html xmlns=”http://www.w3.org/1999/xhtml” xml:lang=”en”>

Antithrombotic Therapy in the ICU: Practical Management Considerations

William E. Dager, PharmD, BCPS (AQ Cardiology), FCSHP, FCCP, FCCM, FASHP

The occurrence of thromboembolic events in critically ill patients is common and carries the potential for long-term consequences, including increased mortality. In the critically ill patient, the presence of multiple risk factors leading to venous stasis, endothelial trauma, and hypercoagulable conditions dramatically increases the risk for a thromboembolic event. Incidence rates for venous thromboembolic events (VTEs) in the ICU range from 20% to 80%.¹ Prevention of thromboembolism requires careful assessment of the risk of bleeding and thrombosis and a decision regarding approaches for thromboprophylaxis. In the presence of acute, recent, or high-risk of thromboembolism, management considerations will depend on the clinical presentation of the patient. In the ICU, therapeutic decisions may need to be altered as the patient’s clinical status changes. Observations from clinical trials can be helpful; however, clinicians may need to individualize management, because critically ill patients are frequently excluded from enrollment in clinical trials. Selected topics regarding the prevention and treatment of thromboembolism in the ICU are discussed in this chapter, including applications into practice.

Thromboprophylaxis in the ICU

Decisions to initiate pharmacological thromboprophylaxis in critically ill patients care will depend on multiple factors, including the risk of thrombosis balanced with the risk of bleeding. In the surgical setting, it is recommended that some form of prophylaxis be initiated within 24 hours postoperatively.² In selected situations, bleeding risks may limit the use of pharmacological prophylaxis. The presence of an epidural catheter, the setting of neurosurgical procedures, and concurrent major bleeding risks or active bleeding may limit options for pharmacological prophylaxis for a period of time until the risks diminish. Several tools have been proposed for risk assessment but require further validation in critically ill patients (Table 10-1).³

Pharmacological prophylaxis is a common approach for VTE prevention, either alone or in addition to mechanical devices or placement of inferior vena cava filters. Removing any risk factors for VTE and encouraging mobility for the patient should be considered. The selection of a pharmacological prophylaxis regimen should be based on the individual presentation of the patient. Considerations include what sites are available for injections and the potential that other conditions such as obesity or renal failure, may require revision of the dosing regimen.

Randomized trials comparing the various options for prophylaxis in critical care are limited. In general, some form of pharmacological prophylaxis is better than placebo.⁴ Indications for the use of low-molecular-weight heparin (LMWH) are found in selected surgical settings such as major trauma, ischemic stroke, and elective orthopedic procedures. It is not clear whether there is a difference in outcomes between the utility of LMWH and that of low-dose subcutaneous heparin in the general surgical or medical populations. Low-dose heparin, however, may be preferred where bleeding risks are high, such as in the presence of neuraxial anesthesia, or when institutional decisions are made for cost-related reasons when there is a lack of compelling evidence to use more expensive options. Low-dose heparin, typically given 2 or 3 times daily, is a common option but may be associated with missed doses, for example, when patients are off the floor for procedures. Low-molecular-weight heparins are frequently a favored option because they may be given less often and have been associated with a lower risk of developing heparin-induced thrombocytopenia (HIT).⁵

Choosing a dose of LMWH may depend on the patient’s risk for thrombosis or bleeding. This class of agents was originally marketed as having a wide therapeutic range without the need for laboratory monitoring. Treatment regimens should consider what agents are available, the frequency with which the patient can tolerate injections, and patient weight and renal function. In the presence of HIT or antithrombin (AT) deficiency, unfractionated heparin (UFH) and LMWHs may not be effective, and other options such as a direct thrombin inhibitor (DTI) may be considered.⁶ Alterations in the pharmacokinetic response following subcutaneous injections may occur in the critically ill. Lower anti–factor Xa activity levels as the result of expanded volumes of distribution and poor absorption are seen with obesity, fluid overload, and concomitant use of vasopressor agents.^7-9 Injections of the LMWH enoxaparin into the thigh instead of the abdomen may reduce its bioavailability.¹⁰ Wide variability in anti–factor Xa activity has been observed with little change from baseline activity in the critically ill.^11,12 In the PROTECT trial comparing the impact of dalteparin 5,000 U once daily to low-dose UFH 5,000 U twice daily, the incidences of proximal deep vein thrombosis, pulmonary embolism (PE), and major bleeding were similar (5.8% vs 5.1%, 2.3% vs 1.3%, and 5.6% vs 5.5%, respectively). Of note is that trauma, orthopedic surgery, and neurosurgical patients were excluded from the trial.¹³

One option considered in some centers is the use of a continuous infusion of UFH for prophylaxis in selected ICU situations such as major trauma or morbid obesity. Targeting a small increase in the activated partial thromboplastin time (aPTT) (eg, 10-15 seconds over baseline) would suggest that some anticoagulation effect is being delivered. Outcomes for thrombosis, bleeding, or mortality relative to using UFH infusions in this manner for prophylaxis have not been studied. Another option being explored is dose-adjusted LMWH based on anti–factor Xa activity levels. The challenge with either approach is that notable variations and lack of standardization exists between laboratory methods along the requirement for additional resources without a clear understanding on how outcomes are improved with such monitoring. Further trials are needed to determine the most optimal approach to pharmacological VTE prophylaxis in the critically ill. Drug choice, dose, route, frequency, and duration of therapy may depend on the individual presentation of the patient.

