How the ECMO Procedure Works for Heart and Lung Failure

Extracorporeal membrane oxygenation, universally known as ECMO, is a form of life support that takes over some or all of the work your heart and lungs normally do. Blood is drawn out of the body through large tubes called cannulas, pumped through a machine that adds oxygen and removes carbon dioxide, and then returned to the body. It is reserved for the sickest patients in intensive care, people whose hearts, lungs, or both have failed so severely that a ventilator or medications alone cannot keep them alive. The technology has been used in humans since the early 1970s and has evolved considerably since then, but it remains a high-risk intervention with serious complications and no guarantee of survival.

How the Machine Works

An ECMO circuit has a few core components: large-bore cannulas inserted into major blood vessels, flexible tubing, a centrifugal pump that propels the blood, and a membrane oxygenator that functions as an artificial lung. The oxygenator is a bundle of hollow fibers through which oxygen flows; blood passes along the outside of these fibers, and gas exchange happens across the membrane. A heat exchanger keeps the blood at body temperature before it re-enters the patient. The whole setup sits on a cart at the bedside, monitored continuously by specialized nurses and perfusionists.

Two main configurations exist, and the choice between them depends on whether the problem is in the lungs, the heart, or both. Veno-venous (VV) ECMO drains blood from a large vein, oxygenates it, and sends it back into the venous system. The patient’s own heart still does the pumping, so VV-ECMO is suited for isolated lung failure. Veno-arterial (VA) ECMO, by contrast, returns oxygenated blood into an artery, effectively bypassing both the lungs and the heart. That makes it the choice for cardiac failure or combined heart-and-lung failure.

Beyond those two standard setups, hybrid configurations have gained traction. Triple-cannulation strategies such as VV-A (two venous drains plus an arterial return) or VA-V (a venous drain with both arterial and venous return) attempt to address situations where both organs are failing simultaneously. These hybrids carry their own trade-offs. Simulations show that full VA-ECMO support can increase the workload on the left side of the heart by raising the pressure it pumps against, which sometimes worsens the very problem the machine is trying to solve.¹

ECMO for Severe Lung Failure

The most studied use of ECMO is for acute respiratory distress syndrome (ARDS), a condition in which the lungs become so inflamed and fluid-filled that even maximum ventilator settings cannot deliver enough oxygen. An individual-patient-data meta-analysis of the two major randomized trials found that death by day 90 was significantly lower in patients placed on VV-ECMO compared with those managed conventionally, with roughly three-quarters as many deaths in the ECMO group.² A separate pooled analysis of the same trials looking at 60-day mortality reported a similar benefit, but also flagged that ECMO carried a moderate risk of major bleeding.³ These numbers are specific to carefully selected patients in expert centers, and they underscore the central tension of ECMO: it can meaningfully improve survival in the right patient, but it introduces its own serious dangers.

ECMO for Heart Failure and Cardiac Arrest

VA-ECMO has emerged as a rescue option for cardiogenic shock, the state in which the heart is too weak to pump enough blood to sustain organ function. It provides circulatory support and buys time, either for the heart to recover on its own, for medications to take effect, or as a bridge to a more durable solution like a ventricular assist device or heart transplant.⁴ It can even be deployed during or after cardiac arrest that does not respond to standard resuscitation, a use called ECPR (extracorporeal cardiopulmonary resuscitation).

The outcomes for cardiac ECMO are sobering. A large meta-analysis including over 12,000 patients with refractory cardiogenic shock found a pooled in-hospital mortality of about 62%. Older age, the presence of infection, and shorter ECMO support duration were each independently linked to higher mortality. Complications during support were frequent and carried the potential for permanent injury or death.⁵ Those numbers reflect the reality that patients who need cardiac ECMO are already extremely ill, and the machine is buying time in a desperate situation rather than curing the underlying problem.

Who Gets ECMO and How Doctors Decide

Not every critically ill patient is a candidate. ECMO teams weigh the likelihood that the underlying condition is reversible (or that a bridge destination like transplant exists), the patient’s age and baseline health, and whether the expected benefits outweigh the near-certain complications. Several scoring systems have been developed to predict who will survive on ECMO and who will not. Scores like the RESP and PRESERVE were designed for respiratory ECMO patients and use variables such as blood gas values and the duration of mechanical ventilation before ECMO initiation.⁶

Newer models try to improve on these. The PRESET-Score incorporates factors beyond the lungs, including blood pH, blood pressure, lactate levels, platelet counts, and how long the patient was hospitalized before ECMO started. In its development cohort, PRESET outperformed older scores substantially, though its real-world accuracy drops when applied to different hospitals and patient populations.⁷ A recent scoping review noted that while these scoring tools help guide decisions, none are reliable enough on their own to dictate whether a patient should or should not be placed on ECMO.⁸ The decision ultimately rests on multidisciplinary clinical judgment.

