American Thoracic Society - We help the world breathe - PULMONARY :: CRITICAL CARE :: SLEEP
| | More

Extracorporeal membrane oxygenation (ECMO) in Patients with ARDS

Pauline K. Park M.D., James M. Blum M.D., Lena M. Napolitano, M.D.,
Gail Annich, M.D., Jonathan W. Haft, M.D., and Robert H. Bartlett, M.D.
University of Michigan Health System

Extracorporeal membrane oxygenation (ECMO) provides continuous cardiopulmonary support on a long-term basis, typically on the order of days to weeks, as adjunctive management of severe respiratory and cardiac failure.  The goal of therapy is to minimize ventilator-induced lung injury while allowing additional time to treat the underlying disease process and to permit recovery from acute injury.  ECMO is a complex technique and requires a dedicated team, appropriate equipment, institutional commitment and thorough preparation.  Potential complications are significant and its use is advocated only in patients who have substantial risk of death.

Historically, ECMO for influenza has been performed in neonatal and pediatric populations, with overall survival to discharge of 50% (1).  Recent reports of successful ECMO support in older H1N1 influenza patients with severe respiratory failure (2, 3) raise questions as to the role of extracorporeal support for adolescents and adults. 

During the 2009 Australia and New Zealand outbreak, the majority of H1N1 influenza cases receiving ECMO support were over 18 years of age, with a median age of 36 years (3).  At the time of the most recent report, 32% of the cohort remained alive in the hospital, 47% had survived to discharge home and 21% had died.  ECMO utilization was estimated at 2.6 per million population, or potentially 800 or 1300 cases if extrapolated to the U.S. and European populations. 
ECMO in Adult Respiratory Failure:

ECMO is established as standard of care for the management of neonatal and pediatric respiratory failure and lung transplantation, but is not widely used in adult acute respiratory distress syndrome (ARDS).  A randomized trial published in 1979 showed only 10% survival in both ECMO and control groups (4).  Over the ensuing 20 years, uncontrolled case series have reported 50% survival in ECMO for adult respiratory failure.  The most recent data include 20-year observational data from the Extracorporeal Life Support Organization (ELSO) registry (5), the UK CESAR randomized trial (6), the recent ANZ ECMO H1N1 experience (3) and the current ELSO H1N1 registry.

Historical Data:  1974 – 1994

Between 1974-77, the NIH sponsored a multicenter prospective, randomized trial of venoarterial (VA) ECMO versus conventional mechanical ventilation for adult patients with severe acute respiratory failure4. The overall survival for 686 hypoxemic patients was 34%; 90 of the sickest patients were entered into a randomized trial (7, 8).  Forty-eight were managed using conventional ventilation (including high FiO2 and high pressure) and 42 patients received conventional ventilation and venoarterial ECMO.  Survival was low in both treatment arms (9.5% vs. 8.3%) 

In 1986, Gattinoni reported a series of 43 patients managed with membrane lung support with low-flow extracorporeal CO2 removal (ECCO2R) combined with low frequency “rest-ventilation”. The survival was 48.8% (9).  Morris reported a randomized, controlled clinical trial of pressure-controlled inverse ratio ventilation and ECCO2R  which showed no difference in survival between the two arms (42% vs. 33%), however overall survival was significantly higher than in the previous decade (10, 11).

Given the negative results from these trials, enthusiasm for the use of ECMO in adult respiratory failure waned in the 1980s, and ECMO research continued primarily in pediatric and cardiac populations.  Advances in overall ICU care, ventilator management and ECMO technology render these early trial results not applicable to current practice.

Recent Data:  1986-2009

Retrospective review of the Extracorporeal Life Support Organization (ELSO) registry from 1986-2006 yielded data on 1,473 adult patients who received ECMO for severe respiratory failure (5).  Patients were not randomized to specific protocols and reporting was on a voluntary basis.  Advanced patient age, pre-ECMO arterial blood pH < 7.18, increased duration of pre-ECMO ventilation, decreasing patient weight, underlying cause of respiratory failure and complications on ECMO were associated with increased mortality.  Venovenous (VV) bypass increased survival compared to VA ECMO.  The median patient age was 34 years; the median time on ECMO was 154 hours.  Approximately 9% of patients sustained radiographic evidence of infarction, hemorrhage or brain death.  Overall survival to discharge was 50%.

