Procedures in the Setting of Anticoagulation

Introduction

Performing an invasive procedure on an anticoagulated emergency department (ED) patient can be challenging for the emergency provider (EP). Regardless of the procedure, some anticoagulated patients are at potential significant risk of hemorrhage from the procedure. However, emergency reversal of anticoagulation in order to perform the procedure may also place the patient at risk for serious thrombotic complications. When deciding to perform a procedure on an anticoagulated ED patient, the EP must weigh the risks of bleeding with the risk of disrupting anticoagulation, emergency reversal of anticoagulation, and delaying a potentially critical intervention. If the procedure is not needed for life-saving therapy, postponing the procedure or providing empiric treatment may be a reasonable choice. In contrast, emergency procedures to reverse an imminent life-threatening condition should never be withheld and emergency reversal of anticoagulation may be required. This chapter reviews current literature and recommendations for select ED procedures such as lumbar puncture, central venous catheterization, arthrocentesis, paracentesis, thoracentesis, and tube thoracostomy in the anticoagulated patient.

Assessing Risk of Bleeding and Thrombosis

All anticoagulants inhibit stable clot formation and increase the risk of bleeding. Routine coagulation testing, that is prothrombin time (PT), partial thromboplastin time (PTT), international normalized ratio (INR), and platelet count, in anticoagulated patients, is recommended prior to performing an invasive procedure. Although laboratory tests to determine drug presence, drug concentration, and level of anticoagulant effect can be useful in the assessment of bleeding risk, standard coagulation assays accurately monitor the degree of anticoagulation for only a few agents. PT and INR conveniently assay the extrinsic pathway to monitor the quantitative effect of warfarin therapy. PTT or activated partial thromboplastin time (aPTT) accurately monitors the anticoagulant effect of heparin. Newer nonvitamin K antagonist oral anticoagulants (NOACs) were developed to have predictable pharmacokinetics with minimal interactions to avoid the need for routine laboratory testing. Coagulation assays to quantify the effect or presence of a NOAC would be useful in many clinical circumstances such as in active bleeding, suspected overdose, and the perioperative period, but do not exist for many. The direct measurement of drug concentration is not suitable in clinical practice because of the time required to perform the laboratory analysis. Standard coagulation assays can be useful in extreme scenarios, but are generally insufficient in determining drug concentration or degree of anticoagulant effect of NOACs ( Table 72.1 ).

Normal thrombin time is suitable for excluding significant dabigatran levels but too sensitive for determining the degree of anticoagulant effect. Dilute thrombin time, ecarin chromogenic assay, and ecarin clotting time may be useful to determine the degree of anticoagulant effect, but is not standardized across laboratories. Although using standard coagulation assays to quantify dabigatran effect is not recommended, an elevated PT/INR suggests a supratherapeutic dabigatran effect, and a normal PTT excludes it. Unfortunately, none of these tests are accurate predictors of bleeding risk with dabigatran and patients may have bleeding when these tests are within the normal range.

Anti-factor Xa activity is valuable to determine the effect of factor Xa inhibitors such as rivaroxaban and apixaban, but this test is not widely available. Although it is not recommended to use standard coagulation assays to quantify drug effect, a normal PT and INR level virtually excludes supratherapeutic rivaroxaban effect.

For patients on warfarin, understanding how INR correlates to the risk of bleeding is important. Spontaneous bleeding events are uncommon when the INR is normal or within a therapeutic range (i.e., 2 to 3). The relative risk for bleeding with an INR between 3 and 5 is 2.7 (95% confidence interval [CI] 1.8–3.9). However, when the INR is above 5, the relative risk of a spontaneous hemorrhagic event increases dramatically to 21.8 (95% CI 12.1–39.4). These findings are similar to a previous retrospective review in 1995 that also showed adverse events occurred in 75 per 100 patient-years for an INR greater than or equal to 6.5. Patients with severe coagulopathy (INR greater than 9) have a poor prognosis, with 67% experiencing spontaneous hemorrhagic events and a 74% mortality rate.

