ECMO: How Extracorporeal Membrane Oxygenation Works, What the Evidence Shows

The Scene

The call comes from the ICU at 11 PM. A 47-year-old woman from Naperville has been on a ventilator for six days with COVID-19 pneumonia, her second pandemic infection, and she is failing despite prone positioning, neuromuscular blockade, and inhaled nitric oxide. Her PaO2 on 100% FiO2 is 52 mmHg. Her right ventricle, visible on bedside echo, is dilated and nearly akinetic. She has no cardiac history. She is a kindergarten teacher, a mother of two school-age children, and she was healthy eight days ago.

The ARDS team has been managing her correctly; there is nothing more to do within conventional ventilator management that has not already been tried. The question now is whether to initiate veno-venous ECMO: bypass the lungs entirely, drain deoxygenated blood from her venous circulation, pass it through an artificial membrane oxygenator outside the body, and return oxygenated blood to her right atrium or inferior vena cava while her native lungs are given time to recover.

The ECMO team at Northwestern Memorial in Chicago is on the phone within 10 minutes. Transport to a capable center is arranged. She is cannulated 3 hours later. Her PaO2 climbs to 88 mmHg within 20 minutes of ECMO initiation. Her cardiac output, which had collapsed from RV failure in the setting of severe hypoxia, improves as her RV afterload drops with improved oxygenation.

This is ECMO at its most straightforward: a patient with reversible respiratory failure, without primary cardiac failure, placed on veno-venous ECMO as a bridge to lung recovery. The ethical situation is relatively clear. The technical situation is demanding but defined.

Seven hours earlier, in a different room in the same ICU, a different patient had a different conversation. A 72-year-old man with ischemic cardiomyopathy and LVEF 18%, admitted for cardiogenic shock following an anterior MI, had been placed on veno-arterial ECMO when his pressure collapsed despite PCI and maximal vasopressors. His coronary anatomy was revascularized. His right heart catheterization showed CI of 1.1 L/min/m2 and a pulmonary artery saturation of 35%. He was on ECMO, and it was keeping him alive. But his repeat echo at 12 hours showed no meaningful recovery of the akinetic anterior wall. His creatinine had risen to 4.1. His lactate remained 8.1. He was not going to recover his myocardium. He was not a transplant candidate. He had no LVAD indication at his age and comorbidities. He was on a machine that was keeping his blood circulating until his kidneys and liver failed completely.

This is the “bridge to nowhere.” It is the most difficult ethical reality in ECMO medicine.

What It Is

ECMO (extracorporeal membrane oxygenation) is a form of life support that uses an external circuit containing a pump and a membrane oxygenator to perform the gas exchange and, in its cardiac configuration, the circulatory pumping function of the heart and lungs.

Two distinct configurations exist:

Veno-venous ECMO (VV-ECMO): provides respiratory support only. Blood is drained from the venous circulation (typically the femoral vein or internal jugular vein), passes through the membrane oxygenator (where oxygen is added and carbon dioxide is removed), and is returned to the venous circulation (typically the right atrium via a dual-lumen cannula or a separate jugular vein cannula). VV-ECMO does not provide cardiac output; the patient’s heart must still pump. It is used for severe respiratory failure (ARDS, severe pneumonia, refractory hypoxemia).

Veno-arterial ECMO (VA-ECMO): provides both respiratory and hemodynamic support. Blood is drained from the venous circulation and returned to the arterial circulation (typically the femoral artery), bypassing the pulmonary circuit and the left ventricle. VA-ECMO provides cardiac output directly and is used for cardiogenic shock, cardiac arrest, and conditions causing both cardiac and respiratory failure.

The oxygenator is the central component: a hollow-fiber membrane module through which blood flows on one side and a sweep gas (oxygen-enriched air) flows on the other. Oxygen diffuses from the sweep gas across the membrane into the blood; carbon dioxide diffuses from the blood into the sweep gas and is exhausted. This is the same gas exchange principle as the alveolus, but accomplished outside the patient’s body.

