VV ECMO — short for veno-venous extracorporeal membrane oxygenation — pulls deoxygenated blood from a large vein, runs it through an oxygenator and pump, then returns the oxygenated blood back into the venous system. The native heart still does all the cardiac work. The lungs get a chance to rest.
That single sentence carries the whole concept. Drainage is venous. Return is venous. There is no arterial cannula, no retrograde aortic flow, and no direct hemodynamic support. If the heart fails on VV ECMO, you do not have a backup pump — you have a patient. This is the cleanest mental separation from VA ECMO, which adds arterial return and gives the circulation a parallel highway when the heart cannot.
Most adult VV ECMO runs are for severe, reversible hypoxemic or hypercarbic respiratory failure. Think ARDS that is no longer responding to lung-protective ventilation, severe pneumonia (COVID-19 made this familiar to nearly every ICU on the planet), status asthmaticus that cannot be ventilated, or a patient awaiting lung transplant whose oxygenation is collapsing despite every escalation.
You will see VV ECMO described as a bridge. Bridge to recovery is the most common use. Bridge to transplant runs longer — sometimes weeks. Bridge to decision is the polite phrase for "we need 48 hours to figure out whether this is salvageable." None of these are casual decisions. ELSO registry data still shows in-hospital mortality around 40% for adult respiratory VV ECMO, and that number climbs sharply for older patients, immunocompromised hosts, and runs that begin late.
The indication list is short but the judgement is long. Adult VV ECMO is typically considered when a patient has severe, potentially reversible respiratory failure and conventional therapy is failing — usually after optimal ventilator settings, prone positioning, neuromuscular blockade, and inhaled pulmonary vasodilators have been tried.
Classic triggers:
The EOLIA trial and its post-hoc Bayesian re-analyses moved the field toward earlier referral rather than later. Centers using protocolized VV ECMO for severe ARDS report better outcomes than ad-hoc programs. The Extracorporeal Life Support Organization publishes consensus guidelines that most modern ECMO centers follow closely.
What disqualifies a patient? Multi-organ failure that is not lung-driven, age and frailty that make meaningful recovery unlikely, advanced malignancy with poor prognosis, intracranial hemorrhage, and uncontrolled bleeding all push the risk-benefit calculation against cannulation. None of these are absolute, but every one of them lowers the chance the run ends in extubation rather than withdrawal.
VV ECMO = lungs only. Venous drain, venous return, native heart still pumping. VA ECMO = lungs and heart. Venous drain, arterial return, parallel circulation. If the heart fails on VV, you do not have ECMO support for it — you have a developing emergency. Recognizing this distinction is the most commonly tested concept on certified ECMO specialist exams.
VV ECMO can be assembled in three practical configurations. Each carries different recirculation behavior, different mobility tradeoffs, and different anatomical risks. The choice is rarely arbitrary — it is driven by what the patient needs the run to look like over the next two weeks.
Femoral-femoral (bi-femoral) is the workhorse for emergency adult cannulation. A multi-stage drainage cannula sits in the inferior vena cava via one femoral vein, and a return cannula in the other femoral vein delivers oxygenated blood with its tip advanced near the cavo-atrial junction. It is fast, familiar, and bed-side feasible. The cost is recirculation if cannulas overlap and immobility of the patient.
Femoro-jugular places drainage in the femoral vein and return in the internal jugular. This separates drainage and return geographically, which usually drops recirculation below 10%. It permits more upper-body mobility for awake ECMO, and it is the configuration most centers prefer for runs expected to last more than a few days.
Dual-lumen single cannula (Avalon Elite or ProtekDuo) is inserted through the right internal jugular vein and crosses the right atrium so that drainage occurs from the superior and inferior cava while return is directed through the tricuspid valve. Done well, this allows a fully ambulatory patient on ECMO — important for transplant bridging. Done poorly, malposition causes catastrophic recirculation or right-ventricle injury. Echocardiographic or fluoroscopic guidance is mandatory at insertion. Detailed ECMO cannulation technique varies by center but the anatomy does not.