Patients without a history of bleeding or anticoagulant use do not require routine coagulation studies prior to performing a procedure unless history or physical examination suggest a bleeding disorder or use of anticoagulation. A 2005 systematic review failed to demonstrate the utility of routine coagulation studies on nonanticoagulated patients prior to the performance of a procedure. Abnormal laboratory findings such as thrombocytopenia, an elevated INR, or other coagulation abnormalities are not necessarily absolute contraindications to performing an invasive procedure. Numerous studies have demonstrated that select procedures can safely be performed even in the setting of anticoagulation with a vitamin K antagonist (VKA). Determining a patient's risk for bleeding requires examination of both the procedure to be performed, the anticoagulant medication, and the level of anticoagulation. Unfortunately, there are no validated systems available to quantify risk of bleeding.

The risk of thrombosis after anticoagulant reversal is significant for a subset of ED patients. These patients include those with a recent diagnosis of pulmonary embolism, significant clot burden, or those with mechanical hardware such as a prosthetic cardiac valve. When clinically feasible, it is best to allow the anticoagulant effect of a medication to wane rather than emergently reversing the medication in these patients. This approach is rarely possible in the ED, so timing of the procedure should be coordinated with the inpatient provider if the patient requires admission.

Many procedures report contraindications with severe coagulopathy (INR greater than 9) and disseminated intravascular coagulation (DIC). Proper coagulation is significantly disrupted in these clinical settings and invasive procedures may lead to severe bleeding. However, life-saving measures such as thoracostomy, central venous catheterization, and endotracheal intubation are sometimes required even in these severe settings. Additional life-saving procedures that may carry a high risk of bleeding in the setting of anticoagulation, but should not be withheld, include defibrillation and pericardiocentesis. The gravity of the clinical scenario is important to keep in mind when weighing the risk of performing a procedure in the setting of anticoagulation.

Limited data are available on the safety of performing procedures on patients taking newer NOACs. These NOACs may confer the same, less, or increased risk of bleeding as traditional VKA agents. The lack of an effective reversal agent for many of the NOACs is another factor that complicates the decision to perform a procedure in the setting of anticoagulation. Fortunately, many of the NOACs have a duration of action that is less than the traditional VKAs.

Management of Anticoagulant Associated Bleeding

Although periprocedural bleeding is typically not life threatening and can often be controlled with direct pressure, predicting which anticoagulated patients will bleed from a procedure is difficult. Furthermore, the simple act of controlling periprocedural bleeding may not entirely end the risk of serious harm. For instance, periprocedural bleeding with percutaneous coronary intervention is associated with an increased short- and long-term morbidity and mortality, including major adverse cardiovascular events and readmission rates well after the bleeding is controlled.

When managing spontaneous, periprocedural, or postprocedural bleeding associated with anticoagulation, there are a few principles to consider. Most importantly, never withhold emergency life-saving procedures such as endotracheal intubation, tube thoracostomy, cardiac defibrillation, pericardiocentesis, or vascular access when necessary.

Stabilization with supportive treatments such as oxygenation, intravascular volume resuscitation, and repletion of blood products via transfusion are the initial steps of assessment and management of bleeding in the anticoagulated patient. Patients with prolonged or severe bleeding may have presenting symptoms in various stages of circulatory shock. Poor blood supply and intravascular volume leads to poor tissue perfusion. Cellular hypoxia, damage, and resulting inflammatory response from hypoperfusion can exacerbate the patient's clinical status. Adequate resuscitation and stabilization is key to the initial phase of management.

Hemostatic measures such as direct compression of a bleeding or oozing vessel, wound, or region should be performed concurrently with stabilization to prevent further blood loss. Additional measures may be needed to control bleeding when compression fails or when the bleeding originates from a noncompressible site. These measures include the use of topical hemostatic agents, systemic hemostatic agents (i.e., tranexamic acid), procedural intervention, operative management, or intraarterial embolization.

The clinician can assess the severity of coagulopathy with laboratory analysis. Patients may experience bleeding at subtherapeutic, therapeutic, or supratherapeutic levels of anticoagulation. Determining qualitative level of anticoagulation may help guide therapy. However, many anticoagulants do not have an associated diagnostic test that is accurate, timely, and clinically relevant to determine severity of coagulopathy like warfarin or heparin. Obtain routine coagulation assays (i.e., PT, PTT, and INR) when no relevant study exists. Routine coagulation assays may sometimes be helpful in determining the lack of a supratherapeutic anticoagulant effect but are often difficult to interpret.

When bleeding is severe, life-threatening, refractory to hemostatic efforts, or the coagulopathy is determined to be severe, restoring the ability to generate an effective clot by administration of an antidote, coagulation factor, blood product, or removal of the offending anticoagulant (i.e., hemodialysis) may be necessary. Consideration for the anticoagulant's duration of action is also important when deciding whether to reverse anticoagulation.