ECMO circuits are manufactured by several companies. The leading platforms in clinical use include:

• Maquet Cardiohelp (Getinge, FDA 510(k) cleared): a highly miniaturized portable ECMO system widely used at major centers for both VV and VA configurations
• LivaNova ROTA and Sorin Revolution circuits (LivaNova, FDA 510(k) cleared)
• Medtronic CARMEDA BioActiveSurface circuits (FDA 510(k) cleared)
• Terumo ECMO systems (FDA 510(k) cleared)

ECMO circuits, oxygenators, and cannulae are Class II devices cleared through the 510(k) pathway. The clinical management protocols that determine who receives ECMO, how it is managed, and when it is discontinued are institutional protocols rather than device approvals.

The Mechanism

3.1 VV-ECMO: Lung Bypass

In VV-ECMO, the native lungs are bypassed for gas exchange while maintaining perfusion. The circuit operates at 3 to 6 liters per minute of blood flow. Key physiologic parameters:

• ECMO fraction (FiO2 of sweep gas): typically 1.0 (pure oxygen) initially, weaned as lung recovery occurs
• Sweep gas flow rate: determines CO2 removal (higher sweep gas = lower PaCO2)
• Blood flow rate: determines O2 delivery (higher blood flow = more oxygenation)
• Recirculation: the major technical challenge specific to VV-ECMO. When the return cannula is positioned too close to the drainage cannula, oxygenated blood may be immediately drained back into the circuit before it reaches the heart (recirculation), reducing effective oxygenation. Monitoring is essential.

During VV-ECMO, the lungs are managed with “ultra-protective” ventilation: extremely low tidal volumes (2-4 mL/kg), low respiratory rates, and low PEEP, allowing inflamed, injured lung tissue to rest and recover without the barotrauma of conventional ventilation.

3.2 VA-ECMO: Cardiac and Respiratory Bypass

In VA-ECMO (peripheral femoro-femoral configuration, the most common emergency setup), the femoral vein drains desaturated blood into the circuit; oxygenated blood is returned through the femoral artery into the descending aorta in a retrograde direction (against normal blood flow direction). The circuit provides flow of 3 to 6 liters per minute, depending on cannula size and patient anatomy.

VA-ECMO provides systemic perfusion, but it also significantly increases left ventricular afterload: the retrograde blood from the femoral artery pushes against the left ventricle’s ejection, requiring the LV to work against a higher pressure. In a severely compromised LV, this afterload augmentation can cause left ventricular distension (LV dilation), increased wall stress, pulmonary edema, and myocardial oxygen demand that may exceed supply. This is the Harlequin syndrome problem (also called North-South syndrome): the upper body is perfused by the failing native heart with poorly oxygenated blood while the lower body receives well-oxygenated ECMO blood, creating an oxygen saturation gradient between the right arm and the lower extremities.

LV venting strategies address left ventricular distension during VA-ECMO:

• Addition of an Impella device in the LV (the “ECPELLA” or “IMPELLA-ECMO” combination): the Impella drains from the LV into the ascending aorta, reducing LV volume and pressure
• Intra-aortic balloon pump (IABP) for modest afterload reduction
• Surgical LV vent through the LVAD inflow position

3.3 The Oxygenator as Artificial Lung

The hollow-fiber oxygenator contains thousands of microporous hollow fibers bundled together. Blood flows outside the fibers; sweep gas flows inside. The fiber wall is gas-permeable but liquid-impermeable, allowing gas exchange without blood-gas mixing. The oxygenator also functions as a heat exchanger: the circuit includes a heater/cooler that maintains blood temperature during support. In cardiac arrest with targeted temperature management, the circuit can cool the blood to 33-36 degrees Celsius.

How It Is Used

4.1 Indications for VV-ECMO

VV-ECMO is indicated for severe, refractory ARDS or other respiratory failure causes when conventional ventilatory support (including prone positioning, PEEP optimization, neuromuscular blockade, and inhaled vasodilators) has failed to maintain acceptable oxygenation or ventilation. The Berlin criteria for severe ARDS (PaO2/FiO2 ratio below 100 on PEEP at least 5 cmH2O) in patients below age 65 without end-stage organ disease constitutes the typical threshold at most centers.

A validated prediction tool for VV-ECMO candidacy is the RESP score (Respiratory ECMO Survival Prediction), which uses age, immunocompromised status, hospitalization duration before ECMO, etiology of respiratory failure, and other factors to predict in-hospital survival on VV-ECMO 4 / Promising . A RESP score below -2 predicts less than 33% survival; centers vary in whether they implant ECMO in such patients.