Therapeutic Anticoagulation in the ICU

Once a thromboembolic event has occurred and a decision has been made to initiate anticoagulation, clinicians will need to explore available options for anticoagulation based on the individual presentation of the patient. In the presence of high bleeding risk or pending procedures, UFH infusions may be the choice for initial therapy. Typically, a bolus dose is administered to rapidly achieve an adequate level of anticoagulation followed by a weight-based continuous infusion.¹⁴ If anticoagulation has already been established or the risk for acute thrombosis is low, a bolus dose may not be necessary. In the setting of an acute stroke, the use of bolus dosing is discouraged because of the potential for increased bleeding.¹⁵ Within a few hours of initiating the infusion, the clinician should measure either the aPTT or anti–factor Xa activity level to assess whether the infusion rate is achieving target goals. Laboratory draws shortly after the bolus (eg, 4 hours) may result in higher values because effects of the bolus are still present, leading to little or no change in the infusion rate and a delay in reaching targets. Delays in achieving targets with continuous infusion of heparin in the setting of VTE have been associated with diminished outcomes.¹⁶ Thus, when aPTT/anti–factor Xa activity levels should be considered at 4 hours if no bolus is administered and up to 8 hours out if a bolus is administered. This allows results to reflect the response to the infusion and more accurately, guide dosing adjustments so that targets are reached earlier. Heparin also follows a circadian elimination pattern where aPTT or anti–factor Xa values may be higher during sleep cycles compared with awake periods.¹⁷ Delays in transferring withdrawn blood into a citrated collection tube can also increase reported aPTT values.

Heparin activity is commonly measured using either the aPTT or anti–factor Xa activity. Benefits of either monitoring approach for bleeding or thrombosis are unknown. In the setting of AT deficiency, elevated factor VIII, or fibrinogen, the aPTT may not respond, suggesting potential heparin resistance (Table 10-2). For AT deficiency, the aPTT has a potential advantage over anti–factor Xa assays that incorporate AT into the assay and may leave true resistance to heparin unrecognized. If an aPTT or anti–factor Xa assay is questioned (eg, heparin infusion rate >25 U/kg/h and no response from baseline), consider checking the opposite assay, activated clotting time if available, or thrombin times to determine whether there is any signal suggesting poor performance of the assay in that particular patient.

Oral Anticoagulants

Once initial anticoagulation has been established and concerns for bleeding diminish, longer acting anticoagulants including vitamin K antagonists (VKAs) such as warfarin can be initiated. Typically, 5 days of parenteral anticoagulant therapy is desired prior to discontinuing.¹⁸ A baseline international normalized ratio (INR) should be drawn to assess changes with subsequent draws. Presence of drug interactions, decreased cardiac, liver, or renal function which are common in the critical care setting may warrant lower initial doses, even if the regimen used prior to admission was higher. Typically, steady-state INR values of 2 to 3 are targeted for VKAs but may be adjusted upward or downward in selected situations. The INR draw should be delayed until 10 to 12 hours post dose to allow for a response. Initial increases (or decreases from holding the VKA) are weighted more on the decrease (or increase) in Factor VII, which has the shortest elimination half-life.¹⁹ Changes in Factor II will lag behind. Thus, a rapidly increasing INR value will reflect a lower intensity of anticoagulation compared to same number that is declining. Notable changes in reported INR values outside expectations can be misleading. One example involves line draws that become hemodiluted; the INR and aPTT increase and hematocrit declines, falsely suggesting a bleed. In such situations, the test should be repeated by peripheral phlebotomy, if possible, for confirmation. The presence of UFH or DTI can independently increase the INR.^20,21 For UFH, some laboratories may perform an additional step to neutralize the heparin prior to measuring the INR, potentially yielding lower values (especially with higher concurrent aPTT values) than samples that are not neutralized.

Several strategies regarding the overlap period with a parenteral anticoagulant and VKA initiation have been proposed. One such strategy is to wait for the INR to reach a value of 2 prior to stopping the parenteral anticoagulant; another suggestion is to extend the parenteral anticoagulant for an additional day or two once the INR is in target range. Most clinical trials have used either approach in the warfarin cohorts; however, no trial has directly compared the two strategies and determined any advantages with the additional overlap. If the patient is not able to take warfarin orally, it can be given intravenously. Since almost the entire oral warfarin dose is absorbed, the same dose may be considered regardless of the route. However, this has not been studied in the critically ill, and the possibility of reduced oral bioavailability creating difference in response between routs may exist.