What ECMO Does to the Blood

The moment blood leaves the body and enters the ECMO circuit, it encounters conditions that do not exist inside a healthy blood vessel. The centrifugal pump is the biggest culprit. It generates high mechanical shear stress that physically damages red blood cells and activates platelets, with the pump head posing a far greater risk of blood trauma than the tubing or the oxygenator membrane.⁹ Computational studies show this damage scales steeply with pump speed: increasing revolutions roughly 2.5-fold can produce a 50-fold increase in a standard measure of red blood cell destruction.¹⁰

That shear stress also triggers clotting. Platelet activation is highest in the pump head, followed by the cannulas, with long exposure time and high shear rates acting together to drive clot formation.¹¹ Meanwhile, the non-biological surface of the circuit activates the immune system broadly, setting off an inflammatory cascade that, if unchecked, can contribute to organ damage beyond whatever disease landed the patient on ECMO in the first place.¹²

One of the subtler consequences of all this shear stress is acquired von Willebrand syndrome. Von Willebrand factor is a protein that helps blood clot by linking platelets to damaged blood vessel walls. Inside the ECMO circuit, shear forces break apart the large, functional forms of this protein, leaving patients with a clotting defect that resembles a known bleeding disorder. Almost all ECMO patients develop some degree of this within hours of being connected, and it persists for as long as the machine is running.¹³ In one cohort study, about 23% of patients with this syndrome experienced significant bleeding, with blood type O, longer ECMO duration, and veno-venous cannulation all raising the risk.¹⁴

Neurological Risks

Brain injury during ECMO is uncommon in percentage terms but devastating when it occurs. Among over 15,000 patients on VV-ECMO in the Extracorporeal Life Support Organization registry, about 5% suffered some form of acute brain injury, including ischemic stroke (around 1.4%) and hemorrhagic stroke (around 3.1%). In-hospital mortality for those who had a stroke was roughly 70%, compared with 36% overall.¹⁵

The mechanisms differ by stroke type. Ischemic strokes are driven primarily by blood clots forming in the circuit or elsewhere and traveling to the brain, as well as by reduced blood flow to the brain during periods of low cardiac output.¹⁶ Hemorrhagic strokes, or brain bleeds, are linked to the anticoagulation required to prevent the circuit from clotting, combined with platelet dysfunction and low platelet counts. Risk factors for intracranial bleeding include pre-ECMO cardiac arrest, sepsis, kidney failure, and the need for dialysis while on ECMO.¹⁷ The challenge for clinical teams is walking a tightrope: too little anticoagulation and the circuit clots, too much and the patient bleeds.

Limb Complications in VA-ECMO

When a large cannula is placed into the femoral artery for VA-ECMO, it can obstruct blood flow to the leg on that side, potentially causing ischemia that, at its worst, leads to tissue death and amputation. A systematic review and meta-analysis found that placing a smaller secondary tube called a distal perfusion cannula to supply blood downstream of the main cannula reduced the rate of limb ischemia by a substantial margin, from about 25% to about 10%.¹⁸ Whether to place this extra tube at the time of ECMO initiation or wait and add it only if ischemia develops remains a matter of institutional preference. One study found that even without preemptive placement, delayed insertion resolved symptoms in the patients who developed problems, though the overall limb complication rate in that cohort was still about 15%.¹⁹

Infection During ECMO

ECMO patients are exceptionally vulnerable to hospital-acquired infections. The reported rates vary widely, from under 9% to as high as 64%, depending on the population and how infection is defined.²⁰ The most common culprits are ventilator-associated pneumonia and bloodstream infections. Risk factors identified across a meta-analysis of 23 studies and nearly 3,000 patients include immunosuppression, the need for dialysis, red blood cell transfusions, longer time on ECMO, and longer time on mechanical ventilation.²¹ Given that ECMO patients often have multiple invasive lines, open surgical sites, and suppressed immune function from the circuit itself, the high infection burden is not surprising, but it contributes meaningfully to both the length of stay and mortality.

ECMO in Newborns and Children

ECMO was originally developed for neonates, and its longest track record is in newborns with severe breathing problems. One of the classic pediatric indications is congenital diaphragmatic hernia (CDH), a birth defect in which abdominal organs push up into the chest and prevent the lungs from developing properly. Survival to discharge for CDH babies placed on ECMO has historically hovered around 50-60%.²² Interestingly, as centers have refined their selection criteria, ECMO use in CDH has dropped sharply without a change in survival rates, suggesting that earlier guidelines were placing some babies on ECMO who did not need it.

Long-term development is a concern. CDH infants who required ECMO scored lower on cognitive, language, and motor assessments compared to CDH infants who did not need ECMO. Among those who needed a second ECMO run, motor scores were significantly worse than in single-run patients, with longer ventilator dependence and hospital stays.²³ The timing of ECMO initiation and how it interacts with the timing of surgical repair are thought to influence these outcomes, though data on long-term quality of life in pediatric survivors remain scarce.²⁴

ECMO During Pregnancy

Pregnant and recently postpartum patients occasionally develop ARDS severe enough to warrant ECMO, as happened notably during the 2009 H1N1 influenza pandemic. A systematic review of ECMO use in pregnancy found that maternal survival has improved over time, with larger and more recent case series reporting survival above 80%. The main complications mirrored those in non-pregnant adults: bleeding, blood clots, infection, and kidney injury. Fetal outcomes largely depended on how far along the pregnancy was when ECMO started.²⁵