4.2 Indications for VA-ECMO

VA-ECMO is used in:

1. Refractory cardiogenic shock unresponsive to vasopressors and intraaortic balloon pump support, from any cause (AMI, myocarditis, decompensated cardiomyopathy, post-cardiotomy shock)
2. Refractory cardiac arrest (E-CPR: ECMO-facilitated CPR): VA-ECMO initiated during cardiac arrest to maintain perfusion during or after failed ACLS, as a bridge to identifying and treating the underlying cause
3. Massive pulmonary embolism with cardiac arrest or severe hemodynamic compromise: VA-ECMO maintains perfusion while thrombolytics lyse the embolic burden
4. Severe pulmonary hypertension crises causing acute RV failure
5. Bridge to VAD or transplant in patients with refractory cardiogenic shock who are potential candidates for definitive support

4.3 Cannulation Strategies

Peripheral femoro-femoral VA-ECMO (most common emergency setup): femoral vein drainage, femoral artery return. Can be performed percutaneously (Seldinger technique) at the bedside during resuscitation. Risk: limb ischemia from the arterial cannula occluding the femoral artery’s distal perfusion; a distal perfusion catheter (DPC) inserted into the superficial femoral artery distal to the access site is used to preserve limb flow.

Dual-lumen VV-ECMO (Avalon or OriGen BiCaval dual-lumen cannula): a single large-bore cannula placed in the right internal jugular vein, with drainage holes in the superior and inferior vena cava and a return jet directed at the tricuspid valve. Requires fluoroscopic or echocardiographic guidance for positioning. Allows patient mobilization during VV-ECMO support, which is critical for lung recovery.

Central VA-ECMO (in cardiac surgery settings): direct right atrial drainage and ascending aortic return cannulation, used in post-cardiotomy shock when the chest is already open.

4.4 Anticoagulation

All ECMO circuits require anticoagulation to prevent thrombus formation in the circuit. Unfractionated heparin continuous infusion, titrated to an activated clotting time (ACT) of 180 to 220 seconds or anti-Xa level of 0.3 to 0.5 IU/mL, is the standard. Bleeding from anticoagulation in an already critically ill patient is common; the management requires continuous titration of the heparin dose against the competing risks of circuit clotting and patient hemorrhage.

The Evidence

5.1 VV-ECMO in ARDS: CESAR and EOLIA

The evidence for VV-ECMO in ARDS comes from two major RCTs, neither of which provides definitive evidence of mortality benefit for ECMO itself.

CESAR trial (UK, 2009): randomized 180 adult ARDS patients to transfer to an ECMO center versus conventional management at referring hospitals 5 / Solid 61069-2). At 6 months, 63% of ECMO-center patients survived without severe disability versus 47% of conventional management patients (RR 0.69, p=0.03). The CESAR trial has the critical limitation that it randomized to transfer to an ECMO center, not to ECMO per se; patients at the ECMO center received ECMO-center-level care even if not placed on ECMO. Thus, CESAR demonstrates the benefit of ECMO-center transfer but not ECMO itself.

EOLIA trial (multinational, 2018): randomized 249 severe ARDS patients to immediate VV-ECMO versus target conventional management with ECMO rescue allowed for conventionally managed patients who met predefined deterioration criteria 5 / Solid . ECMO reduced 60-day mortality from 57% to 35% in the intention-to-treat analysis (RR 0.76, 95% CI 0.55-1.04; p=0.09). The trial did not reach statistical significance despite the absolute mortality reduction of 22 percentage points, because of trial underpowering and because 28% of conventionally managed patients crossed over to ECMO, diluting the between-group difference. A Bayesian analysis of the EOLIA data found a 97% posterior probability that ECMO reduces mortality 4 / Promising .

The practical consensus: for severe ARDS unresponsive to target conventional ventilation, VV-ECMO at a high-volume experienced center improves outcomes, but the strongest evidence is for ECMO-center transfer rather than ECMO itself, and not all severe ARDS patients benefit equally.

