Oxygenator
On this page
Direct answer
The oxygenator is the artificial lung: blood films over the outside of microporous hollow fibres while oxygen-rich sweep gas flows through them, and gas crosses the membrane down partial-pressure gradients. Two independent dials govern it — FiO2 sets arterial oxygenation, sweep gas flow sets carbon dioxide clearance — and every oxygenator question in the viva separates candidates on exactly that distinction. In modern integrated devices the oxygenator bundles the heat exchanger (placed upstream of the fibres, because warming or cooling blood before the membrane reduces plasma leakage), and its health is judged not by a single reading but by a trend: arterial gases at fixed settings, the transmembrane pressure gradient, sweep requirements, and the gas outlet port.
What you must remember
- Evolution: bubble oxygenators (direct blood-gas contact, protein denaturation, defoaming needed) are obsolete; microporous hollow-fibre membrane oxygenators are the standard; true non-microporous silicone diffusion membranes serve long runs such as ECMO.
- Hollow-fibre design: blood flows outside the fibres (extraluminal) for low resistance and small prime volumes, typically 150-300 mL for the integrated device.
- Oxygenation is governed by FiO2 (and flow delivery); CO2 removal is governed by sweep gas flow rate — raise FiO2 for a low PaO2, raise sweep for a high PaCO2, never the reverse.
- Initial sweep roughly equals blood flow (1:1); FiO2 starts high and weans on the first in-line or arterial-line gas.
- The integrated heat exchanger sits before the fibre bundle: temperature swings enlarge micropores, so tempering blood first reduces plasma leak; water bath temperature never exceeds about 42 degrees C (haemolysis risk).
- Performance surveillance: arterial gas every 30 minutes at fixed settings, in-line venous saturation above 60-65%, the pressure gradient across the oxygenator (typically well under 50 mmHg at baseline — a doubling from baseline is the warning), inspection for fibrin strands and clot.
- Oxygenator failure signs: falling PaO2 at unchanged FiO2, escalating sweep needed for the same CO2, a climbing transmembrane gradient, visible clot, plasma appearing at the gas outlet, and falling platelets over hours.
- Prolonged runs (commonly quoted beyond about six hours for microporous devices) raise plasma leak risk — plan the change-out window in long cases.
Fifty minutes in, the numbers move
An hour into a mitral replacement, the in-line venous saturation reads 58%, the arterial-line gas shows PaO2 140 mmHg at FiO2 1.0 where it was 320, sweep has crept from 4 to 6 litres for the same CO2, and the transmembrane gradient has risen from 40 to 110 mmHg. Pink-tinged fluid wets the gas outlet port. Each finding alone has innocent explanations; together they describe fibrin deposition in a failing oxygenator — in this case with suboptimal anticoagulation.
The sequence is rehearsed, not improvised: notify the surgeon and anaesthetist; confirm the ACT and correct heparinisation; optimise flow, haematocrit and temperature. If gases still fail, the decision is a change-out — a controlled pause, the circuit cut into with a prepared bridge or replacement segment, the new oxygenator primed and de-aired, then resumption with full checks. The drill is memorised in advance because the window between "numbers drifting" and "cannot oxygenate" may be minutes.
Where students slip
The two-dial confusion dominates: raising sweep gas for a hypoxaemic patient does nothing for PaO2, and raising FiO2 for hypercarbia does nothing for PaCO2 — examiners ask this question precisely because the error is common on the pump. Second, plasma leak gets mistaken for harmless condensation: water vapour at the gas outlet is normal; plasma — proteinaceous, often pink, flowing rather than beading — means the membrane is failing. Third, the in-line venous saturation gets misread as an oxygenator reading: a low SvO2 with perfect arterial gases is a flow or demand problem (raise flow, transfuse, cool, deepen anaesthesia), not a membrane problem — the oxygenator can only be judged by what it does to the blood passing through it. Candidates who separate delivery from transfer, in one sentence each, close the topic cleanly.
Frequently asked questions
What controls carbon dioxide removal on bypass?
Sweep gas flow rate across the membrane; oxygenation is controlled separately by FiO2.
Which oxygenator type is used for long-term support such as ECMO?
True diffusion-membrane (silicone or similar non-microporous) oxygenators, which resist the plasma leakage that limits microporous devices.
What is a normal transmembrane pressure gradient and what change alarms you?
Baseline is commonly under 50 mmHg at full flow; a doubling from the device's own baseline signals clotting within the bundle.
Why does the heat exchanger sit upstream of the fibres?
Tempering the blood before the membrane avoids temperature-driven enlargement of the micropores, reducing plasma leak.
What is plasma leak and how is it recognised?
Plasma crossing enlarged micropores, appearing as proteinaceous (often pink) fluid at the gas outlet — distinct from normal condensation.
What does a falling in-line venous saturation indicate?
Oxygen delivery is not meeting consumption — a pump flow, haemoglobin or demand problem, not proof of oxygenator failure.