Veno-arterial ECMO (VA ECMO) is the configuration of extracorporeal membrane oxygenation that delivers both circulatory and respiratory support. Blood is drained from the venous system, passed through a membrane oxygenator, and returned under pressure to the arterial circulation. That single change in return cannula location, compared with veno-venous ECMO, is what makes VA ECMO the workhorse for cardiogenic shock, refractory cardiac arrest, and a long list of perioperative cardiac emergencies.
If you are preparing for ECMO specialist certification, perfusionist boards, or critical care fellowship exams, you need to know VA ECMO at a level that goes well beyond the diagram. Cannula sizing decisions, North-South syndrome, distal perfusion catheters, LV venting strategies, and the difference between central and peripheral cannulation all show up in question banks. This guide walks through the physiology, the clinical setup, the pitfalls you will be asked about, and the monitoring targets that separate a stable run from a code situation.
What follows mirrors the way exam writers actually frame VA ECMO content. You will see indication categories, hemodynamic targets with specific numbers, weaning criteria, and the complications that account for the majority of mortality on circuit. Use the components below to anchor the facts, then drill them with the practice tests linked throughout this guide.
Think of VA ECMO as a parallel circuit. The patient's native heart is still beating (most of the time), and the ECMO pump is pulling deoxygenated blood from the right atrium or vena cava and returning oxygenated blood under arterial pressure. The two circuits meet somewhere in the aorta, and the meeting point matters more than most trainees realize.
With peripheral femoral VA cannulation, oxygenated blood flows retrograde up the descending aorta. If native cardiac function is poor, the ECMO flow dominates and the entire aorta is perfused by the circuit. If native function is preserved but the lungs are bad, you can run into the classic North-South syndrome, also called Harlequin syndrome: deoxygenated blood ejected from the left ventricle perfuses the coronaries, brain, and right arm, while ECMO-oxygenated blood perfuses the lower body. Right radial arterial line monitoring is therefore mandatory on femoral VA ECMO.
Central VA ECMO, with the return cannula in the ascending aorta, avoids North-South syndrome entirely because oxygenated blood is delivered above the aortic valve. The trade-off is an open chest or sternotomy, higher bleeding risk, and the need for surgical decannulation. Most institutions reserve central cannulation for patients who cannot be weaned off cardiopulmonary bypass after cardiac surgery.
VV ECMO supports the lungs only and requires a working heart; VA ECMO supports both heart and lungs but adds afterload to the failing left ventricle and carries higher stroke, limb ischemia, and bleeding risk. Memorize that single distinction and most exam stems become trivial to classify.
The Extracorporeal Life Support Organization (ELSO) groups VA indications into a handful of buckets. You should be able to recognize each on a vignette.
Cardiogenic shock remains the headline indication. This includes acute myocardial infarction with shock, fulminant myocarditis, decompensated chronic heart failure as a bridge to transplant or LVAD, and drug-induced cardiomyopathy. The hemodynamic threshold most centers use is a cardiac index under 2.0 L/min/m squared with rising lactate despite two inotropes and either an intra-aortic balloon pump or Impella.
Refractory cardiac arrest, commonly delivered as ECPR (extracorporeal cardiopulmonary resuscitation), is the fastest-growing indication. Selection criteria are tight: witnessed arrest, bystander CPR, initial shockable rhythm, low-flow time under 60 minutes, and no obvious contraindication. Outcomes correlate strongly with cannulation speed.
Postcardiotomy failure covers the patient who cannot wean from cardiopulmonary bypass. Central cannulation is convenient because the chest is already open. Massive pulmonary embolism with hemodynamic collapse is increasingly managed with VA ECMO as a bridge to thrombolysis or thrombectomy. Bridge to decision covers situations where the team is buying time to evaluate neurologic status, transplant candidacy, or reversibility of the underlying insult.