The Conventional Ventilation or ECMO for Severe Adult Respiratory Failure (CESAR) trial was conducted in the United Kingdom between 2001 and 2006.  A “pragmatic“ randomized controlled trial design was used based on the previous trial of ECMO in neonatal respiratory failure conducted by the Oxford Clinical Trials group.  The trial randomized patients with severe, but potentially reversible, ARDS (Murray score > 3 or pH < 7.20) to either conventional care at one of 68 tertiary care centers or to a single center employing a treatment protocol that included ECMO.  Patients were excluded if they had been on high pressure or high FiO2 ventilation for more than 7 days, had signs of intracranial bleeding, contraindications to limited heparinization or contraindication to continuation of active treatment.  The trial was stopped for efficacy after 180 patients.

Of 90 patients randomized to the ECMO center, 71 (75%) ultimately received ECMO and 45 of these survived.  Three patients died in or pending transfer to the ECMO center.  Twenty two improved and recovered without ECMO.  Mean time of ventilation prior to study entry in the ECMO and conventional groups was 35 and 37 hours respectively; mean Murray scores were 3.5 and 3.4.  Median critical care days and hospital days were higher in the ECMO group (median, [interquartile range], 24 [13.0 – 40.5] and 35.0 [15.6 – 74.0]) than in the conventional group (13.0 [11.0-16.0] and 17.0 [4.8 – 45.3]).   Survival without severe disability at 6 months was 63% at 6 months for those patients referred to the ECMO center and 47% for those treated at tertiary care centers.  The “pragmatic” trial design has been criticized for inability to standardize mechanical ventilation management in the conventional care group and the fact that survival benefit is diminished if patients transferred, but not treated with ECMO, are excluded from the analysis.  However, CESAR did demonstrate that protocolized care including ECMO in an expert ARDS center yielded higher survival than the best standard care in tertiary ICUs in the UK.

Experience with ECMO in 2009 Influenza A (H1N1) ARDS was reported by the Australia and New Zealand Extracorporeal Membrane Oxygenation (ANZ ECMO) Influenza Investigators (3).  This observational study detailed their experience treating 201 ICU patients with confirmed or suspected 2009 Influenza A (H1N1) infection between June 1 and August 31, 2009.  Sixty-eight patients received ECMO for failure to improve on conventional treatment.  Forty-nine (72%) of these patients had treatment initiated at the referring hospital and transport on ECMO to the tertiary care center.  Prior to commencing ECMO, the patients were quite ill, with median [IQR] Murray score 3.8 [3.5-4.0], PaO2/fraction of inspired oxygen (FiO2) of 56 [48-63], lowest pH 7.2 [7.1-7.2], PEEP 18 [25-20] cm H20 and peak airway pressure 36 [33-38] cm.  The median time on ventilator prior to initiating ECMO was 2 days and median duration of ECMO support was 10 days.  Other salvage therapies (prone ventilation, HFOV, inhaled nitric oxide, prostacyclin) were utilized in 5 – 32% patients. 

Significant resource allocation was required, with 828 patient days of ECMO support utilized for the 68 patients.  Compared to the 133 patients who improved with conventional care, median days of mechanical ventilation were longer in patients treated with ECMO (18 [9-27] vs. 8 [4-14] days, p = .001), median ICU days were higher (22  [13-32] vs. 12 [7-18] days; p = .001) and ICU mortality was higher (23% vs. 9%; p = 0.01).  At the time of reporting, 48 (71%) of the ECMO patients had survived to ICU discharge, 14 (21%) had died, and 6 remained in the ICU.  Of the 22 patients still remaining in the hospital, 16 had survived to ICU discharge.  For the 46 patients for whom final discharge status was known, survival to hospital discharge was 70%. 

As of November 6, 2009, the ELSO H1N1 registry reported 91 additional H1N1 cases receiving ECMO support, approximately half pediatric and half adult. With outcomes reported in 65% of patients, overall survival was 64%.  In this series, survival was 75% when ECMO was instituted before 7 days of mechanical ventilation, compared with 30% when ECMO was instituted after 7 days of mechanical ventilation; survivors have been reported when ECMO was instituted after 2 weeks of intubation. 

