ECMO for Respiratory Support
On this page
Direct answer
When the lung fails as an organ rather than as a dial to turn, venovenous ECMO drains venous blood, pumps it through a membrane oxygenator and returns it to the right atrium, letting the team rest the lung on minimal tidal volumes while the membrane does oxygenation and carbon dioxide removal. Referral thresholds in severe ARDS follow the EOLIA pattern: PaO2/FiO2 below 50 for more than 3 hours, below 80 for more than 6 hours, or pH under 7.25 with PaCO2 of at least 60 mmHg for over 3 hours, all despite optimised ventilation, proning and paralysis. Blood flow (roughly 3-6 L/min in adults) governs oxygenation support, sweep-gas flow governs CO2 removal, anticoagulation keeps the circuit from clotting, and bleeding is the countervailing price.
What you must remember
- Configuration logic: venovenous for isolated lung failure (gas exchange, native haemodynamics); venoarterial for cardiac support or combined failure (unloads the heart but risks differential hypoxaemia).
- EOLIA-style entry criteria (per centre protocol): PaO2/FiO2 below 50 mmHg for over 3 hours, below 80 for over 6 hours, or pH below 7.25 with PaCO2 at least 60 for over 3 hours, despite optimised ventilation, proning and blockade; beyond about 7 days of high settings weighs against futility.
- Two independent knobs: blood flow rate (commonly 3-6 L/min in adults) drives oxygen delivery; sweep-gas flow (typically 100% oxygen on the oxygenator) drives CO2 clearance — a rising PaCO2 is a sweep problem, a falling saturation a flow, cannula or membrane problem.
- Lung rest on ECMO: continued low-stretch ventilation — commonly tidal volumes around 4-6 mL/kg predicted weight (or pressure-limited), low rate, PEEP per unit protocol — the lung is immobilised, not abandoned.
- Anticoagulation: unfractionated heparin titrated to APTT roughly 1.5-2 times control per protocol, with platelets, fibrinogen and haemolysis markers monitored daily.
- Circuit surveillance: oxygenator clotting, rising transmembrane pressure drop, cannula drainage problems (chattering, flow swings), and recirculation in venovenous configs (oxygenated blood drawn straight back into the drainage cannula).
- Complications: bleeding (the commonest), circuit component failure, haemolysis, air embolism, infection, limb ischaemia in femoral VA, and neurological events.
- Weaning: improve native-lung settings stepwise; trial off sweep gas; decannulation when the lung sustains protective ventilation unaided.
The therapist's day around one circuit
A 30-year-old with influenza ARDS reached PaO2/FiO2 55 on FiO2 1.0, proned and paralysed, and was cannulated femoro-jugular for venovenous support. The morning round is a fixed sequence: flow 4.2 L/min, sweep 6 L/min, ventilator set to lung rest (rate 8, pressure-limited, PEEP 10), anticoagulation within target, plasma-free haemoglobin not rising, no ooze at the cannulation sites, and the drainage cannula not chattering when the patient is repositioned. A sudden desaturation with the circuit running triggers its own drill — check flow and cannula position first (kinking, migration), then membrane function (pre- and post-oxygenator blood gases), then the native lung (pneumothorax, plugging) — because the saturation alarm does not say which of the three organs (native lung, circuit, cannula) has failed. As compliance improves over ten days, sweep falls to zero, ventilator settings climb, and a two-hour trial off sweep decides decannulation.
Where students slip
The configuration question is asked first and answered worst: venovenous supports gas exchange with the patient's own circulation; venoarterial adds cardiac support and changes everything about monitoring (arterial pulse, differential cyanosis in respiratory failure with femoral VA). The second slip is attributing every derangement to "the ECMO" — the native lung can still develop a pneumothorax or a mucous plug while the circuit runs, and the therapist who keeps assessing breath sounds and secretion load earns the role. Third, quoting EOLIA as "ECMO reduces mortality": the trial stopped early, missed its primary endpoint, and left a 28% cross-over debate — the exam-safe phrasing is "considered for refractory severe ARDS per centre protocols, supported by meta-analysis and observational data". The Indian framing is candid: ECMO remains a scarce, high-cost, referred therapy concentrated in a handful of centres, so the realistic exam task is recognition, referral criteria and transport thinking.
Frequently asked questions
When should venovenous ECMO be considered in severe ARDS?
At EOLIA-type thresholds — PaO2/FiO2 below 50 for over 3 hours, below 80 for over 6 hours, or pH below 7.25 with PaCO2 at least 60 mmHg — despite optimised ventilation, proning and blockade.
What is the difference between venovenous and venoarterial ECMO?
Venovenous supports gas exchange alone with native circulation; venoarterial also supports or replaces cardiac output, with cannulation and monitoring implications including differential hypoxaemia.
What controls oxygenation versus carbon dioxide removal on ECMO?
Blood flow rate primarily supports oxygen delivery; sweep-gas flow across the oxygenator controls CO2 clearance — a rising PaCO2 is addressed on the sweep, not the pump.
How is the lung ventilated while on ECMO?
With lung-rest settings — very low tidal volumes (about 4-6 mL/kg predicted weight) or pressure limitation, low rate, and PEEP per protocol.
Name four complications of ECMO.
Bleeding (commonest), oxygenator or circuit clotting, haemolysis and infection — plus air embolism and limb ischaemia with femoral arterial cannulation.