Contraindications you will be tested on include uncontrolled bleeding, severe aortic regurgitation (the LV cannot eject against retrograde flow, distends, and infarcts), aortic dissection involving the cannulation territory, and irreversible neurologic injury. Severe peripheral vascular disease can also preclude femoral cannulation and may push the team toward axillary or central access.
Drainage: femoral vein into RA. Return: femoral artery. Fast to place, percutaneous, no sternotomy. Risks: limb ischemia, North-South syndrome, LV distension. Best for: ECPR, cardiogenic shock outside the OR.
Return cannula in axillary or subclavian artery via graft. Antegrade flow to aortic arch avoids Harlequin. Allows ambulation as bridge to transplant. Risks: hyperperfusion of right arm, graft infection.
Drainage: RA. Return: ascending aorta. Antegrade flow, no limb ischemia, easy LV venting. Requires open chest, higher bleeding, surgical decannulation. Standard for postcardiotomy failure.
Femoral VA with an additional return cannula in the internal jugular vein. Splits oxygenated return between aorta and right atrium to fix North-South physiology while keeping circulatory support.
Cannula selection drives flow. For a 70 kg adult you typically need 4-5 L/min of support, which usually means a 21-25 Fr multistage venous drainage cannula and a 15-19 Fr arterial return cannula. Cannulae are sized by the Poiseuille equation in spirit: bigger and shorter beats smaller and longer for any given pressure gradient. Pre-cannulation ultrasound of the femoral vessels is now standard to confirm patency and minimum vessel diameter (typically the arterial cannula outer diameter should not exceed 75-80% of the artery's inner diameter to allow distal flow).
The Seldinger technique under ultrasound and fluoroscopic guidance is the default percutaneous approach. The drainage cannula tip should sit at the cavoatrial junction or just inside the right atrium; the arterial return tip should sit in the common iliac artery or distal aorta. Confirm position with transesophageal echo whenever possible, particularly the venous tip relative to the tricuspid valve. Misplaced venous tips cause poor drainage, chatter, and inability to reach target flows.
Distal perfusion catheter (DPC) placement is a non-negotiable element of modern peripheral VA cannulation. A 5-7 Fr sheath is inserted antegrade into the superficial femoral artery distal to the arterial cannula and connected back to the arterial limb of the circuit. Without it, limb ischemia rates approach 20-30%. With it, rates drop below 15%, and most centers consider DPC placement a quality metric audited in their ECMO program reviews.
Target 60-80 mL/kg/min in adults (typically 4-6 L/min). Lower flow risks circuit thrombosis; higher flow risks hemolysis. Adjust to achieve adequate end-organ perfusion measured by lactate clearance and urine output.
Mean arterial pressure 60-80 mmHg. Higher MAP increases afterload on the left ventricle and worsens distension; lower MAP risks under-perfusion of brain and kidneys. Titrate vasopressors to stay in band.
Mixed venous oxygen saturation greater than 70 percent indicates adequate oxygen delivery relative to consumption. Sustained values under 60 percent suggest under-pumping, anemia, or shivering/agitation increasing metabolic demand.
Pulse pressure 10-15 mmHg confirms the native LV is still ejecting. A flat arterial line (pulse pressure under 5 mmHg) signals impending LV distension and the need for venting strategy escalation.
Activated clotting time 180-220 seconds with unfractionated heparin, or anti-Xa 0.3-0.7 IU/mL. Bivalirudin is an increasingly common alternative when HIT is suspected or anti-Xa response is erratic.
One of the most heavily tested complications of VA ECMO is left ventricular distension. The mechanism is simple in principle: retrograde flow up the aorta raises afterload, the failing LV cannot eject against it, end-diastolic pressure climbs, the mitral valve fails to close cleanly, and the lungs flood. Echocardiographically you see a dilated, sluggish LV with smoke in the chamber, a stagnant aortic root, and rising pulmonary capillary wedge pressure.
