Secondary Logo

Journal Logo

Management of COVID-19 Patients

Specific Considerations for Venovenous Extracorporeal Membrane Oxygenation During Coronavirus Disease 2019 Pandemic

Guihaire, Julien*,†,‡; Owyang, Clark g.§; Madhok, Jai; Laverdure, Florent; Gaillard, Maïra*; Girault, Antoine#; Lebreton, Guillaume**,††; Mercier, Olaf†,‡,#

Author Information
doi: 10.1097/MAT.0000000000001251
  • Free


The coronavirus disease 2019 (COVID-19) global pandemic has created an unprecedented burden of acute respiratory distress syndrome (ARDS). Recent recommendations from major societies including the World Health Organization (WHO) and Extracorporeal Life Support Organization (ELSO) guide the consideration of extracorporeal membrane oxygenation (ECMO) for select populations.1,2 Here, we present our preliminary results for severe COVID-19 respiratory failure requiring ECMO support focusing on technical considerations, resource allocation, and specific anticoagulation approach.

Preliminary Experience of VV ECMO for Severe COVID-19 at Marie Lannelongue Hospital

From March 23 to May 5, we cannulated 24 patients (48.8 ± 8.9 years, 83% males) on venovenous (VV) ECMO for severe COVID-19. The Institutional Ethical Committee reviewed and approved the study. Main characteristics and outcomes are listed in Table 1. We considered VV ECMO in COVID-19 when respiratory failure persisted despite optimum management including 6 ml/kg controlled ventilation with a plateau pressure maintained below 30 cm H2O, use of neuromuscular blockers, high-positive end-expiratory pressure (PEEP), and repeated prone positioning sessions. Inhaled nitric oxide could be used awaiting VV ECMO initiation in case of severe oxygenation impairment. Contraindications for ECMO were an age greater than 60 years, chronic respiratory insufficiency, chronic heart failure, or any other comorbidities associated with life expectancy of less than 5 years. The average delay between onset of mechanical ventilation and VV ECMO was 6.3 ± 3.6 days. We used only polymethylpentene oxygenators and centrifugal pumps. Large bore (25–29 Fr) multiport cannulae were considered in COVID-19 patients to allow for optimal drainage as flow is proportional to the fourth power of the radius of the cannula (i.e., Poiseuille’s Law). Ultrasonographic evaluation was necessary to assess for vascular patency and facilitate percutaneous cannulation. A short (20 cm, single end hole) 17 or 19 Fr cannula was suitable for return in all cases. We did not use bicaval double-lumen cannulae because they usually provide limited blood flow (<5 L/min) that may not be suitable for optimal oxygenation in COVID-19 patients. Mean ECMO blood flow 24 hours after cannulation ranged from 4.3 to 7.2 L/min in our series with a mean sweep gas flow of 5.9 L/min (Table 1).

Table 1. - Demography, ECMO Settings and Outcomes
Demographics and Comorbidities
Males 20 (83%) High blood pressure 5 (21%)
Age (y–o) 48.8 [28–60] Diabetes mellitus 5 (21%)
BMI (kg/m2) 29.4 [22–44] Tobacco use 2 (8.3%)
BSA (m2) 2.05 [1.70–2.54] Dyslipidemia 4 (17%)
Characteristics of Severe COVID-19 and Management of Acute Respiratory Distress Syndrome
Delay between onset of symptoms and mechanical ventilation (days) 7.4 [2–14]
Delay between onset of mechanical ventilation and VV ECMO (days) 6.3 [1–11]
Lowest P/F ratio 67 [52–78]
Duration of mechanical ventilation (days) 34 [15–52]
ECMO Configuration and Settings
Femoro–jugular 22 (91.7%) Admission cannula
Femoro–femoral 1 (4.2%) 25 Fr 5 (21%) 29 Fr 19 (79%)
Bifemoro–jugular 1 (4.2%) Return cannula
17 Fr 9 (37%) 19 Fr 15 (62%)
ECMO flow at day 1 (L/min) 5.4 [4.3–7.2]
Sweep gas flow at day 1 (L/min) 5.9 [4–8]
ECMO FiO2 at day 1 (%) 95.6 [80–100]
Duration of ECMO support (days) 19 [1–39] Causes of death
Successful weaning 17 (71%) Massive PE 3
Pulmonary hemorrhage 4 (17%) Septic shock 3
Hemorrhagic stroke 1 (4.2%) Mesenteric ischemia 1
PE 6 (25%) Disabling stroke 1
BMI, body mass index; BSA, body surface area; COVID-19, coronavirus disease 2019; ECMO, extracorporeal membrane oxygenation; PE, pulmonary embolism; VV ECMO, venovenous extracorporeal membrane oxygenation.