Fastest emergency option. Drain in IVC via one femoral vein, return in the other femoral vein.
Workhorse for runs expected to last more than a few days. Drain from femoral, return via right IJ.
Single 27–31 Fr cannula in right IJ crossing the right atrium. Drainage from SVC+IVC, return toward tricuspid.
On VV ECMO, arterial oxygen content is determined by mixing. The ECMO circuit returns highly oxygenated blood into the venous system. That blood mixes with whatever the native cardiac output is doing and with whatever residual venous return is escaping the drainage cannula. The mixed blood goes through the lungs — which may add nothing — and then to the systemic circulation.
Two numbers drive most bedside thinking. The first is the ratio of ECMO flow to cardiac output. When pump flow approaches or exceeds cardiac output, you maximize the fraction of total venous return that the circuit can oxygenate. Below about 60% of cardiac output, oxygenation stalls because too much native deoxygenated blood bypasses the circuit. This is why a hyperdynamic septic patient with cardiac output of 10 L/min can be very hard to oxygenate on 4 L/min of ECMO flow.
The second number is recirculation — the fraction of oxygenated return blood that re-enters the drainage cannula without ever reaching the body. High recirculation means the circuit is oxygenating its own blood, the patient is starving, and pre- and post-oxygenator saturations look almost identical. Repositioning a cannula by 2 cm under ultrasound can drop recirculation from 50% to under 10%.
This is also why VV ECMO does not need a target SaO2 of 100%. Most centers accept 80–88% saturation if tissue perfusion (lactate, mixed venous saturation, capillary refill) is adequate. The lung-protective payoff of a low FiO2 and a low driving pressure is worth more than chasing a normal SpO2.
Arterial SaO2 target: 85–92%. Some centers accept 80–85% if perfusion is good. Driven by ECMO flow as a fraction of cardiac output, sweep gas FiO2, and recirculation.
Pre-membrane SvO2 should be 65–75%. Pre/post oxygenator saturation gap <5% suggests recirculation. Lactate trend matters more than the single SpO2 number.
Sweep gas (gas flow across the membrane) controls CO2. Increase sweep to drop PaCO2. The relationship is almost linear over the working range.
Avoid rapid pH swings — a brisk PaCO2 drop in a chronically hypercarbic patient can trigger intracranial hemorrhage. Adjust sweep by 0.5–1 L/min, not 5.
Pump flow drives oxygenation. Sweep gas drives CO2. Two separate dials.
Adult flow targets: 3–6 L/min, generally 60–80 mL/kg/min. Increase flow if SaO2 falls; increase sweep if PaCO2 rises. RPM is adjusted only as needed to achieve the flow target while keeping negative drainage pressure above −80 to −100 mmHg.
Rest settings. Once on ECMO, drop to lung-protective "rest" ventilation: tidal volume 3–4 mL/kg, plateau pressure <25 cmH2O, PEEP 10–14, respiratory rate 10, FiO2 0.30–0.40.
The goal is to give the lungs time to heal without inflicting ventilator-induced injury. Daily spontaneous breathing assessment, awake ECMO if tolerated, early tracheostomy if run will exceed 10–14 days.
Unfractionated heparin is the default, targeted to aPTT 50–70 or anti-Xa 0.3–0.7 IU/mL. Some centers use bivalirudin for HIT or unstable patients.
Daily anti-Xa, ACT q4–6h, platelet trend (HIT screen if drop >50%), fibrinogen, D-dimer for circuit clot suspicion. Tolerate sub-therapeutic aPTT if bleeding is active.
Two scoring tools dominate VV ECMO candidacy discussions. They do not replace clinical judgement, but they structure it.
The Murray Lung Injury Score is calculated before cannulation. It uses PaO2/FiO2 ratio, chest radiograph quadrants involved, PEEP, and lung compliance. A Murray score above 3.0 — combined with failure of conventional therapy — places a patient firmly in ECMO-consideration territory. Below 2.5, the evidence for ECMO over optimal mechanical ventilation thins out quickly.
