;10:182–186

177 Ueda Y, Nanto S, Komamura K, et al. Neointimal coverageof stents in human coronary arteries observed by angioscopy.J Am Coll Cardiol 1994; 23:341–346

178 Cattaneo M. Platelet P2 receptors: old and new targets for

antithrombotic drugs. Expert Rev Cardiovasc Ther 2007;5:45–55

179 Grines CL, Bonow RO, Casey DE Jr, et al. Prevention ofpremature discontinuation of dual antiplatelet therapy in pa-tients with coronary artery stents: a science advisory from theAmerican Heart Association, American College of Cardiology,Society for Cardiovascular Angiography and Interventions,American College of Surgeons, and American Dental Associa-tion, with representation from the American College of Physi-cians. J Am Coll Cardiol 2007; 49:734–739

180 Brilakis ES, Banerjee S, Berger PB. Perioperative manage-ment of patients with coronary stents. J Am Coll Cardiol2007; 49:2145–2150

181 Schulman S, Kearon C. Definition of major bleeding in clinicalinvestigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost 2005; 3:692–694

182 Borea G, Montebugnoli L, Capuzzi P, et al. Tranexamic acid asa mouthwash in anticoagulant-treated patients undergoing oralsurgery: an alternative method to discontinuing anticoagulanttherapy. Oral Surg Oral Med Oral Pathol 1993; 75:29–31

183 Evans IL, Sayers MS, Gibbons AJ, et al. Can warfarin becontinued during dental extraction? Results of a randomizedcontrolled trial. Br J Oral Maxillofac Surg 2002; 40:248–252

184 Sacco R, Sacco M, Carpenedo M, et al. Oral surgery inpatients on oral anticoagulant therapy: a randomized com-parison of different INR targets. J Thromb Haemost 2006;4:688–689

185 Al-Belasy FA, Amer MZ. Hemostatic effect of n-butyl-2-cyanoacrylate (histoacryl) glue in warfarin-treated patientsundergoing oral surgery. J Oral Maxillofac Surg 2003;61:1405–1409

186 Carter G, Goss A. Tranexamic acid mouthwash–a prospec-tive randomized study of a 2-day regimen vs 5-day regimento prevent postoperative bleeding in anticoagulated patientsrequiring dental extractions. Int J Oral Maxillofac Surg 2003;32:504–507

187 Carter G, Goss A, Lloyd J, et al. Tranexamic acid mouthwashversus autologous fibrin glue in patients taking warfarinundergoing dental extractions: a randomized prospectiveclinical study. J Oral Maxillofac Surg 2003; 61:1432–1435

188 Halfpenny W, Fraser JS, Adlam DM. Comparison of 2hemostatic agents for the prevention of postextraction hem-orrhage in patients on anticoagulants. Oral Surg Oral MedOral Pathol Oral Radiol Endod 2001; 92:257–259

189 Ramstrom G, Sindet-Pedersen S, Hall G, et al. Prevention ofpostsurgical bleeding in oral surgery using tranexamic acidwithout dose modification of oral anticoagulants. J OralMaxillofac Surg 1993; 51:1211–1216

190 Campbell JH, Alvarado F, Murray RA. Anticoagulation andminor oral surgery: should the anticoagulation regimen bealtered? J Oral Maxillofac Surg 2000; 58:131–135

191 Cannon PD, Dharmar VT. Minor oral surgical procedures inpatients on oral anticoagulants: a controlled study. AustDent J 2003; 48:115–118

192 Devani P, Lavery KM, Howell CJ. Dental extractions inpatients on warfarin: is alteration of anticoagulant regimenecessary? Br J Oral Maxillofac Surg 1998; 36:107–111

193 Gaspar R, Brenner B, Ardekian L, et al. Use of tranexamicacid mouthwash to prevent postoperative bleeding in oralsurgery patients on oral anticoagulant medication. Quintes-sence Int 1997; 28:375–379

194 Saour JN, Ali HA, Mammo LA, et al. Dental procedures inpatients receiving oral anticoagulation therapy. J HeartValve Dis 1994; 3:315–317

195 Street AM, Leung W. Use of tranexamic acid mouthwash indental procedures in patients taking oral anticoagulants.Med J Aust 1990; 153:630




196 Zanon E, Martinelli F, Bacci C, et al. Safety of dentalextraction among consecutive patients on oral anticoagulanttreatment managed using a specific dental managementprotocol. Blood Coagul Fibrinolysis 2003; 14:27–30

197 Blinder D, Manor Y, Martinowitz U, et al. Dental extrac-tions in patients maintained on oral anticoagulant therapy:comparison of INR value with occurrence of postoperativebleeding. Int J Oral Maxillofac Surg 2001; 30:518–521