Though ECMO has not clearly been demonstrated to improve outcome compared to standard of care management of adult ARDS, modern experience with ECMO for adult respiratory failure continues to accrue, with recent reported hospital survival rates of 50-70%.  Advances in ICU care, ventilator management and ECMO devices and technology all contribute to the improvement in overall outcomes.

Methods of ECMO support:

Thorough discussion of ECMO indications, management, complications, and outcome are found on the ELSO web site:

Components utilized in ECMO include 1) pump, 2) oxygenator and 3) vascular access to circuit with the patient’s native circulation.  Currently, the most efficient systems utilize a small centrifugal pump and a low-resistance polymethylpentene-oxygenator. Based on type of access, there are two primary forms of ECMO, venovenous (VV) and venoarterial (VA).  The artificial lung is used in series (VV) or in parallel (VA) with the native lungs, depending on the indications.

In VV ECMO, venous blood is withdrawn through a large bore venous cannula (21-25 Fr). Oxygen is added and CO2 removed, and oxygenated blood is returned to the venous circulation close to the right atrium.  VV ECMO is accomplished through either two single lumen catheters, typically placed via the right internal jugular and femoral veins, or one bicaval dual lumen catheter (VVDL, 27 – 31 Fr) placed via the right internal jugular vein.  This technique offers respiratory support with reduced hemodynamic effects, a low risk of ischemia from embolic phenomena or reduced extremity flow, and is utilized in the majority of adult patients with acute respiratory failure.

VA ECMO provides both respiratory and cardiac support.  Blood is withdrawn from the venous circulation, oxygen is added and CO2 removed, and returned to the patient's arterial circulation.  Depending on the age and condition of the patient, a variety of access choices are available.  In the adult, femoral venous and femoral arterial cannulation is preferred.  A number of additional potential complications exist compared to VV ECMO.  Ischemic complications (arterial ischemia of the extremity or other organs) may accompany arterial cannulation.  VA ECMO directs oxygenated blood retrograde through the aorta; if flow is insufficient, it may not reach the proximal circulation, and thrombotic or air emboli may be perfused into the systemic circulation.

Candidates for ECMO:

Timing of referral in  patients: If ECMO is to be used in an adult with ARDS, the decision to transfer to an ECMO center should be considered only if the patient is not responding to optimal management and has a mortality risk judged to be greater than 50%. This remains difficult to quantitate since patients can progress from healthy status to moribund septic shock in 24 hours and transportation on ECMO is not generally available as in the ANZ ECMO study.  Accordingly, if transfer to an experienced ECMO center is to be considered, it should be done early in the patients’ course. 

For reference, the inclusion criteria from the CESAR trial were:  Adult patients 18-65 years old, with Murray score > 3 or uncompensated hypercarbia with pH < 7.20.

The Murray score (12) uses the average score of 4 elements graded on a 0 - 4 scale to establish ARDS severity.  It is calculated as follows:







(FiO2 at 1.0 for at least 20 minutes)







Number of quadrants with infiltrates












(ml/cm H2O)






The following website contains a calculator to assist in the computation:

Absolute contraindications to ECMO include:
Ongoing terminal disease that will not resolve or stabilize, contraindication to anticoagulation, intracranial hemorrhage, refusal to receive blood products

Relative contraindications to ECMO (due to historically poor survival rates): Mechanical ventilatory support for > 10 days and high pressure mechanical ventilatory support for > 7 days

Management on ECMO

Supportive therapy is continued, including prompt provision of appropriate antivirals and antibiotics, nutritional support and volume management to approximate dry weight.  Ventilator settings, including pressures, tidal volume, rate and FiO2 are lowered as support permits.  Systemic anticoagulation with heparin to prevent circuit clotting is required, titrated by bedside anticoagulation monitoring.  Oximetric assessment of venous circuit “SvO2” can be utilized to guide support. The use of continuous renal replacement therapy for fluid overload or renal failure is common.