Aside from using the intravenous DTIs in the management of heparin-induced thrombocytopenia discussed below, 2 new oral anticoagulants have been marketed in the United States recently. Dabigatran is an oral DTI approved for use in nonvalvular atrial fibrillation to prevent strokes.²² Rivaroxaban is an oral factor Xa inhibitor approved for stroke prevention in the setting of non-valvular atrial fibrillation, VTE prophylaxis following elective orthopedic joint replacement and for treatment of VTE. Additional studies have explored these agents in the setting of acute coronary syndromes and treatment of deep vein thrombosis or PE.²² In general, the use of these agents in critically ill patients has not been investigated.

Challenges with the use of the new oral anticoagulants include lack of a clear reversal strategy and limited ability to measure the intensity of the anticoagulation present.²² Dabigatran is renally eliminated, requiring dosing adjustments as renal function declines. The bioavailability of dabigatran, which is in a liquid capsule, is around 6%; however, opening the capsule prior to administration increases the bioavailability to more than 80%, and safety measures should be put into place to prevent this. The effects of dabigatran can be measured using an undiluted thrombin time, which is very sensitive to the presence of dabigatran and may serve as a signal to continued drug presence.²³ This may be a helpful tool when invasive procedures are planned. In general, the aPTT is more sensitive to dabigatran than the prothrombin time or INR.²⁴ The INR can also be elevated, and values drawn using point-of-care devices can be notably higher than those drawn in the central laboratory.²⁵ Other approaches such as using dilute thrombin times and ecarin clotting times to assess the concentration of dabigatran are under development.

Rivaroxaban is partially eliminated by the kidney, and renal dosing adjustments are recommended.²⁶ The ecarin clotting time and thrombin time cannot be used to measure rivaroxaban effects. The prothrombin time is more sensitive than the aPTT, and INR measurements may be inconsistent because of differences in calibration methods.²⁷ The measurement of anti–factor Xa activity is being explored; however, rivaroxaban-specific calibrators should be used, and considerable variability between laboratories can exist. With both agents, clinicians should interpret results with caution until they understand the limitations of the test. In addition, it is not known how values predict risk of thrombosis or bleeding, especially when an invasive procedure is planned. Careful attention to renal function and administration time of the last dose should be included in management decisions.

Special Management Populations

Dosing Weights

Patient body weight can be difficult to accurately measure in the ICU. Clinicians should give careful attention to the dosing weight when initiating weight-based dosing regimens. Presence of morbid obesity (e.g. total body weights over 110 – 150 kg) alone is frequently not a population included or independently studied in clinical trials involving anticoagulants. The use of the patient’s total body weight excluding excess fluid from anasarca should be considered to guide weight based dosing of selected anticoagulants. Obesity is also an independent risk factor for PE, and dose adjustments for low-dose heparin or LMWH that take into consideration the excess weight are unclear. Unless notable bleeding risks are present, occurrence of a major PE and the potential for hemodynamic collapse may be of greater concern than acute unexpected bleeding suggesting caution with regimens that adjust or cap UFH or LMWH dose in the obese. Dose capping or weight related adjustments may be considered in extreme cases of obesity.

Organ Failure

Renal insufficiency is common in the critically ill, and a third of patients are reported to have creatinine clearance (CrCl) values less than 30 mL/min, a common cut-off value for adjusting the dose of anticoagulants.²⁸ Limited evidence suggests that there is no need to adjust the dose of dalteparin if the estimated steady-state CrCl is greater than 20 mL/min and that the dose of enoxaparin should be reduced by half if the CrCl is between 20 and 30 mL/min.²⁹ Information on the use of LMWH for the systemic treatment of thromboembolism and concurrent severe renal impairment (CrCl <20 mL/min or requiring hemodialysis) is limited. Other anticoagulants, such as DTIs, may require renal dosing adjustments (Table 10-3).

When hemodialysis is used, several approaches are available, including intermittent, extended, and continuous methods. If acute kidney injury is present, renal replacement therapy when used may change on a daily basis. Anticoagulation regimens may need to address both prevention of thrombosis in the dialysis circuit and meet necessary, systemic anticoagulation goals.

In the presence of heart failure, diminished cardiac output may reduce hepatic and renal elimination of drugs, resulting in diminished drug clearance. As cardiac function recovers, dose–response relationships with anticoagulants may change.

Liver impairment can reduce hepatic elimination of selected anticoagulants, including warfarin and argatroban. Reductions in clotting factor production can then result in a hypocoagulable state and increased risk for bleeding, creating challenges in decisions to anticoagulate the patient. Patients may have nutritional deficiencies and low albumin levels, which influence the response to warfarin.

Related posts:

Media, Technology, and Education in the ICU Water, and Brain–What Should the Intensivil Know? Paradigms in Sepsis Initiatives in Emergency Preparedness of the Injured Patient: The Enemy of Good Is Better on Sepsis Resuscitation

Stay updated, free articles. Join our Telegram channel

Join