The RESP Score (Respiratory ECMO Survival Prediction) is calculated at cannulation to estimate post-ECMO survival. It uses 12 pre-cannulation variables — age, immunocompromised status, mechanical ventilation duration before ECMO, acute respiratory diagnosis group, central nervous system dysfunction, acute associated infection, neuromuscular blockade, nitric oxide use, bicarbonate, cardiac arrest, PaCO2, and peak inspiratory pressure. The score sorts patients into five risk classes, with class I survival around 92% and class V survival around 18%.
The PRESERVE score performs similarly. The PRESET score is COVID-specific. The point is not to pick a favorite — it is to communicate. A team that documents "RESP score 4, predicted survival 57%" is making a very different conversation possible with a family than "we want to try ECMO." Honest probabilities help the room more than optimism.
Background reading on the broader physiology and indications lives in our ECMO overview.
The daily rhythm of a VV ECMO patient settles into a pattern after the first 24–48 hours. Hourly cannula site checks for bleeding and migration. Pump flow, sweep, RPM, and pre/post-membrane pressures documented at least q1h. Drainage pressure trends watched like an EKG — a sudden swing to more negative values means the patient is volume-down or the cannula is compromised.
Sedation is kept light when possible. The awake ECMO strategy — extubating the patient on ECMO and letting them participate in physical therapy — improves long-term outcomes and shortens runs in selected candidates. It also requires a team that is comfortable with a patient breathing spontaneously on full ECMO support, because the first time a patient asks for water is unsettling if your last ECMO patient was paralyzed and prone.
Fluid balance matters more than most teams initially appreciate. Diuresis to dry weight or below consistently improves oxygenation by reducing pulmonary edema, and the ECMO circuit tolerates lower preload than the native heart would. Daily renal replacement (CRRT) is added in many runs, often through an integrated circuit access port to avoid an additional line.
Nutrition follows standard ICU protocol. Enteral feeding works on ECMO. Bowel rest is not routine. Infection surveillance is aggressive because cannulas are large foreign bodies in critically ill patients — line cultures if temperature spikes, broad-spectrum antibiotics if circuit-associated infection is suspected. Detailed ECMO machine behavior and circuit alarms have their own rhythm too.
The ECMO circuit is a thrombogenic foreign body. Without anticoagulation, clot forms on the oxygenator membrane, in the pump head, and at cannula tips — sometimes within hours. With too much anticoagulation, the patient bleeds — from cannulation sites, the airway, the GI tract, and most worryingly inside the skull.
Unfractionated heparin is the default agent in most adult VV ECMO programs. It is titrated to either aPTT (typical target 50–70 seconds) or anti-Xa (typical target 0.3–0.7 IU/mL). Anti-Xa is increasingly preferred because aPTT can be confounded by coagulopathy of critical illness. ACT (activated clotting time) is monitored at the bedside every 4–6 hours but is not the primary target.
Bivalirudin, a direct thrombin inhibitor, is gaining ground for several reasons: it does not depend on antithrombin, it has a more predictable dose-response, and it bypasses the catastrophe of HIT. Centers using bivalirudin titrate to aPTT 1.5–2.5x baseline or to a specific protocol concentration. Half-life is short (25 minutes), but the drug accumulates in renal failure.
The hardest decisions arrive when the patient is bleeding. A run can continue with sub-therapeutic anticoagulation — sometimes briefly off heparin entirely — if circuit flows are high enough to limit stasis and the team accepts the rising risk of circuit thrombosis. Modern heparin-bonded circuits make this more tolerable than it was a decade ago. Daily reassessment of bleeding vs. clotting risk is non-negotiable. Heparin-induced thrombocytopenia must be screened for any time platelets fall >50% from baseline.