Complications of ECMO:

Complications on ECMO are common and are associated with increased mortality.  The most frequent complication during ECMO is hemorrhage.  Fatal vascular perforation can occur during cannulation.  Bleeding is managed by decreasing or stopping heparin and infusion of platelets and clotting factors. Vascular access should be attempted with ultrasound guidance. Careful hemostasis with electrocautery is used for invasive procedures such as tracheostomy.  Avoidance of invasive procedures such as tube thoracostomy, if possible, is ideal and limits complications. 

Pulmonary hemorrhage is seen commonly in patients on ECMO.  Management includes bleeding control as above, and frequent bronchoscopy to clear the airway over time.

Intracerebral hemorrhage or infarction occurs in approximately 10-15% of ARDS  patients on ECMO.  Forty-three percent of the deaths in the ANZ ECMO series were related to intracranial hemorrhage.  Due to the requirement for ongoing anticoagulation and the condition of patients, this complication is usually fatal. 

Hemolysis does not occur during ECMO unless there is a problem in the circuit or the patient.  Plasma free hemoglobin should be checked daily; values over 10 mg% require further investigation in identifying and repairing the cause.

Thromboembolism:  Following completion of the patient’s ECMO run, continued anticoagulation of the H1N1 patient should be considered due to reports of high incidence of pulmonary thromboemoblic disease (2).  If the patient has had a femoral venous cannula in place for VV-ECMO, insertion of an IVC filter should be considered at the time of decannulation.


Successful management of severe respiratory failure requires high-level, intensive care support.  The patient population is young, often with much salvageable quality of life, but at the cost of significant critical care resource utilization.  Though ECMO has not clearly been demonstrated to be better than the standard of care for ARDS, referral to a specialized center with ECMO experience should be considered early after the initiation of high-level ventilator support. Centers considering implementing ECMO as a rescue therapy should actively seek training and guidance as early as possible.  The ELSO website contains a member list with contacts, management guidelines, references and training/education materials.  In the event epidemic conditions arise, substantial resource and triage decisions will likely be required.


ELSO: The ELSO website contains a member list with contacts, management guidelines, references and training/education materials.

ELSO H1N1 registry:


  1. ELSO. unpublished
  2. Intensive-care patients with severe novel influenza A (H1N1) virus infection - Michigan, June 2009. MMWR Morb Mortal Wkly Rep 2009;58:749-52.
  3. Extracorporeal Membrane Oxygenation for 2009 Influenza A(H1N1) Acute Respiratory Distress Syndrome. JAMA 2009.
  4. Zapol WM, Snider MT, Hill JD, et al. Extracorporeal membrane oxygenation in severe acute respiratory failure. A randomized prospective study. JAMA 1979;242:2193-6.
  5. Brogan TV, Thiagarajan RR, Rycus PT, Bartlett RH, Bratton SL. Extracorporeal membrane oxygenation in adults with severe respiratory failure: a multi-center database. Intensive Care Med 2009.
  6. Peeak GJ, Mugford M, Tiruvoipati R, et al. Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial. Lancet 2009.
  7. Bartlett RH, Morris AH, Fairley HB, Hirsch R, O'Connor N, Pontoppidan H. A prospective study of acute hypoxic respiratory failure. Chest 1986;89:684-9.
  8. National Heart Lung and Blood Institute (NIH) Public Health Service.  Extracorporeal support for respiratory insufficiency:  A collaborative study. DHEW Publication 1979.
  9. Gattinoni L, Pesenti A, Mascheroni D, et al. Low-frequency positive-pressure ventilation with extracorporeal CO2 removal in severe acute respiratory failure. JAMA 1986;256:881-6.
  10. Morris AH, Wallace CJ, Menlove RL, et al. Randomized clinical trial of pressure-controlled inverse ratio ventilation and extracorporeal CO2 removal for adult respiratory distress syndrome. American journal of respiratory and critical care medicine 1994;149:295-305.
  11. Suchyta MR, Clemmer TP, Orme JF, Jr., Morris AH, Elliott CG. Increased survival of ARDS patients with severe hypoxemia (ECMO criteria). Chest 1991;99:951-5.
  12. Murray JF, Matthay MA, Luce JM, Flick MR. An expanded definition of the adult respiratory distress syndrome. Am Rev Respir Dis 1988;138:720-3.

Critical Care