Patients’ ventilation was adjusted in an ultraprotective fashion during ECMO support. Volume-controlled ventilation was switched to a pressure-controlled mode. Positive inspiratory pressure was set to 25–30 cm H2O and PEEP to 10–15 cm H2O. In our experience, static pulmonary compliance dropped dramatically in these patients. Therefore, tidal volumes were comprised between 1 and 4 ml/kg. Despite optimal ECMO settings, some patients remained hypercapnic and necessitated respiratory rates above 25 cycles/min. Ventilator FiO2 was set to the minimal value. Neuromuscular blockers were maintained, and repeated prone positioning was applied to all ECMO patients unless poor tolerance of this maneuver. Weaning from ECMO was considered if pulmonary compliance could reach tidal volumes of 6 ml/kg at least and if the patient’s oxygenation permitted to switch off the gas sweep flow for more than 24 hours.

At the beginning of our experience of VV ECMO in critical COVID-19 patients, we observed early clot formation inside the cannulae immediately after cannulation. Similarly, we had to deal with early membrane thrombosis within the first 24 hours in some cases. Therefore, we decided to administrate a bolus of 5,000 IU of intravenous heparin after vascular access but before cannulation, followed by the administration of an additional 3,500 IU of heparin directly into the circuit. Before circuit connection, we carefully flushed the cannulae with heparinized saline using 50-ml syringes to prevent in situ thrombosis. After ECMO initiation, a continuous infusion of heparin was rapidly started to maintain a target of anti-Xa activity from 0.4 to 0.6 IU/ml.

Three patients died early after ECMO implantation due to septic shock or multiorgan failure. Seventeen patients (71%) were successfully weaned from ECMO after a mean duration of 19.0 ± 10.1 days. Sixteen have been discharged from ICU, and one died from acute pulmonary embolism 6 days after weaning (Figure 1). Pulmonary hemorrhage under ECMO support was observed in four patients and was successfully managed by temporary disruption of heparin therapy for 3–5 days combined with bronchial artery embolization in one case. Circuit exchanges were performed for evidence of clot formation, consumption within the circuit (low fibrinogen < 2 g/L, thrombocytopenia < 50,000), or postoxygenator PaO2 < 200 mm Hg indicating impaired blood oxygenation. In our cohort of 24 patients, the average durability of the first circuit was 8.5 ± 5.1 days (Figure 1). Among the overall pool of 42 circuits implanted in our series, we experienced five circuit thromboses despite meeting target anticoagulation (anti-factor Xa activity >0.4 UI/mL).

Figure 1.
Figure 1.:
Representation of outcomes and circuit consumption over time of VV ECMO support in a preliminary cohort of 24 patients with severe COVID-19 respiratory failure. VV ECMO, venovenous extracorporeal membrane oxygenation; COVID-19, coronavirus disease 2019.

Recommendations and Published Literature for ECMO Use in Severe COVID-19

ECMO Team Protection.