The first sign at the bedside is loss of arterial pulsatility. If pulse pressure on the arterial line is consistently under 10 mmHg, you have a venting problem brewing. Chest X-ray will show worsening pulmonary edema despite adequate flow. Untreated LV distension leads to subendocardial ischemia, pulmonary hemorrhage, and circuit-induced myocardial stunning that prevents weaning.
Venting strategies form a tested hierarchy. Lowest invasiveness first: optimize ECMO flow (sometimes reducing flow improves ejection), add inotrope or chronotrope to recruit residual LV function, add an intra-aortic balloon pump to reduce afterload. Mid-tier: Impella CP or 5.5 as an LV vent (the so-called ECPELLA configuration). Highest invasiveness: percutaneous atrial septostomy or surgical LV vent through the right superior pulmonary vein or LV apex.
VA ECMO has a defined complication profile that you should be able to rank by frequency for exam purposes. Bleeding leads the list. Surgical site bleeding, gastrointestinal bleeding, intracranial hemorrhage, and pulmonary hemorrhage together account for the largest share of morbidity. Anticoagulation targets are deliberately lower than they used to be (ACT 180-220 rather than 220-260) precisely because bleeding outweighs thrombosis in modern circuits with heparin-bonded surfaces.
Thrombotic complications include stroke (3-8 percent in adult cardiogenic shock cohorts), circuit thrombosis, and limb ischemia. Pump head thrombosis is detected by rising P-pre (pre-oxygenator pressure), falling D-dimer trend reversing, visible clot on transillumination, and worsening hemolysis (LDH rising above 2.5 times the upper limit, free hemoglobin over 50 mg/dL).
Infection on ECMO is common and underdiagnosed. Cannula site infection, ventilator-associated pneumonia, and bloodstream infection from circuit colonization all increase mortality. Empiric antibiotic policies vary by institution but surveillance cultures every 48-72 hours are standard.
Neurologic injury includes ischemic stroke from embolic phenomena, hemorrhagic stroke from anticoagulation, and hypoxic-ischemic encephalopathy from pre-cannulation low-flow state. Daily sedation interruption (when safe) and serial neurologic exams are essential. Many centers add EEG monitoring for patients who remain unresponsive after sedation lightening.
Weaning is a structured trial, not a wish. You begin when the underlying insult is reversing: vasoactive requirements falling, lactate normalizing, kidney function improving, and echocardiographic signs of recovery (LVEF over 25 percent, aortic VTI greater than 12 cm, pulse pressure over 25 mmHg on stable flow). Native lung function should also be adequate, because VA ECMO weaning unmasks the native gas exchange capacity.
The classic weaning trial reduces flow stepwise (often 0.5-1.0 L/min decrements) every 10-30 minutes while monitoring MAP, lactate, ScvO2, and echocardiographic indices. At each step the team checks whether the native heart can maintain cardiac output. A successful trial typically demonstrates LVEF over 25 percent, aortic VTI greater than 10 cm, TDI lateral S prime over 6 cm/s, and MAP over 60 mmHg with minimal inotrope at flows of 1.0-1.5 L/min for at least 30-60 minutes.
If the trial passes, decannulation proceeds: percutaneous arterial cannulae often require surgical repair of the femoral artery, venous cannulae can frequently be removed bedside with manual compression. Watch for post-decannulation arterial thrombosis, pseudoaneurysm, and reactive hyperemia in the previously cannulated limb.
Failed weaning trials should prompt reassessment for durable mechanical circulatory support (LVAD), transplantation evaluation, or withdrawal-of-support discussions. The bridge-to-nothing scenario, prolonged VA ECMO without a defined destination, is associated with poor outcomes and is increasingly recognized as a quality issue at high-volume centers.
Pediatric VA ECMO uses scaled cannulae (typically 8-19 Fr arterial, 12-23 Fr venous) and frequently relies on neck cannulation through the right internal jugular vein and right carotid artery. Neonates and infants tolerate carotid ligation surprisingly well due to circle of Willis collateralization, but long-term neurodevelopmental follow-up is essential and any program offering pediatric ECMO should have a structured follow-up clinic.
