Recent pressures on heparin supply have prompted many ECMO programs to take a closer look at their anticoagulation protocols, but the conversation around alternatives is one worth having regardless of the current landscape.  

For experienced ECMO clinicians, the “right” anticoagulation approach is far from settled science. The decision requires balancing circuit integrity, bleeding risk, patient physiology, and drug pharmacokinetics in a population where the margin for error is narrow. 

This post breaks down the primary anticoagulation options available for ECMO patients, the clinical reasoning behind each, and the trade-offs that make this decision so difficult. 

The See-Saw of Bleeding and Clotting 

For most clinicians reading this, the see-saw of clotting and bleeding risks on ECMO is well understood, but the tension between them is exactly what makes anticoagulation management so challenging with these patients. For this reason, it’s worth reviewing before getting into the drugs. 

When blood contacts the foreign surfaces of the ECMO circuit (tubing, oxygenator, cannula), it triggers the coagulation cascade and activates platelets, creating a prothrombotic environment. In the absence of systemic anticoagulation, this significantly elevates the risk of circuit clotting, oxygenator failure, and thromboembolic complications. 

It is worth noting that ECMO circuits have evolved over time, with more biocompatible surface coatings, improved oxygenator design, and more stable flow characteristics. These changes have reduced, though not eliminated, the thrombogenicity of the circuit compared to earlier generations. As a result, some programs have become more comfortable targeting lower levels of anticoagulation than would have been considered safe historically, as we’ll discuss later.  

Although ECMO patients are a higher risk of clotting, they are simultaneously at high risk for bleeding. Cannulation sites, recent surgical procedures, underlying coagulopathy, and the consumption of clotting factors and platelets by the circuit itself all contribute to a heightened risk of bleeding complications. Anticoagulation management on ECMO is therefore a constant balancing act: enough anticoagulation to protect the circuit and patient, not so much that you precipitate a catastrophic bleed. 

Circuit Configuration in Anticoagulation Strategy 

ECMO configuration is an important factor shaping anticoagulation strategy, and the differences between VV and VA have real implications for how intensively most programs approach it. In general: 

VV-ECMO patients are more likely to be managed at lower anticoagulation targets (or sometimes none at all) because: 

• Clots tend to embolize to the pulmonary circulation, not the systemic arterial system 

• Pulmonary emboli, while not trivial, are generally more manageable than arterial emboli; the ECMO circuit continues to support gas exchange even if pulmonary vascular resistance worsens 

• The native heart continues to eject, maintaining pulsatility and reducing arterial stasis 

VA-ECMO patients generally require more consistent, tightly managed anticoagulation because: 

• Blood is returned directly into the arterial system, meaning clots can go straight to the brain, coronaries, or limbs 

• Unlike a pulmonary embolus, arterial embolism is immediate, often catastrophic, and frequently irreversible; a stroke or MI from a circuit clot leaves no room for recovery 

• Reduced or absent native cardiac output creates stasis in the LV and aortic root, adding an additional source of thrombus with direct access to the systemic circulation 

The Anticoagulants: Options, Mechanisms, and Trade-offs 

Heparin 

Unfractionated heparin remains the most widely used anticoagulant in ECMO, and for most clinicians, will be the most familiar.  

Heparin works by binding antithrombin (AT), dramatically accelerating AT’s inhibition of thrombin and Factor Xa. This mechanism is both its strength and its limitation: patients with reduced AT levels (common in critically ill patients, particularly after prolonged illness or cardiac surgery) will have a blunted heparin response that isn’t captured by standard aPTT or ACT monitoring unless AT levels are specifically checked and repleted. 

The other major concern is heparin-induced thrombocytopenia (HIT). In ECMO patients, distinguishing HIT from other causes of thrombocytopenia is notoriously difficult, and the consequences of missing it are severe (circuit thrombosis, limb ischemia, stroke, etc). Any unexplained drop in platelet count should prompt a HIT workup and serious consideration of transitioning to another anticoagulant. 

Pros: 

• Established safety profile, widely understood 

• Inexpensive 

• Reversible with protamine 

• Easy to titrate with familiar monitoring (aPTT, ACT, anti-Xa) 

Cons: 

• Efficacy dependent on antithrombin levels and blunted response in AT-depleted patients is easy to miss 

• Variable and sometimes unpredictable dose-response 

• Risk of HIT 

• Heparin resistance not uncommon in critically ill patients 

Bivalirudin 

When heparin is contraindicated, bivalirudin is one of the primary alternatives. It is a direct thrombin inhibitor (DTI), meaning it binds directly to thrombin itself, independent of antithrombin. This makes its anticoagulant effect far more predictable across patients, eliminates the confounding variable of AT levels, and removes any risk of HIT. 

