Anticoagulation and coagulation management during ECMO is a challenge. Bleeding and clotting are major sources of morbidity and mortality. The currently used strategies are of poor evidence. This observational study wants to evaluate the use of automated thromboelastography (TEG 6s) to guide the management of coagulation and anticoagulation in patients supported by ECMO

ECMO induces both a risk of clotting and bleeding. Therapeutic anticoagulation is recommended during ECMO. While preventing clotting, anticoagulation triggers a higher risk of bleeding. However, anticoagulation does not completely avoid clotting due to complex interactions between the circuit, the patients' blood and the anticoagulant. Unfractionated heparin (UFH) is the most commonly used anticoagulant and requires monitoring. No clear recommendation exists. The activated partial thromboplastin time (aPTT) as well as the anti-factor Xa activity (anti-Xa) are plasma-based tests used to monitor UFH. The evidence for their use is poor and they do no correlate well. Moreover, these two tests do not take into account all the elements of blood and do not reflect the entire coagulation process.

Thromboelastography is a whole-blood point-of-care test that describes each phases of the clotting process from the activation of coagulation factors to the later lysis to the thrombus. TEG is also sensitive to UFH and can be used to quantify its effect. The use of heparinase during TEG also permits to evaluate the coagulation process during the use of heparin. This could give an important understanding of the effect of the ECMO itself on the circuit and help to develop a strategy to prevent bleeding and clotting as well as monitor heparin treatment.

The hypothesis for this study is that the use of thromboelastography will identify situations with high risk of bleeding and allow interventions to reduce hemorrhagic events and blood products transfusions. It is also hypothesized that the monitoring of unfractionated heparin (UFH) with TEG is feasible and could lead to the use of less UFH during the course of ECMO.

The objectives of this study are 1) to determine and calibrate the TEG R-time values corresponding to aPTT therapeutic range for patients under therapeutic UFH during ECMO course, 2) to determine the level of correlation of TEG parameters with other anticoagulation tests [Prothrombin Time (PT), anti-Xa, activated clotting time (ACT)], fibrinogen level, platelets count and d-dimers, 3) to determine the TEG values that are associated with the administration of blood-derived coagulation products in clinical practice 4) to identify TEG cutoff points that are associated with an increased risk of bleeding/clotting in ECMO patients 5) use these observational data to develop a TEG-based algorithm for anticoagulation management and blood products replacement that could be tested in a future study.

This study will you the last version of TEG, TEG-6s which is a fully automated point-of-care device.

Inclusion Criteria:

• Consecutive patients admitted in Toronto General Hospital Medical Surgical ICU (MSICU) supported with ECMO. Patients cannulated before admission (ECMO retrieval) can also be included

Exclusion Criteria:

• Refusal of consent, unavailability or impossibility to process blood samples before ECMO cannulation

Adult patients suffering from cardiorespiratory failure and needing Extracorporeal Membrane Oxygenation