198 Della Valle A, Sammartino G, Marenzi G, et al. Preventionof postoperative bleeding in anticoagulated patients under-going oral surgery: use of platelet-rich plasma gel. J OralMaxillofac Surg 2003; 61:1275–1278

199 Martinowitz U, Mazar AL, Taicher S, et al. Dental extractionfor patients on oral anticoagulant therapy. Oral Surg OralMed Oral Pathol 1990; 70:274–277

200 Ramli R, Abdul Rahman R. Minor oral surgery in anticoag-ulated patients: local measures alone are sufficient forhaemostasis. Singapore Dent J 2005; 27:13–16

201 Russo G, Corso LD, Biasiolo A, et al. Simple and safemethod to prepare patients with prosthetic heart valves forsurgical dental procedures. Clin Appl Thromb Haemost2000; 6:90–93

202 Zusman SP, Lustig JP, Baston I. Postextraction hemostasisin patients on anticoagulant therapy: the use of a fibrinsealant. Quintessence Int 1992; 23:713–716

203 Barrero MV, Knezevic M, Martın MT. Oral surgery in thepatients undergoing oral anticoagulant therapy. MedicinaOral 2002; 7:63–70

204 Cieslik-Bielecka A, Pelc R, Cieslik T. Oral surgery proce-dures in patients on anticoagulants: preliminary report.Kardiologia Polska 2005; 63:137–140

205 Keiani Motlagh K, Loeb I, Legrand W, et al. Prevention ofpostoperative bleeding in patients taking oral anticoagulants:effects of tranexamic acid. Rev Stomatol Chir Maxillofac2003; 104:77–79

206 Garcia-Darennes F, Darennes J, Freidel M, et al. Protocol foradapting treatment with vitamin K antagonists before dentalextraction. Rev Stomatol Chir Maxillofac 2003; 104:69–72

207 Blinder D, Manor Y, Martinowitz U, et al. Dental extrac-tions in patients maintained on continued oral anticoag-ulant: comparison of local hemostatic modalities. OralSurg Oral Med Oral Pathol Oral Radiol Endod 1999;88:137–140

208 Bodner L, Weinstein JM, Baumgarten AK. Efficacy of fibrinsealant in patients on various levels of oral anticoagulantundergoing oral surgery. Oral Surg Oral Med Oral PatholOral Radiol Endod 1998; 86:421–424

209 Ardekian L, Gaspar R, Peled M, et al. Does low-dose aspirintherapy complicate oral surgical procedures? J Am DentAssoc 2000; 131:331–335

210 Madan GA, Madan SG, Madan G, et al. Minor oral surgerywithout stopping daily low-dose aspirin therapy: a study of51 patients. J Oral Maxillofac Surg 2005; 63:1262–1265

211 Allard RH, Baart JA, Huijgens PC, et al. Antithrombotictherapy and dental surgery with bleeding. Ned TijdschrTandheelkd 2004; 111:482–485

212 Alcalay J. Cutaneous surgery in patients receiving warfarintherapy. Dermatol Surg 2001; 27:756–758

213 Billingsley EM, Maloney ME. Intraoperative and postoper-ative bleeding problems in patients taking warfarin, aspirin,and nonsteroidal antiinflammatory agents: a prospectivestudy. Dermatol Surg 1997; 23:381–383

214 Kargi E, Babuccu O, Hosnuter M, et al. Complications ofminor cutaneous surgery in patients under anticoagulanttreatment. Aesthetic Plast Surg 2002; 26:483–485

215 Syed S, Adams BB, Liao W, et al. A prospective assessmentof bleeding and international normalized ratio in warfarin-

anticoagulated patients having cutaneous surgery. J AmAcad Dermatol 2004; 51:955–957

216 Bartlett GR. Does aspirin affect the outcome of minorcutaneous surgery? Br J Plast Surg 1999; 52:214–216

217 Shalom A, Wong L. Outcome of aspirin use during excisionof cutaneous lesions. Ann Plast Surg 2003; 50:296–298

218 Alam M, Goldberg LH. Serious adverse vascular eventsassociated with perioperative interruption of antiplatelet andanticoagulant therapy. Dermatol Surg 2002; 28:992–998

219 Hirschman DR, Morby LJ. A study of the safety of contin-ued anticoagulation for cataract surgery patients. Nurs Fo-rum 2006; 41:30–37

220 Kallio H, Paloheimo M, Maunuksela EL. Haemorrhage and riskfactors associated with retrobulbar/peribulbar block: a prospectivestudy in 1383 patients. Br J Anaesth 2000; 85:708–711

221 Katz J, Feldman MA, Bass EB, et al. Risks and benefits ofanticoagulant and antiplatelet medication use before cata-ract surgery. Ophthalmology 2003; 110:1784–1788

222 Lumme P, Laatikainen LT. Risk factors for intraoperativeand early postoperative complications in extracapsular cata-ract surgery. Eur J Ophthalmol 1994; 4:151–158