5.2 VA-ECMO in Cardiogenic Shock: The Absence of RCT Evidence

As of 2026, there is no completed randomized controlled trial demonstrating that VA-ECMO improves mortality in cardiogenic shock 5 / Solid . The ECMO-CS trial (Czech Republic, planned enrollment 120 patients with cardiogenic shock) and the ECLS-SHOCK trial (Germany, multinational) are ongoing or recently completed.

ECLS-SHOCK trial (presented 2023, now published): randomized 420 patients with acute MI cardiogenic shock to VA-ECMO plus revascularization versus revascularization alone 5 / Solid . The primary outcome (30-day all-cause mortality) did not differ between groups (47.8% ECMO vs 49.0% control, HR 0.98, p=0.81). Major bleeding and peripheral vascular complications were higher in the ECMO group.

This is a critically important finding. VA-ECMO in AMI-cardiogenic shock, when initiated after PCI for hemodynamic support, does not reduce 30-day mortality compared to revascularization alone, and it increases bleeding and vascular complications. This finding mirrors the RECOVER II/DMCS trial finding for Impella (see DEVI-019). The implication: hemodynamic support with VA-ECMO or Impella after PCI may stabilize hemodynamics without improving myocardial recovery or survival in established AMI-cardiogenic shock.

The important question, not yet answered by available RCTs, is whether early initiation (before or during PCI rather than after) might change the equation by providing hemodynamic stability during revascularization. This is the hypothesis being tested in ongoing trials.

5.3 ECMO in Massive Pulmonary Embolism

In massive PE with cardiac arrest or severe hemodynamic collapse, VA-ECMO provides a bridge to treatment while systemic or catheter-directed thrombolytics lyse the embolic burden. The evidence is entirely observational (registry-level), driven by the impossibility of RCTs in a situation where withholding support is not ethically feasible 4 / Promising . Survival rates of 50 to 65% have been reported in small series at experienced centers.

The Patient Experience

6.1 What ECMO Looks Like From the Bedside

A patient on VA-ECMO is almost always sedated and intubated. The circuit is visible at the bedside: red tubing (arterial) and blue tubing (venous), the small centrifugal pump head spinning visibly, the hollow-fiber oxygenator with blood flowing through its translucent housing. The machine makes a low continuous whirring sound. The patient’s color is often better than expected given their hemodynamic state; the machine is doing the work.

Family members in the ICU room see the circuit and have questions. “Is that machine her heart now?” The honest answer: it is doing part of the work her heart would do. It is buying time. The question we do not yet know the answer to is how much time, and what will be at the end of that time.

For the patient on VV-ECMO who is awake and breathing independently (ambulatory ECMO, increasingly practiced for longer-duration support in lung failure), the experience is different: they sit up, walk with assistance, eat, and engage in respiratory therapy to maintain muscle function while their lungs recover. Ambulatory VV-ECMO has been associated with better transplant outcomes in patients bridged to lung transplant 4 / Promising .

6.2 The ECMO Team

ECMO management is a team sport. The ECMO specialist (usually a perfusionist or an intensivist-trained nurse or respiratory therapist with specialized ECMO training) monitors the circuit continuously. The intensivist manages the mechanical ventilation, hemodynamics, anticoagulation, and sedation. The cardiology or cardiac surgery team evaluates daily whether escalation (to LVAD or transplant), de-escalation (ECMO weaning), or palliation is appropriate.

At Northwestern Memorial Hospital in Chicago, University of Illinois Hospital, Rush University Medical Center, and Loyola University Medical Center, ECMO teams operate 24/7 with institutional protocols for initiation, management, and weaning. Regional ECMO programs accept transfer patients from non-ECMO hospitals throughout Illinois, including from Carle Foundation Hospital in Urbana for cases that exceed the capabilities of a complete community cardiology program.

6.3 Sex Differences

Women represent a minority of cardiogenic shock ECMO recipients (approximately 25 to 30%) in large registries, reflecting the lower prevalence of ischemic cardiomyopathy in women presenting with cardiogenic shock in the acute phase 4 / Promising . Among VV-ECMO recipients for ARDS, sex distribution is more equal. In available ECMO registry data, women on VA-ECMO have higher rates of limb ischemia from femoral artery cannulation due to smaller femoral artery caliber, underscoring the importance of distal perfusion catheter use and appropriate cannula sizing.