Some studies have also suggested lower rates of circuit thrombosis with bivalirudin compared to heparin, though findings are mixed and largely based on observational data. 

Unlike heparin, bivalirudin has no reversal agent. However, its short half-life (~25 minutes vs. 60–90 minutes for heparin) means its anticoagulant effect dissipates quickly once the infusion is stopped. 

It is primarily metabolized in the bloodstream rather than cleared renally, which is particularly relevant in ECMO patients where renal failure is common and renally cleared drugs can accumulate unpredictably. One caveat is that in low-flow states, such as severe cardiac dysfunction, drug breakdown may be impaired, raising the potential for accumulation. 

Pros: 

• Direct thrombin inhibition independent of antithrombin 

• No risk of HIT 

• More predictable pharmacokinetics than heparin 

• Short half-life (~25 min) provides some functional offset in bleeding situations 

• May reduce circuit thrombosis compared to heparin 

Cons: 

• Significantly more expensive than heparin 

• No reversal agent 

• Potential for accumulation in very low-flow states 

• Monitoring can be more complex at higher doses 

• Less long-term outcome data in ECMO compared to heparin 

Argatroban 

Argatroban is another DTI used when heparin is contraindicated. Some programs reach for it before bivalirudin, others after; institutional preference and familiarity often guide selection. 

Like bivalirudin, argatroban binds directly to the thrombin active site without requiring antithrombin. Its half-life is approximately 45 minutes—longer than bivalirudin but still relatively short compared to many other anticoagulants. Argatroban also does not have a reversal agent. Its anticoagulant effect will decline after stopping the infusion, but not as rapidly as with bivalirudin given its longer half-life. 

It is hepatically metabolized, which can be a significant limitation. In patients with hepatic dysfunction, argatroban can accumulate in the body, leading to supratherapeutic anticoagulation that is difficult to manage. 

A notable characteristic as well is that argatroban artificially elevates the INR, which can be problematic when the patient is on or being bridged to warfarin. 

Pros: 

• Direct thrombin inhibition independent of antithrombin 

• No risk of HIT 

• Moderate half-life (~45 min) 

• Predictable dosing in patients with normal hepatic function 

Cons: 

• Significantly more expensive than heparin 

• No reversal agent 

• Hepatic metabolism makes it unreliable and potentially dangerous in liver failure or significant hepatic congestion 

• Artificially elevates INR, which can complicate coagulation monitoring 

The No Anticoagulation Approach 

It may surprise some clinicians to hear that a growing number of centers are running ECMO patients with little to no systemic anticoagulation.  

This reflects a broader shift in how some programs conceptualize thrombosis risk on ECMO. Rather than anticoagulating everyone and only holding it for things like active bleeding, some programs have flipped that model and only anticoagulate when there’s a clear reason to, such as conditions like atrial fibrillation, evidence of circuit thrombosis, or low ECMO flows that increase thrombotic risk. 

Observational data from these centers suggest lower rates of bleeding complications and reduced transfusion requirements, without a clear increase in thrombotic events. However, these findings are derived from retrospective, single-center experiences and should be interpreted with appropriate caution. 

This practice is most commonly described in VV-ECMO, however, it is increasingly being applied in VA-ECMO at select centers as well. In these programs, anticoagulation is often deferred as long as circuit flows remain sufficiently high and stable, with the understanding that flow itself helps mitigate thrombotic risk. Anticoagulation may be initiated or resumed if flows decrease, if there are signs of circuit thrombosis, or if another clinical indication emerges. 

Importantly, “no anticoagulation” does not mean “no monitoring.” Programs using this strategy maintain rigorous circuit surveillance: visual inspection for clot, close attention to transmembrane pressure trends, and careful management of circuit flows. The approach shifts from pharmacologic anticoagulation to vigilant mechanical and physiologic management, but the level of attention required is no less. 

No Right Answer, Yet 

It would be convenient if there were consensus on how to best anticoagulate ECMO patients (if at all). But despite a growing body of research, the field is not there yet. Which drug to use, what targets to aim for, how aggressively to anticoagulate based on configuration, and whether to anticoagulate at all — these are all actively debated questions with variation across centers.  

What drives institutional protocols is a combination of available data, clinical experience, and local preference. Knowing the mechanisms, trade-offs, and failure modes of each option is what allows clinicians to navigate that uncertainty.