As appropriate personal protective equipment remains a priority with this pandemic, the Surviving Sepsis Campaign and ELSO guidelines outline the use of fitted respirator masks, negative pressure rooms for aerosol-generating procedures, and the maximally experienced operators for airway management.3 An ASAIO Journal case series of ECMO in COVID-19 from China reported the utilization of level 3 infection control precautions consisting of powered air-purifying respirators, protective suits, surgical gowns, and three layers of sterile gloves.4 Infection control recommendations from the international multidisciplinary ELSO group further optimize safety via a number of systems-level interventions. These include colocalization of COVID-19 ECMO patients and positioning the ECMO control panel within view of the treatment team from the patient’s door.1

Indications for ECMO and Management of Mechanical Ventilation.

Consistent with ELSO guidelines recommending no deviation from usual indications for ARDS, the indication for VV ECMO were a partial pressure of arterial oxygen [PaO2] to the fraction of inspired oxygen [FiO2] ratio less than 50 mm Hg for more than 3 hours, a PaO2/FiO2 of less than 80 mm Hg for more than 6 hours, or an arterial blood pH of less than 7.25 with a partial pressure of arterial carbon dioxide [PaCO2] of ≥60 mm Hg for >6 hours.1 Recent observational work on mechanical ventilation in ECMO via the LIFEGARDS study has shown ultraprotective lung ventilation with ECMO can reduce the driving pressure and mechanical power delivered to the respiratory system.5

Cannulation Configurations and Circuit Considerations.

While optimal ECMO flow varies among patients, typically ECMO blood flows greater than 60% of cardiac output may be necessary to achieve adequate oxygenation.6 Configurations for cannulation are described in Figure 2 and do not differ from usual indications of VV ECMO. The high incidence of deep venous thromboses (DVT) and coagulation derangements have been well-described in current reports.7 Along with the increasing demand for ECMO support, there was potential for critical material shortage during the COVID-19 pandemic especially for large multiport drainage cannulae (≥27 Fr). Bifemoral cannulation using smaller cannulae (23 and 25 Fr) for venous drainage can be an alternative configuration (Figure 2), but may be associated with an increased risk of DVT due to impaired bilateral venous return from the lower limbs. Recirculation can be avoided when a 15 cm distance between the tips of two cannulae is respected. Persistent severe hypoxemia (PaO2 < 45 mm Hg) despite optimal blood flow (>5 L/min), and in absence of membrane thrombosis or recirculation, can also be managed using a second oxygenator. An entirely separate new circuit is therefore necessary, from which the second oxygenator is isolated and connected in parallel to the first oxygenator.8 This approach should, however, be applied with caution due to limited material resources during COVID-19 pandemic. Additionally, the parallel connection may increase turbulences within the circuit leading to increased risk of oxygenator thrombosis.

Figure 2.
Figure 2.:
A: Femoro–internal jugular configuration for venovenous extracorporeal membrane oxygenation (VV ECMO) with a long (>50 cm) and large multistage drainage cannula (25–29 Fr) inserted in the femoral vein (right or left) and a shorter (20 cm) and smaller single-stage return cannula (17–23 Fr) inserted the internal jugular vein (preferentially the right one) to be advanced at the junction between the superior vena cava and the right atrium. Recirculation of perfusate blood can be avoided when a minimum distance of 5 cm between the drainage and return cannula. B: Bifemoral-internal jugular configuration with two multistage drainage cannula in the femoral veins may be necessary in case of insufficient drainage by a single cannula. In bifemoral configurations, we recommend placing both the guidewires in the inferior vena cava (IVC) before introducing the first cannula. C: Femoro-femoral configuration for both drainage and return can be performed in absence of jugular venous access (thrombosis, local sepsis). In this configuration, a long (>50 cm) return cannula (21–23 Fr) with a single end hole is placed in the right atrium via a femoral vein, while a larger multistage drainage cannula is advanced into the distal IVC via the contralateral femoral vein.

Coagulopathy Profile of COVID-19 and Anticoagulation Protocol for VV ECMO.