223 Narendran N, Williamson TH. The effects of aspirin andwarfarin therapy on haemorrhage in vitreoretinal surgery.Acta Ophthalmol Scand 2003; 81:38–40

224 Wirbelauer C, Weller A, Haberle H, et al. Cataract surgeryunder topical anesthesia with oral anticoagulants. KlinMonatsbl Augenheilkd 2004; 221:749–752

225 Roberts CW, Woods SM, Turner LS. Cataract surgery inanticoagulated patients. J Cataract Refract Surg 1991; 17:309–312

226 Rotenstreich Y, Rubowitz A, Segev F, et al. Effect ofwarfarin therapy on bleeding during cataract surgery.J Cataract Refract Surg 2001; 27:1344–1346

227 Kumar N, Jivan S, Thomas P, et al. Sub-Tenon’s anesthesiawith aspirin, warfarin, and clopidogrel. J Cataract RefractSurg 2006; 32:1022–1025

228 Davies B. Combined aspirin and clopidogrel in cataractsurgical patients: a new risk factor for ocular haemorrhage?Br J Ophthalmol 2004; 88:1226–1227

229 Baker RI, Coughlin PB, Gallus AS, et al. Warfarin reversal:consensus guidelines, on behalf of the Australasian Society ofThrombosis and Haemostasis. Med J Aust 2004; 181:492–497

230 Dezee KJ, Shimeall WT, Douglas KM, et al. Treatment ofexcessive anticoagulation with phytodione (vitamin K): ameta-analysis. Arch Intern Med 2006; 166:391–397

231 Lubetsky A, Yonath H, Olchovsky D, et al. Comparison oforal vs intravenous phytodione (vitamin K1) in patients withexcessive anticoagulation: a prospective randomized con-trolled study. Arch Intern Med 2003; 163:2469–2473

232 Raj G, Kumar R, McKinney WP. Time course of reversal ofanticoagulant effect of warfarin by intravenous and subcutane-ous phytodione. Arch Intern Med 2000; 159:2721–2724

233 Hongo RH, Ley J, Dick SE, et al. The effect of clopidogrelin combination with aspirin when given before coronaryartery bypass grafting. J Am Coll Cardiol 2002; 40:231–237

234 Vilahur G, Choi BG, Zafar MU, et al. Normalization ofplatelet reactivity in clopidogrel-treated subjects. J ThrombHaemost 2007; 5:82–90

235 Levi MM, Vink R, de Jonge E. Management of bleedingdisorders by prohemostatic therapy. Int J Hematol 2002;76(suppl 2):139–144

236 Porte RJ, Leebeek FW. Pharmacological strategies to de-crease transfusion requirements in patients undergoing sur-gery. Drugs 2002; 62:2193–2211

237 Nutescu E, Helgason C. Outpatient dalteparin peri-proce-dure bridge therapy in patients maintained on long termwarfarin. Stroke 2001; 32:328–329




DOI 10.1378/chest.08-0675 2008;133; 299S-339SChest

C. Spyropoulos, Richard C. Becker and Jack AnsellJames D. Douketis, Peter B. Berger, Andrew S. Dunn, Amir K. Jaffer, Alex

Guidelines (8th Edition)College of Chest Physicians Evidence-Based Clinical Practice

: American*The Perioperative Management of Antithrombotic Therapy

December 25, 2011This information is current as of

http://chestjournal.chestpubs.org/content/133/6_suppl/299S.full.htmlUpdated Information and services can be found at:

Updated Information & Services

http://chestjournal.chestpubs.org/content/133/6_suppl/299S.full.html#ref-list-1This article cites 233 articles, 79 of which can be accessed free at:

References

http://chestjournal.chestpubs.org/content/133/6_suppl/299S.full.html#related-urlsThis article has been cited by 51 HighWire-hosted articles:

Cited Bys

http://www.chestpubs.org/site/misc/reprints.xhtmlfound online at: Information about reproducing this article in parts (figures, tables) or in its entirety can bePermissions & Licensing

http://www.chestpubs.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

Reprints

"Services" link to the right of the online article.Receive free e-mail alerts when new articles cite this article. To sign up, select the

Citation Alerts

PowerPoint slide format. See any online figure for directions. articles can be downloaded for teaching purposes inCHESTFigures that appear in Images in PowerPoint format



http://chestjournal.chestpubs.org/content/133/6_suppl/299S.full.html#ref-list-1

http://chestjournal.chestpubs.org/content/133/6_suppl/299S.full.html#related-urls

http://www.chestpubs.org/site/misc/reprints.xhtml

http://www.chestpubs.org/site/misc/reprints.xhtml


Documents

The Perioperative Management of Antithrombotic Therapy ... · drugs, we suggest against the routine use of platelet function assays to monitor the antithrom-botic effect of aspirin