Decisions and Trade-Offs

7.1 The Bridge to Nowhere: When to Stop ECMO

The most difficult decision in ECMO management is recognizing when continued support is providing biological bridge to nothing recoverable. The indicators:

• Neurological catastrophe: confirmed brain death or severe anoxic brain injury precludes any meaningful recovery; ECMO should not be continued
• Multiorgan failure progression despite support: rising creatinine, worsening bilirubin, rising lactate despite adequate ECMO flow all suggest the metabolic failure is not correctable
• No cardiac recovery: serial echocardiograms showing no improvement in LV or RV function after 5 to 7 days of VA-ECMO support, in a patient who is not a transplant or LVAD candidate, defines the bridge to nowhere

The conversation with the family must be explicit: “The machine is keeping your husband’s blood circulating, but his heart is not recovering. He is not a candidate for a transplant or an artificial heart. If we continue the machine, we will be prolonging his dying rather than his living. We need to talk about what he would want.”

Palliative care involvement is mandatory before ECMO is discontinued. The ECMO weaning process (gradual reduction of flows over hours to days, watching for cardiac response) allows the team to confirm irreversibility before the final step.

7.2 ECMO Before vs. After PCI in Cardiogenic Shock

The ECLS-SHOCK results have driven significant practice reconsideration. If VA-ECMO after PCI does not improve mortality in AMI-cardiogenic shock, should it ever be used? The clinical consensus after ECLS-SHOCK is nuanced:

• VA-ECMO remains appropriate for cardiogenic shock with refractory cardiac arrest (E-CPR)
• VA-ECMO remains appropriate as a bridge to LVAD or transplant in patients who are candidates for definitive support
• Routine VA-ECMO for hemodynamic support in AMI-cardiogenic shock after PCI is not supported by current RCT evidence
• The question of whether ECMO before PCI (unloading before reperfusion) might improve outcomes remains unanswered and is being studied

7.3 Cost and Access Disparities

ECMO is extraordinarily resource-intensive: the circuit and its components cost $25,000 to $50,000; the daily management by specialized ECMO nurses and perfusionists adds $3,000 to $5,000 per day; 14 days of ECMO support can cost $150,000 to $300,000 before considering intensive care costs. These costs are borne by the hospital and reimbursed imperfectly by payers, creating financial stress at ECMO centers even in academic institutions.

Access to ECMO is almost entirely determined by geography: ECMO is available only at centers with the infrastructure and training to manage it. A patient in rural southern Illinois who arrests in a small community hospital has a very different access profile than a patient who arrests at Northwestern Memorial. The ECMO-CS designation system recognizes centers of excellence, but it does not solve the fundamental access gap for patients who cannot be transported in time.

Clinical Synthesis

ECMO occupies the final node in the cardiovascular care graph: it is what happens when everything else has failed. The kindergarten teacher from Naperville who developed ARDS from a second COVID-19 infection is an example of ECMO for a non-preventable acute illness. The 72-year-old man with ischemic cardiomyopathy who arrested after his anterior MI represents ECMO used in a context that this clinical framework would have aimed to prevent.

His MI was his second. His first, five years earlier, had been managed with stenting and discharged without a complete secondary prevention program. His LDL was 148 mg/dL at discharge. He was on a moderate-intensity statin but not a high-intensity statin, and not a PCSK9 inhibitor, and his ApoB was never measured. He was not on an SGLT2 inhibitor. He developed heart failure in the interval between the two MIs but was seen only three times in five years by a cardiologist. He was not in an advanced heart failure program. He was not on sacubitril/valsartan.

The anterior MI that sent him to VA-ECMO at a lactate of 8 did not arrive without warning. It arrived after five years of suboptimally treated atherosclerosis in a patient with a previous MI and established cardiomyopathy, who was not enrolled in a program designed to prevent exactly this event.

That is the clinical argument. ECMO is not preventable in every case. But in a meaningful fraction of cases, the upstream failures that led to the moment of ECMO cannulation were identifiable, addressable, and unaddressed. A structured cardiovascular assessment for his phenotype, five years before, would have identified his LDL/ApoB discordance, flagged the need for high-intensity statin or PCSK9 inhibitor, recommended an SGLT2 inhibitor for secondary prevention, and enrolled him in a quarterly follow-up program that would not have allowed his LVEF to decline from 45% to 18% without intervention.