Coagulopathy has been a prominent focus of COVID-19 given the association of abnormal coagulation parameters with poor outcomes.7 The pathophysiology is unknown, but hypotheses include sepsis-induced disseminated intravascular coagulopathy, and microthromboses in the pulmonary vascular bed with subsequent activation of the fibrinolytic cascade. The phenotype of coagulopathy associated with COVID-19 tends to be more thrombotic than hemorrhagic, since almost one third of ICU patients experience thrombotic complications.9 Furthermore, two independent autopsy studies have revealed thromboembolic events in 18 of 23 deceased patients, suggesting that close attention should be paid to anticoagulation on ECMO.10,11 This continues to shift the focus of anticoagulation on the thrombotic profile in COVID-19. In a case series of eight ECMO patients in Shanghai (seven VV and one VA ECMO), anticoagulation with heparin was titrated to activated clotted time 180–200 seconds and activated partial thromboplastin time 50–80 seconds.4 The overall burden of hemostatic abnormalities in COVID-19 associated with the prothrombotic nature of the ECMO circuit may increase the risk of thrombosis in these patients. Interim guidance on ECMO in COVID-19 by ELSO suggests titrating anticoagulation parameters to target the higher end of normal.1 This statement is based on pathophysiological considerations, while we await evidence-based guidelines and must be balanced with the increased risk of bleeding.

Our early experience of VV ECMO for severe COVID-19 underscore the need for high blood flows to achieve adequate oxygenation. The prothrombotic hematologic profiles of these patients should be considered, along with limitations in resource allocation during the pandemic. Our early results are encouraging but further reports of midterm outcomes are warranted.


The authors want to thank ICU teams involved in the management of severe COVID-19 patients at Stanford University Hospital and Marie Lannelongue Hospital.


1. Shekar K, Badulak J, Peek G, et al.: Extracorporeal Life Support Organization COVID-19 interim guidelines. ASAIO J. 2020.66:707–721
2. World Health Organization: Clinical Management of Severe Acute Respiratory Infection (SARI) When COVID-19 Disease Is Suspected: Interim Guidance, 13 March 2020. 2020.Geneva: World Health Organization;
3. Alhazzani W, Møller MH, Arabi YM, et al.: Surviving Sepsis Campaign: Guidelines on the management of critically ill adults with coronavirus disease 2019 (COVID-19) Intensive Care Med. 2020.46:854–887.
4. Li X, Guo Z, Li B, et al.: Extracorporeal membrane oxygenation for coronavirus disease 2019 in Shanghai, China. ASAIO J. 2020.66:475–481.
5. Schmidt M, Pham T, Arcadipane A, et al.: Mechanical ventilation management during extracorporeal membrane oxygenation for acute respiratory distress syndrome. An international multicenter prospective cohort. Am J Respir Crit Care Med. 2019.200:1002–1012.
6. Schmidt M, Tachon G, Devilliers C, et al.: Blood oxygenation and decarboxylation determinants during venovenous ECMO for respiratory failure in adults. Intensive Care Med. 2013.39:838–846.
7. Tang N, Li D, Wang X, Sun Z: Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia. J Thromb Haemost. 2020.18:844–847.
8. Sidebotham D, Allen SJ, McGeorge A, Ibbott N, Willcox T: Venovenous extracorporeal membrane oxygenation in adults: Practical aspects of circuits, cannulae, and procedures. J Cardiothorac Vasc Anesth. 2012.26:893–909.
9. Klok FA, Kruip M, van der Meer NJM, et al.: Incidence of thrombotic complications in critically ill ICU patients with COVID-19 Thromb Res. 2020.191:1148–1150.
10. Wichmann D, Sperhake JP, Lütgehetmann M, et al.: Autopsy findings and venous thromboembolism in patients with COVID-19 Ann Intern Med. 2020 [M20-2003, Epub ahead of print]
11. Lax SF, Skok K, Zechner P, et al.: Pulmonary arterial thrombosis in COVID-19 with fatal outcome: Results from a prospective, single-center, clinicopathologic case series. Ann Intern Med. 2020 [M20-2566, Epub ahead of print]

acute respiratory distress syndrome; extracorporeal membrane oxygenation; COVID-19; hemocompatibility; anticoagulation

Copyright © 2020 by the ASAIO