The best anticoagulation strategy during Extracorporeal Membrane Oxygenation (ECMO) is unknown. Actual recommendations suggest to use unfractionated heparin infusion and monitor the effect with either activated Partial Thromboplastin Time (aPTT) or Activated Clotting Time (ACT). Since hemorrhage is still the major adverse effect of ECMO with impact on mortality and morbidity, the investigators raised a question whether an alternative monitoring technique namely Thromboelastography (TEG) could allow a more accurate management of anticoagulation in this setting. To test this hypothesis the investigators designed a pilot study to test safety and feasibility of an anticoagulation monitoring algorithm based on TEG versus aPTT.

Extracorporeal membrane oxygenation (ECMO) is frequently used as a supportive tool in the settings of acute severe respiratory and/or circulatory failure. The presence of the extracorporeal circuit and oxygenator bring along the need to anticoagulate the patients in order to prevent activation of the coagulation cascade that can lead to thrombosis of the artificial surfaces and losses of coagulation factors, fibrinogen and platelets. Despite tight monitoring with different anticoagulation essays bleeding remain the main threat for the patients on ECMO, leaving thrombosis as an anecdotic event.

In our institution, anticoagulation during ECMO is monitored mainly with aPTT ratio in a range of 1.5 to 2 (patient to normal). Recently, the investigators started to match Thromboelastography (TEG) assays to activated Partial Thromboplastin Time (aPTT) and Activated Clotting Time (ACT) routine measurements during ECMO. Data from a preliminary analysis of thirty-two patients show that in nearly half of TEG assays, the R (reaction time) parameter which is sensitive to inhibition of fibrin formation by heparin, was consistently prolonged despite aPTT and ACT were in the normal range. In the same patients the investigators registered high rate of bleeding (nearly 40% of patients suffered from clinically significant bleeding, i.e. bleeding that required heparin dose reduction, blood transfusions or interruption of the extracorporeal support). These findings suggests the possibility that the actual target of aPTT ratio might be overrated.

The purpose of this study is to test safety and feasibility of standardized protocol based on TEG versus current practice (aPTT) to manage anticoagulation during ECMO.

40 consecutive adults patients undergoing ECMO will be randomized in either the TEG-driven anticoagulation group or the aPTT-driven group. Written informed consent will be obtained according to national legislation.

In both groups, when the patient is connected to ECMO, an heparin bolus of 50-70 UI/kg (depending on baseline aPTT value) will be administered followed by a continuous infusion of 18 UI/kg/h started; for the first 12 hours thereafter proper anticoagulation level will be monitored with Activated Clotting Time (ACT, therapeutic range 180-210 seconds, performed every 1 or 2 hours according to the standardized protocol). From the thirteenth hour onwards each group will follow either aPTT monitoring or TEG according to randomization. Every morning in both groups aPTT, TEG, Antithrombin and anti-Factor Xa activity will be assessed.

• aPTT-driven group: aPTT ratio target is 1.5 - 2.0 as for actual clinical practice; frequency of aPTT measurements may vary from a minimum of 3 times per day to a maximum of 6 depending on the standardized protocol. When aPTT value falls well below the desired range an heparin bolus will be administered. When too high, infusion will be stopped for either 30 or 60 minutes.
• TEG-driven group: target range of R parameter at Kaolin activated-TEG (R K-TEG) is 16 - 24 minutes; frequency of TEG assays may vary from a minimum of 3 times per day to a maximum of 6 depending on the standardized protocol. When TEG value falls well below the desired range an heparin bolus will be administered, when is too high infusion will be stopped for either 30 or 60 minutes.

Daily management requires haemoglobin, platelets, PT ratio, D-dimer and fibrinogen to be checked 2 to 3 times daily, and maintained at specific values (Hb > 10 g/dl; PLT > 45.000/mm3; Fibrinogen > 150 mg/dl; Antithrombin activity > 70%; PT ratio > 1.5) with haemocomponents transfusion.

Data from each patient will be collected on a daily basis: laboratory exams, hemodynamic status, transfusions, surgery or invasive interventions, extracorporeal lung performance.

End of the study will be at ECMO discontinuation, after that deep vein thrombosis on patients' cannulation sites will be assessed with doppler ultrasound.

Inclusion Criteria:

• All patients that initiate ECMO support for acute respiratory failure at our centre or by our equipe at another centre will be screened for eligibility.

Exclusion Criteria:

• Patients under the age of 18, with proven Heparin-induced Thrombocytopenia or with a platelets count of less than 30.000/mm3 will be excluded from the study.