Weaning a VV ECMO patient is conceptually simple. You progressively turn off the sweep gas — the gas flow across the oxygenator — while leaving blood flow unchanged. If the native lungs can carry the oxygenation and CO2 removal load on tolerable ventilator settings, the patient is ready for decannulation. If they cannot, the sweep is restored and weaning is paused.
Typical readiness criteria before a sweep trial: PaO2/FiO2 >150 with the ventilator at FiO2 ≤0.5, PEEP ≤10, plateau pressure <28, and a reasonable spontaneous tidal volume. Lung compliance should be trending up. Chest imaging should show resolution rather than progressive consolidation.
The sweep-off trial: drop the sweep gas to 0 L/min while keeping pump flow constant for 30–60 minutes. Watch the arterial blood gas, ventilator mechanics, work of breathing, and the patient's subjective comfort. If gas exchange is preserved without escalation of ventilator support, you are watching native lung function carry the patient. Document the trial. Repeat at 6–24 hours. If sustainable, plan decannulation.
Decannulation itself is usually bedside for femoral cannulas — pull, hold pressure, suture. Jugular and dual-lumen cannulas may require operative removal. Some centers heparinize down to anti-Xa 0.1–0.3 in the hours before decannulation to lower bleeding while still discouraging fresh thrombosis. Post-decannulation, Doppler ultrasound at 24–48 hours screens for deep vein thrombosis at the cannulation site — incidence is substantial and often clinically silent.
ECPR — extracorporeal cardiopulmonary resuscitation — is technically VA ECMO, not VV, but it overlaps with ECMO programs and deserves a mention. ECPR is the rapid initiation of ECMO during ongoing CPR for refractory cardiac arrest, typically when arrest is witnessed, the rhythm is shockable, time-to-flow is short, and the underlying cause is potentially reversible. Survival to neurologically intact discharge is 20–30% in selected centers — far better than conventional CPR for refractory arrest but only with strict patient selection and a system that can establish ECMO within 60 minutes of arrest.
The ELSO Registry — the Extracorporeal Life Support Organization's international database — is the single most useful outcome resource in the field. As of the most recent ELSO report, adult respiratory ECMO survival to discharge sits around 58–62% across reporting centers. The number masks wide variation: high-volume centers (≥30 runs per year) consistently outperform low-volume programs. The same registry has tracked VV ECMO through H1N1, MERS, and COVID-19, providing real-time outcomes data during each.
Long-term outcomes for VV ECMO survivors are sobering. Most have meaningful neurocognitive recovery. Pulmonary function returns toward baseline over 3–12 months in patients who started with previously healthy lungs. Physical reconditioning is slow. PTSD, depression, and post-intensive-care syndrome are common. The decisions made on rounds during day three of a run reach further into a patient's life than they look at the time. The detailed ECMO procedure story does not end at decannulation — rehab and follow-up are part of it.
VV ECMO is the modality you reach for when the lungs have failed and the heart is still working. The physiology rewards humility — recirculation, cardiac output, and mixed venous saturation matter more than any single arterial gas. The cannulation choice you make on day one defines what the patient's day fourteen looks like. The hardest skill is not knowing when to cannulate but knowing when to wean and when to withdraw, and the team that gets honest about those decisions earlier delivers better care.
If you are studying for a certified ECMO specialist credential, the high-yield concepts cluster around three areas: the difference between VV and VA support, recognizing and managing recirculation, and titrating sweep gas vs. blood flow correctly. Daily rounding pearls — pre/post-membrane saturation gap, anticoagulation targets, oxygenator failure signs, hemodynamic interactions — make up most of the case-based portion of exams. Sharpen these against our ECMO physiology questions and the weaning and decannulation set.
If you are at the bedside, the basics outperform the exotic interventions. Optimize ventilator rest settings. Keep cannulas where you put them. Watch the pre/post saturations and the oxygenator pressure gradient. Reassess every day. Talk to families honestly. Most of what makes VV ECMO work is not glamorous — it is the patient discipline of a team that knows what each number means and what it does not.