Preventive cardiology cannot prevent death. It can prevent the specific sequence of omissions that turned a patient with a manageable chronic condition into a patient on a machine at the center of a bridge to nowhere conversation.

Appendix: Extended Clinical Notes

A.1 VA-ECMO Circuit Components and Their Clinical Implications

The VA-ECMO circuit consists of five functional components: drainage cannula (venous), return cannula (arterial), centrifugal pump head, membrane oxygenator, and heat exchanger. Each component has specific clinical implications that become relevant to bedside management 5 / Solid 61069-2).

The drainage cannula is typically placed in the right femoral vein (for peripheral VA-ECMO) and advanced to the right atrium under fluoroscopic or echocardiographic guidance. Target position has the cannula tip at the junction of the inferior vena cava and right atrium, enabling drainage of both inferior and superior vena caval return. A cannula positioned too far into the right atrium will drain preferentially from the superior vena cava, reducing total venous return and limiting circuit flow. A cannula pulled back into the common femoral vein will have higher resistance and will cause IVC suction events (characterized by high-pitched pump chattering and abrupt flow reduction) 4 / Promising .

The membrane oxygenator contains several thousand hollow fibers through which blood flows while oxygen passes countercurrent on the exterior surface. Gas transfer efficiency is measured by the sweep gas flow rate and the fraction of inspired oxygen (FiO2) in the sweep gas. Sweep gas flow primarily controls CO2 removal: higher sweep flow increases CO2 clearance; lower sweep flow allows CO2 to accumulate (used to maintain permissive hypercapnia in some VV-ECMO weaning protocols). FiO2 in the sweep gas controls the post-oxygenator blood oxygen content 5 / Solid .

Oxygenator failure (plasma leak through the fiber membranes) is recognized by a progressive increase in the pressure drop across the oxygenator (pre- to post-membrane), decreased post-membrane PO2 at constant sweep gas settings, and visible plasma exudation into the sweep gas circuit. Oxygenator replacement requires brief circuit pause or changeover, a procedure that requires practice and coordination and cannot be improvised at the bedside 4 / Promising Guidelines 2021; elso.org).

A.2 The North-South Syndrome (Differential Hypoxia) in VA-ECMO

The north-south syndrome (also called Harlequin syndrome or differential hypoxia) is a unique complication of peripheral femoral VA-ECMO that occurs when cardiac function recovers partially during ECMO support. As the native heart begins to eject, well-oxygenated blood from the ECMO return cannula (at the iliac artery level) competes with poorly oxygenated blood ejected by the recovering heart through the aortic valve 5 / Solid .

The clinical presentation is striking: the lower body (below the aortic arch, perfused by the ECMO return cannula) is well-oxygenated with normal SpO2 readings from a pulse oximeter on the foot, while the upper body (perfused by the native heart ejecting through the coronary arteries and carotid vessels) receives hypoxic blood from the injured lungs, causing cerebral and myocardial hypoxia that is not detected by standard pulse oximetry on the finger 5 / Solid .

Detection requires: continuous right-hand or ear pulse oximetry (upper body), arterial blood gas from the right radial artery (as a surrogate for coronary/cerebral perfusion), and regular echocardiographic assessment of aortic valve opening. Prevention and management involve: (1) improving lung function on the ventilator to increase pulmonary venous blood oxygen content; (2) adding a superior return cannula (right internal jugular vein drain, right subclavian artery return) to create a VAV-ECMO configuration that delivers oxygenated blood to the ascending aorta; or (3) transitioning to central (trans-thoracic) cannulation 4 / Promising . At Northwestern Memorial Hospital in Chicago, a right radial arterial line is placed as standard practice for all VA-ECMO patients within the first hour of cannulation to enable continuous monitoring of upper body oxygenation.

A.3 Weaning VA-ECMO: Protocols and Decision Points

Weaning from VA-ECMO requires demonstration that the native cardiopulmonary system can maintain adequate perfusion and oxygenation without circuit support. Weaning is both a diagnostic process (does the heart tolerate reduced support?) and a therapeutic decision (should we remove support, and when?) 5 / Solid Guidelines 2021; elso.org).

