The Pre-brief
Extracorporeal Membrane Oxygenation (ECMO) features human blood sloshing against an artificial circuit composed of cannulas, tubing, a circulatory pump and an oxygenator. Blood does not like interacting with these foreign membranes and will clot if given a chance. As such, anticoagulation is (usually) required for the safe and effective use of ECMO. Unfortunately, shear stress from the circuit and the underlying patient illness contribute to a dueling coagulopathy. Monitoring of anticoagulation needs to balance the dangers of clotting and bleeding.
The Extracorporeal Life Support Organization (ELSO) released a guideline in 2014 to help navigate the challenge of anticoagulation for ECMO. That document is undoubtedly worth a read, but does not give a recommendation on which laboratory test or assay should be used to monitor degree of anticoagulation. For this post’s purpose, I am limiting the discussion to tests of unfractionated heparin as it is the most common anticoagulant used for ECMO. Activated clotting time (ACT), activated partial thromboplastin time (aPTT), anti-Xa (Xa), and thromboelastography(TEG/ROTEM) all are potential means of monitoring heparin anticoagulation.
ACT: ACT is a point of care test performed by adding a surface activator (kaolin, etc) to whole blood. The time for initial fibrin formation is measured in seconds. ACT has ubiquitous use during cardiopulmonary bypass, given its low cost, accessibility, and rapidity of results. Normal targets for bypass are 400-800 seconds, while ECMO targets range from 160-200. Unfortunately, many factors common to the ECMO patient can prolong the ACT independent of the heparin effect. This includes hemodilution, thrombocytopenia, hypothermia, hypofibrinogenemia, and coagulation factor deficiencies. More concerning, it seems that ACT levels correlate poorly with the heparin effect at dose ranges used in ECMO.
aPTT: aPTT is derived by exposing citrated plasma to calcium and a surface activator (kaolin, ellagic acid). Targets in ECMO are 60-80 seconds. Like ACT, aPTT is susceptible to error with hypofibrinogenemia, factor deficiency, or acute phase reactants (particularly prevalent while on ECMO). Due to a wide variety of aPTT laboratory methods, there is great variability in the relationship between given aPTT assay and anti-Xa levels. One study found a range of 48s to 104s across different aPTT assays for a serum anti-Xa level of 0.3. A strategy to combat this is to establish anti-Xa levels and correlate your institution’s aPTT assay to that level.
Anti-Xa: Anti-Xa concentration directly measures the heparin inhibition of factor Xa. Target values for ECMO range from 0.3-0.7 IU/mL. Anti-Xa levels feature a stable reference range and are not affected by factor deficiencies or acute phase reactants. Anti-Xa correlates more closely to heparin dose than aPTT. The Anti-Xa level can be falsely low in the presence of hyperbilirubinemia or high plasma free hemoglobin. Another shortcoming is that anti-Xa levels represent a ratio of inhibition between heparin effect and the amount of thrombin and fibrin being produced. In a particularly prothrombotic human, the reference range may not be sufficient. Further, this assay is more costly than the ACT or aPTT.
TEG/ROTEM: These viscoelastic tests of hemostasis give a measure of time to initial fibrin formation, clot firmness, platelet function, and fibrinolysis. These tests can be used with and without heparinase to assess the underlying hemostasis in the presence of heparin. The difference in R or clotting time between tests with and without heparinase demonstrates the extent of the heparin effect. A pilot RCT compared TEG to aPTT, finding that the TEG-based heparin titration led to a lower amount of heparin being used, with no statistical difference in bleeding or thrombotic events. A larger trial to identify particular TEG/ROTEM reference ranges and powered to detect clinically essential endpoints is needed.

Which test to use? It depends (this is the default answer in critical care). Cost and institutional availability may limit your choices.
My practice is to target Anti-Xa levels depending upon the patient’s bleeding risk and assessment of clots on the ECMO circuit. For example, though my original target is 0.3-0.7 IU/mL, if the patient is at risk of bleeding complications, I will lower the target. Further, if the level is within that range without bleeding and clot is aggressively accumulating on the oxygenator, I will increase the target.
I use TEG (the viscoelastic test at my institution) in times of excessive bleeding (to assess heparin effect and factor/platelet function, evidence of hyperfibrinolysis) or when Anti-Xa levels are not increasing as expected to appropriate heparin dosages. I would likely use TEG more, though the test is only available at my shop when perfusionists are in-house. As such, institutional availability will probably drive many anticoagulation protocols.
Lastly, heparin anticoagulation is not mandatory in all ECMO runs. In instances of severe bleeding, heparinization is withheld. This lack of anticoagulation is tolerated at relatively high ecmo flow rates (anecdotally above four liters) given the heparin-coated surface of most contemporary ECMO circuits.
The Debrief
- There are many options for monitoring heparin anticoagulation while on ECMO
- There is no perfect test. ACT is insensitive, aPTT is affected by acute phase reactants and factor deficiency, Anti-Xa is expensive, and TEG/ROTEM has limited evidence.
- The best test is the one you can use at your shop. My approach is to target Anti-Xa levels and evaluate the patient clinically for bleeding or ECMO membrane clotting. TEG is a useful adjunct.
Title image by Dr. Rahel Gizaw (@PhysicianDoodles).
References
- Chlebowski, M. M., Baltagi, S., Carlson, M., Levy, J. H., & Spinella, P. C. (2020). Clinical controversies in anticoagulation monitoring and antithrombin supplementation for ECMO. Critical Care, 24(1), 1-12.
- Prime, B. ECLS Circuit. “ELSO Anticoagulation Guideline.” (2014).
- Hohlfelder, B., Kelly, D., Hoang, M., Anger, K. E., Sylvester, K. W., Kaufman, R. M., & Connors, J. M. (2019). Activated Clotting Times Demonstrate Weak Correlation With Heparin Dosing in Adult Extracorporeal Membrane Oxygenation. American journal of therapeutics.
- Bates, S. M., Weitz, J. I., Johnston, M., Hirsh, J., & Ginsberg, J. S. (2001). Use of a fixed activated partial thromboplastin time ratio to establish a therapeutic range for unfractionated heparin. Archives of internal medicine, 161(3), 385-391.
- Garcia, D. A., Baglin, T. P., Weitz, J. I., & Samama, M. M. (2012). Parenteral anticoagulants: antithrombotic therapy and prevention of thrombosis: American College of Chest Physicians evidence-based clinical practice guidelines. Chest, 141(2), e24S-e43S.
- Arnouk, S., Altshuler, D., Lewis, T. C., Merchan, C., Smith, D. E., Toy, B., … & Papadopoulos, J. (2020). Evaluation of anti-Xa and activated partial thromboplastin time monitoring of heparin in adult patients receiving extracorporeal membrane oxygenation support. ASAIO Journal, 66(3), 300-306.
- Kostousov, V., Nguyen, K., Hundalani, S. G., & Teruya, J. (2014). The influence of free hemoglobin and bilirubin on heparin monitoring by activated partial thromboplastin time and anti-Xa assay. Archives of Pathology and Laboratory Medicine, 138(11), 1503-1506.
- Fina, D., Matteucci, M., Jiritano, F., Meani, P., Kowalewski, M., Ballotta, A., … & Lorusso, R. (2020). Extracorporeal membrane oxygenation without systemic anticoagulation: a case-series in challenging conditions. Journal of Thoracic Disease, 12(5), 2113.