The H1N1 experience in Australia illustrated both the promise and the peril. Of 12 pregnant or postpartum women placed on ECMO, eight were weaned off and survived to discharge. Quality-of-life assessments at 12 to 18 months showed that survivors had physical and mental health scores comparable to the general population. Among the four who died, causes included hemorrhage and overwhelming fungal infection, underscoring how the bleeding risk of ECMO intersects dangerously with the physiological changes of pregnancy.²⁶

Drug Dosing on ECMO

A problem that receives less public attention but creates daily headaches for ICU teams is that the ECMO circuit absorbs medications. The extensive surface area of the tubing and oxygenator membrane can sequester drugs, particularly fat-soluble ones. Sedatives like fentanyl and midazolam are especially affected, which helps explain why ECMO patients often require much higher doses than expected to stay comfortable.²⁷ Antibiotics, antifungals, and anticonvulsants can also be affected, meaning that standard doses might not achieve adequate blood levels. Clinicians often rely on frequent blood-level monitoring and dose adjustments to compensate, but the pharmacokinetics during ECMO remain an active and somewhat under-researched area.

Cannulation and the Role of Ultrasound

Getting the cannulas into the right place is one of the most critical steps. These are not ordinary IV lines; ECMO cannulas are thick tubes that sit in major vessels and need to be positioned precisely to drain and return blood efficiently. Placement can be surgical (a direct cut-down to expose the vessel) or percutaneous (using a needle-and-wire technique through the skin). Real-time ultrasound guidance is now recommended to reduce complications during insertion and to confirm that the cannula tip ends up in the optimal position.²⁸ Malpositioning can cause poor flow, vessel injury, or cardiac perforation, so imaging is not optional but integral to the procedure.

Mobile ECMO Teams

Most hospitals do not have the infrastructure or expertise to run an ECMO program. When a patient at a community hospital deteriorates to the point of needing ECMO, the standard approach is for a specialized retrieval team to travel to the patient rather than try to transport an unstable patient on a ventilator alone. These mobile ECMO teams typically include cardiac surgeons, a perfusionist, a nurse, and a driver, along with a self-contained ambulance carrying the ECMO machine, a portable ventilator, oxygen supply, surgical instruments, and continuous monitoring equipment.²⁹ The team cannulates the patient at the referring hospital, initiates ECMO on-site, and then transports the now-stabilized patient back to the ECMO center. This process demands expertise not only in ECMO physiology but also in the unique hazards of transporting critically ill patients.³⁰

Life After ECMO

Surviving ECMO is only the beginning. Studies of survivors consistently show that while many regain functional independence, a substantial portion carry lasting effects. Common long-term problems include reduced physical capacity, depression, anxiety, cognitive difficulties, and trouble returning to work or resuming social roles.³¹ One survey found that among VV-ECMO survivors, about three-quarters reported difficulty with at least one daily activity, and over half screened positive for post-traumatic stress disorder. VA-ECMO survivors fared somewhat better on those particular measures, but still about a quarter had high PTSD scores.³²

These outcomes have pushed centers to start rehabilitation earlier. Evidence now supports that physical therapy and early mobilization are safe and feasible even while a patient is still on ECMO. Getting patients sitting, standing, and even walking while connected to the circuit has been shown to reduce ICU-acquired weakness, delirium, and time on mechanical ventilation, while improving functional ability and quality of life after discharge.³³ The image of an ECMO patient walking the hospital hallway with a nurse and a perfusionist pushing the cart behind them is increasingly common in experienced centers.

Why the Hospital You Go to Matters

ECMO is resource-intensive and technically demanding, and not all programs are created equal. A contemporary analysis of ECMO outcomes across the United States found that high-volume hospitals had lower odds of in-hospital death after adjusting for patient risk factors. However, patients at those same high-volume centers spent more days in the hospital and incurred higher costs, likely reflecting the fact that they were kept alive through complications that would have been fatal at lower-volume centers.³⁴ The cost of ECMO programs can be reduced substantially through operational changes. One institution cut its annual ECMO expenditure by 61% after transitioning to longer-lasting circuit technology and reducing the need for on-call perfusionists, without any change in safety outcomes.³⁵

The Ethical Weight of ECMO

Because ECMO can maintain a patient’s blood oxygen levels and circulation almost indefinitely, it creates situations that older generations of physicians never faced. A patient whose lungs have been completely destroyed can look pink, warm, and alert while connected to the machine, obscuring how gravely ill they truly are. When recovery is no longer possible and no transplant option exists, the decision to withdraw ECMO support becomes agonizing for families and clinical teams alike.³⁶ ECMO is often described as a “bridge” therapy, a bridge to recovery, to transplant, or to a more permanent device. But when none of those destinations are reachable, it becomes what ethicists have called a bridge to nowhere. The degree of apparent wellness that ECMO can sustain makes end-of-life conversations harder than they might be for a patient visibly declining on a ventilator, and centers are increasingly recognizing the need for palliative care involvement early in the ECMO course rather than only after all options have been exhausted.