A structured VA-ECMO weaning trial at most U.S. centers involves reducing circuit flow in decrements of 0.5 L/min every 30-60 minutes while monitoring for: maintenance of mean arterial pressure above 65 mmHg without escalating vasopressor doses, cardiac index above 2.0 L/min/m2 by thermodilution or Fick method, pulmonary artery occlusion pressure below 18 mmHg, and absence of new ST changes indicating myocardial ischemia during the reduced-flow period 4 / Promising . Echocardiography during weaning trials assesses left ventricular ejection, right ventricular function, and aortic valve opening frequency at each flow reduction step.

Successful weaning at 1.0-1.5 L/min for 30-60 minutes typically indicates readiness for cannula removal. Failure to wean (hemodynamic deterioration at flows above 2.5 L/min) indicates either insufficient myocardial recovery to support explantation, or the need to assess for transition to durable LVAD therapy or cardiac transplantation listing 5 / Solid Guidelines 2021; elso.org). At Carle Foundation Hospital, ECMO weaning trials are conducted with continuous right heart catheterization monitoring and are scheduled for morning hours when the full cardiac surgery and perfusion team is available to manage any emergent cannulation issue that arises during the wean 3 / Early .

A.4 Anticoagulation on ECMO: The Heparin Tightrope

Anticoagulation during ECMO is essential to prevent circuit thrombosis but creates a competing risk of bleeding in a patient population that is already hemodynamically fragile and frequently post-surgical. The standard anticoagulant is unfractionated heparin (UFH), managed by continuous infusion titrated to a partial thromboplastin time (PTT) of 50-70 seconds or an anti-Xa activity of 0.2-0.4 units/mL 5 / Solid .

The challenge is that ECMO patients have simultaneous consumption coagulopathy from circuit-blood contact (platelet consumption, coagulation factor consumption), frequent ongoing hemorrhage from surgical sites or trauma, and the physiologic consequences of the underlying cardiac or respiratory failure. PTT-based monitoring may not accurately reflect heparin activity in patients with low levels of antithrombin III (the cofactor required for heparin activity), which is common in patients with liver congestion from heart failure 4 / Promising .

Heparin-induced thrombocytopenia (HIT) is a specific and devastating complication when it occurs during ECMO. The platelet count decline from HIT overlaps with the expected consumption thrombocytopenia of ECMO, making clinical diagnosis difficult. If HIT is suspected (platelet count below 50,000 or 50% drop from baseline, with thrombotic events), heparin must be replaced with a direct thrombin inhibitor: bivalirudin is the preferred agent in most ECMO centers because its short half-life enables rapid titration 4 / Promising . At Rush University Medical Center in Chicago, the ECMO anticoagulation protocol specifies bivalirudin as the first-line agent for any patient with a platelet count drop of more than 40% within the first 5 days of ECMO, pending HIT antibody and serotonin release assay results.

A.5 VV-ECMO Versus VA-ECMO: Selecting the Right Configuration

The distinction between veno-venous (VV) and veno-arterial (VA) ECMO is fundamental and determines the clinical scenarios in which each is appropriate. VV-ECMO replaces pulmonary gas exchange without providing hemodynamic support; VA-ECMO replaces both pulmonary gas exchange and cardiac output 5 / Solid 61069-2).

VV-ECMO is the correct choice for isolated severe respiratory failure (ARDS, severe COVID-19 pneumonia, influenza pneumonitis) in a patient with preserved or recoverable cardiac function. The CESAR trial enrolled ARDS patients with Murray lung injury scores above 3.0 or pH below 7.20 and showed improved survival with transfer to a specialist ECMO center (63% alive without severe disability versus 47% in conventional care) 5 / Solid 61069-2). The EOLIA trial provided additional evidence in a more protocolized design, showing a 35% reduction in 60-day mortality with VV-ECMO in severe ARDS 5 / Solid .

VA-ECMO is the correct choice when cardiac dysfunction is the primary driver of shock, with or without coexisting respiratory failure. The ECLS-SHOCK trial tested VA-ECMO plus early revascularization versus intra-aortic balloon pump plus revascularization in acute MI complicated by cardiogenic shock, and found no mortality benefit from ECMO (52.4% versus 44.5% 30-day mortality, with increased complications in the ECMO group) 5 / Solid . This finding does not eliminate VA-ECMO from the AMI-cardiogenic shock pathway, but it does challenge the assumption that maximal hemodynamic support automatically translates to improved survival, and it reinforces the primacy of early coronary revascularization over device selection in AMI-shock 5 / Solid .