VA ECMO in pregnancy carries unique considerations: fetal monitoring, careful anticoagulation (heparin does not cross the placenta and is preferred), and team coordination for emergent delivery. Reported maternal survival in case series is similar to non-pregnant cohorts when used for cardiogenic shock or peripartum cardiomyopathy.
Septic shock with refractory myocardial depression is a growing indication. Septic patients on VA ECMO are vasoplegic, often require vasopressors despite ECMO flow, and have high bleeding and thrombotic complication rates. Patient selection is critical: those with isolated septic cardiomyopathy and preserved peripheral resistance do better than those with vasoplegic shock alone, where ECMO does not address the dominant pathophysiology.
Hypothermic cardiac arrest, particularly from avalanche burial or cold water immersion, has historically excellent outcomes with VA ECMO rewarming despite extended down-times: the so-called "nobody is dead until they are warm and dead" maxim. Serum potassium over 12 mmol/L on initial labs is a recognized poor prognostic marker that helps the team decide whether to proceed.
VA ECMO sits at the intersection of cardiac surgery, critical care, and emergency medicine, which is why exam writers love it. A single question can probe physiology (afterload mismatch), procedure (cannula sizing), complications (Harlequin), and management (weaning criteria) all at once. The high-yield learner builds a mental scaffold for each topic, then layers numerical targets onto it.
The most common knowledge gaps on board exams are the specific numbers: flow targets in mL/kg/min, MAP ranges, ACT goals, pulse pressure thresholds, and SvO2 cutoffs. Memorize those first. Next, recognize the clinical patterns: rising lactate despite adequate flow points to under-perfusion or hemolysis; flat pulse pressure points to LV distension; right radial hypoxia points to Harlequin; rising pre-oxygenator pressure points to circuit clot.
Finally, anchor your understanding in the workflows: who gets cannulated where, what monitors are mandatory, how weaning trials are structured, and when the conversation shifts to LVAD or transplant. Once those patterns are second nature, the question stem details fall into place. Use the linked practice tests to drill until the recall feels automatic, that is what the certifying boards are measuring, and it is also what makes you safer at the bedside when a real cardiogenic shock case rolls into resus.
One pattern that catches even strong test-takers is the multi-step troubleshooting vignette. A typical stem reads: "A 56-year-old man on day three of femoral VA ECMO for ischemic cardiomyopathy develops worsening hypoxemia on his right radial ABG (PaO2 55 mmHg) while the post-oxygenator gas remains 350 mmHg.
His pulse pressure has widened to 25 mmHg over the past 12 hours." The expected answer is Harlequin syndrome from recovering native LV function with persistent lung dysfunction, and the management is conversion to VA-V configuration by adding a venous return limb. Recognize the pattern: widening pulse pressure (sign of LV recovery) plus right radial desaturation (sign of upper body perfusion by native circulation) equals North-South.
Another classic pattern is the falling SvO2 with stable flow. Differential includes anemia, sedation lightening or shivering raising oxygen consumption, recirculation (less common in VA than VV), and inadequate cardiac output despite the circuit. The work-up order matters on exams: check hemoglobin first (fast, reversible), then sedation depth and temperature, then echocardiogram for unexpected LV failure. Avoid the trap of cranking flow without diagnosing the cause.
Finally, the bleeding-versus-thrombosis dilemma is a perennial question. The default rule for boards: if the patient is bleeding clinically (chest tube output, GI bleed, falling hemoglobin without obvious source), reduce or hold anticoagulation, accept slightly higher circuit thrombosis risk, transfuse platelets and cryoprecipitate, and reassess. Modern heparin-bonded circuits tolerate brief anticoagulation interruption better than older surfaces. The wrong answer is almost always to escalate heparin in a bleeding patient on ECMO.
Learn more in our guide on ecmo procedure. Learn more in our guide on ECMO: How Extracorporeal Membrane Oxygenation Works and Who Needs It. Learn more in our guide on ecmo machine.