Withholding further doses of anticoagulation may be necessary after stabilization and disposition. When the clinical decision to withhold anticoagulant is not clear, consultation with the prescribing physician (e.g., cardiologist prescribing aspirin and clopidogrel for cardiac stent) or hematologist may be warranted.

Reversal of Anticoagulation

The decision to discontinue or reverse anticoagulation prior to a procedure is difficult. Although there are no validated data that can accurately stratify risk for peri- or postoperative thromboembolism, the American College of Chest Physicians (ACCP) has published guidelines that stratify patients into low-, moderate-, and high-risk groups. These guidelines are listed in Table 72.2 .

TABLE 72.2

Suggested Patient Risk Stratification for Perioperative Arterial or Venous Thromboembolism

From Douketis JD, Berger PB, Dunn AS, et al: The perioperative management of antithrombotic therapy: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition), Chest 133:299S, 2008.

Indication for VKA Therapy
RISK STRATUM	MECHANICAL HEART VALVE	ATRIAL FIBRILLATION	VTE
High	Any mitral valve prosthesis	CHADS2 score of 5 or 6	Recent (within 3 mo) VTE
	Older (caged-ball or tilting disc) aortic valve prosthesis	Recent (within 3 mo) stroke or transient ischemic attack,	Severe thrombophilia (e.g., deficiency of protein C, protein S or antithrombin, antiphospholipid antibodies, or multiple abnormalities)
	Recent (within 6 mo) stroke or transient ischemic attack	Rheumatic valvular heart disease
Moderate	Bileaflet aortic valve prosthesis and one of the following: atrial fibrillation, prior stroke or transient ischemic attack, hypertension, diabetes, congestive heart failure, age > 75 yr	CHADS2 score of 3 or 4	VTE within the past 3 to 12 mo; Nonsevere thrombophilic conditions (e.g., heterozygous factor V Leiden mutation, heterozygous factor II mutation); Recurrent VTE; Active cancer (treated within 6 mo or palliative)
Low	Bileaflet aortic valve prosthesis without atrial fibrillation and no other risk factors for stroke	CHADS2 score of 0 to 2 (and no prior stroke or transient ischemic attack)	Single VTE occurred 12 mo ago and no other risk factors

CHADS2, Congestive heart failure-hypertension-age-diabetes-stroke; VTE, venous thromboembolism.

Risk factors that increase the likelihood of a thromboembolic event with anticoagulant disruption include a prosthetic heart valve, atrial fibrillation, recent cerebrovascular accident (CVA) or transient ischemic attack, recent venous thromboembolism (VTE), the presence of a VTE risk factor (i.e., protein C/S deficiency, antiphospholipid antibody), and a high CHAD ₂ score (congestive heart failure, hypertension, age, diabetes mellitus, CVA). Patients with mechanical valves are at higher risk for thromboembolic events when anticoagulation is disrupted than patients with bioprosthetic valves. Mitral valves present a higher risk of thromboembolic events than aortic prosthetic valves. The postoperative period just after (less than 3 months) mechanical valve placement is also a high-risk time for thromboembolic events.

Procedures on patients with prosthetic valves typically require a delayed approach to operative management, except when there is an emergency. In general, discontinuing anticoagulation in a patient with a prosthetic valve is safer than reversal of anticoagulation to perform a procedure. Reversal of anticoagulation puts patients at higher risk for thromboembolic events than does simply discontinuing the anticoagulant. When possible, discontinue anticoagulation and delay invasive procedures on patients with prosthetic valves until coagulation status returns to near normal. For patients with low risk bioprosthetic valves that require discontinuation of anticoagulation, warfarin should be discontinued 48 to 72 hours prior to the procedure or until the INR drops below 1.5. Anticoagulation may then be restarted 24 hours following the procedure. High-risk patients with prosthetic valves (e.g., mitral valve) require bridging therapy to reduce the time off anticoagulation.

For patients taking VKAs that require reversal of anticoagulation for an urgent procedure, administration of oral vitamin K is recommended. If more immediate reversal is required for a procedure, administration of fresh-frozen plasma or prothrombin complex concentrate in addition to oral vitamin K is recommended. For patients taking antiplatelet agents such as aspirin or clopidogrel, platelet transfusion is recommended for reversal of the effect ( Fig. 72.1 ).