A.6 ECMO in the Illinois Healthcare Landscape

ECMO capability requires cardiothoracic surgical expertise, perfusionist availability 24 hours a day, and an ICU team experienced in circuit management. These requirements limit ECMO programs to major academic medical centers and a small number of high-volume community hospitals. In Illinois, ECMO programs for adult patients are concentrated at: Northwestern Memorial Hospital, Rush University Medical Center, Loyola University Medical Center, the University of Illinois Hospital, OSF Saint Francis Medical Center in Peoria, and Carle Foundation Hospital in Urbana for select cases 3 / Early .

The geographic consequence is that patients in rural Illinois who develop severe ARDS or refractory cardiogenic shock must be transferred to one of these centers for ECMO consideration. The median transfer time from rural downstate Illinois to Northwestern Memorial or Rush University Medical Center is 2.5-4 hours by ground transport and 60-90 minutes by helicopter. For patients in overt cardiogenic shock, this transfer time is the interval during which end-organ damage (renal tubular necrosis, hepatic necrosis, watershed stroke) accumulates 3 / Early .

The strategy for this access gap involves two interventions. First, early recognition training: emergency physicians and hospitalists in rural Illinois hospitals who are equipped to recognize ECMO-candidacy earlier (before multi-organ failure is established) can initiate transfer while the patient still has physiologic reserve for the journey. Second, spoke-and-hub ECMO transport: Carle Foundation Hospital’s ECMO transport team can cannulate patients at referring hospitals and transport them on ECMO, reducing the hemodynamic deterioration during transfer. This mobile ECMO program is active for select cases and is highlighted in structured remote monitoring physician education module as a resource available to providers in the east-central Illinois corridor.

A.7 ECMO Team Composition and the Human Factor in Circuit Management

ECMO is a team-based intervention in a way that few other cardiac interventions are. The circuit does not run itself: it requires around-the-clock management by a perfusionist (or ECMO specialist, a trained respiratory therapist or registered nurse with advanced perfusion training at programs that have adopted this model), continuous hemodynamic monitoring by the bedside ICU nurse, and regular physician reassessment of circulatory adequacy and weaning trajectory 5 / Solid Guidelines 2021; elso.org).

Perfusionist staffing is the limiting factor in ECMO program volume in Illinois. Registered perfusionists complete a 2-4 year postgraduate clinical perfusion program accredited by the Perfusion Program Accreditation body, and the national supply of perfusionists has not kept pace with the expansion of cardiac surgery and ECMO programs. At Carle Foundation Hospital, ECMO support is provided by a team model that includes a dedicated perfusionist during the first 48 hours of any ECMO run, transitioning to an ECMO-specialist RN model for ongoing circuit management once the patient is hemodynamically stabilized 3 / Early . The ELSO registry data consistently show that center volume above 30 cases per year is associated with lower ECMO mortality, an effect driven in part by team experience and protocol standardization rather than any single technical factor 5 / Solid 62678-5). For patients in rural Illinois who are transferred to Carle Foundation Hospital for ECMO, the transition of care from the referring hospital team to the ECMO team is a structured handoff process managed by the ECMO coordinator, ensuring that no clinical detail about the patient’s arrest circumstances, comorbidities, or prior medications is lost during the transfer.

The handoff protocol at Carle Foundation Hospital includes a structured verbal briefing using the SBAR (Situation-Background-Assessment-Recommendation) framework adapted for ECMO transfer: the referring physician documents the arrest rhythm, duration of low-flow state, CPR quality assessed by end-tidal CO2 readings, initial lactate and creatinine, and the clinical question driving the ECMO referral (bridge to recovery, bridge to LVAD, bridge to transplant, or palliation clarification). This documentation prevents the most common failure mode in ECMO transfers: the receiving team spending the first hour reconstructing the clinical history rather than focusing on hemodynamic optimization 5 / Solid . The program routes physicians in the east-central Illinois region to the Carle Foundation Hospital ECMO consultation line for real-time guidance on ECMO candidacy and transfer logistics before the patient deteriorates beyond transport safety margins.