Gas analysis and capnography,
principle, analyser, waveform.

Oxygen is an elementary, symmetrical molecule with no dipole moment, so it does not absorb infrared radiation. The same is true of nitrogen and the noble gases.

Oxygen is measured instead by paramagnetic analysis, exploiting the two unpaired electrons in its outer shell, or electrochemically by a fuel cell or a polarographic (Clark) electrode.

Gas analysis and capnography →

Because glass absorbs infrared. A glass window would attenuate the beam before it reached the sample and would itself behave as an unknown absorber. Sapphire is transparent at the working wavelength and hard enough to resist scratching in clinical use.

Gas analysis and capnography →

Zeroing exposes the analyser to carbon dioxide-free gas and sets the reference intensity I₀, correcting the offset — it compensates for soiling of the windows and ageing of the source. Calibration exposes it to a mixture of known concentration and sets the gain. An analyser can be correctly zeroed and still read the wrong value, and vice versa.

Gas analysis and capnography →

Collision or pressure broadening. Nitrous oxide molecules colliding with carbon dioxide molecules broaden its absorption peak, so more infrared is absorbed at the analyser's wavelength than the carbon dioxide alone accounts for, and the reading is too high. Nitrous oxide also absorbs close to the carbon dioxide band, adding spectral overlap. Modern analysers correct for this automatically once the gas mixture is entered or measured.

Gas analysis and capnography →

Transit time along the sampling line plus the rise time of the analyser, conventionally 10% to 90% of a step change. In a neonate the respiratory rate is high and the expiratory time short, so if the total delay approaches the duration of expiration the plateau is never reached and end-tidal carbon dioxide is under-read. Shorten and widen the sampling line, reduce the sample flow so fresh gas is not entrained, or use mainstream sampling.

Gas analysis and capnography →

The trade-off runs both ways. Sidestream removes 50–200 mL/min from a circuit whose total minute ventilation may be under 1 L/min, and its transit delay against a short expiratory time means the plateau may never be reached, so end-tidal carbon dioxide is under-read. Mainstream avoids both but adds apparatus dead space and weight at the airway, which matter disproportionately at this size. Low-dead-space neonatal mainstream cuvettes exist and are generally preferred; if sidestream is used, reduce the sample flow and shorten the line.

Gas analysis and capnography →

The trace loses its square outline and becomes a shark fin: phase II is slurred rather than steep, and phase III slopes steadily upward instead of forming a plateau. The α angle is widened and there is often no true plateau at all, so the value read as end-tidal under-reads the true alveolar carbon dioxide.

The cause is uneven emptying. Obstructed units empty late and are richer in carbon dioxide, so gas arriving at the sampling point becomes progressively more concentrated throughout expiration rather than reaching equilibrium. The steeper the upslope, the worse the obstruction — which makes the shape a bedside measure of severity and a way of tracking the response to a bronchodilator.

Gas analysis and capnography →

Normally 2 to 5 mmHg (0.3–0.7 kPa), with PaCO₂ 35–45 mmHg and PETCO₂ 30–40 mmHg.

It exists because alveolar dead space — ventilated but unperfused alveoli — contributes carbon dioxide-free gas that dilutes the alveolar sample.

It widens with hypovolaemia, reduced cardiac output, pulmonary embolism, high airway pressures or excessive positive end-expiratory pressure, chronic lung disease, increasing age and the upright posture.

Gas analysis and capnography →

Only if the arterial to end-tidal gradient is normal. The assumption fails precisely when it matters — in hypovolaemia, low cardiac output, pulmonary embolism, chronic lung disease or with high airway pressures, all of which increase alveolar dead space and widen the gradient. In those patients the arterial value may be considerably higher, and only a blood gas will tell you.

Gas analysis and capnography →

The usual trace is a time capnogram — carbon dioxide against time. A volume capnogram plots carbon dioxide against expired volume, which allows the area under the curve to be used quantitatively: anatomical dead space by Fowler's method, and physiological dead space through the Bohr equation using mixed expired carbon dioxide.

Gas analysis and capnography →

Assume oesophageal intubation until proven otherwise. Call for help, maintain oxygenation, and look: direct or video laryngoscopy to see the tube between the cords.

The differential also includes complete tube obstruction, a disconnected or blocked sampling line, ventilator failure and cardiac arrest — a patient with no cardiac output produces no waveform even through a correctly placed tube. But the safe default is if in doubt, take it out.

Gas analysis and capnography →

Carbon dioxide that entered the stomach during bag-mask ventilation, or from a recent carbonated drink, being washed out over the first few breaths — an oesophageal intubation. Each waveform is small, misshapen and smaller than the last, and the trace then flatlines. This is exactly why confirmation requires a sustained waveform over six breaths rather than the presence of any trace at all.

Gas analysis and capnography →

A fall to a low but non-zero value with a preserved waveform points at the circulation: pulmonary or gas embolism, a sudden fall in cardiac output, severe hypotension or major haemorrhage. A fall to absolute zero with no waveform points at the circuit or the airway: disconnection, extubation, complete obstruction or sampling-line failure.

Look at the patient and the chest first, then the circuit and the tube, then feel for a pulse and check the blood pressure and rhythm. Treat a lost waveform as an emergency until it is proven to be a monitoring fault.

Gas analysis and capnography →

Inspired gas contains carbon dioxide — rebreathing. The causes are an exhausted absorber, an incompetent unidirectional valve, fresh gas flow too low for the breathing system in use, or channelling within the canister. Increase the fresh gas flow immediately, then change the absorbent and check the valves. Consider a self-inflating bag if the circuit is suspect.

Gas analysis and capnography →

A progressive rise that is resistant to increased ventilation indicates increased production rather than reduced elimination, and malignant hyperthermia must be considered — a rising end-tidal carbon dioxide is its earliest and most sensitive sign. Stop the trigger, call for help and dantrolene, and change to a vapour-free breathing system. Other causes of increased production are thyroid storm, phaeochromocytoma, sepsis and prolonged tourniquet release.

Gas analysis and capnography →

Expected: a gradual rise from peritoneal absorption of insufflated carbon dioxide, plateauing after roughly 15–30 minutes and usually correctable by increasing minute ventilation. It is worse with extraperitoneal insufflation and with subcutaneous emphysema.

Worrying: a rise that continues despite increased ventilation, which raises malignant hyperthermia. And the opposite pattern — a sudden fall with hypotension and desaturation — which suggests carbon dioxide embolism. Stop insufflation, release the pneumoperitoneum, give 100% oxygen and place the patient head-down in the left lateral position.

Gas analysis and capnography →

Return of spontaneous circulation. During compressions, end-tidal carbon dioxide is limited by pulmonary blood flow rather than by ventilation, so it reflects the cardiac output compressions generate. A sudden sustained rise without any change in ventilation means pulmonary blood flow has increased. Use it as a prompt to check for a pulse at the next rhythm check rather than interrupting compressions.

Gas analysis and capnography →

Capnography reports gas movement directly and changes on the very next breath. Pulse oximetry reports arterial saturation, which is defended by the oxygen stored in the functional residual capacity and by the flat upper part of the oxyhaemoglobin dissociation curve — so after preoxygenation there may be minutes of delay before the saturation falls, and by then the patient is already hypoxaemic. Capnography warns; oximetry reports the consequence.

Gas analysis and capnography →

Accepting a raised PaCO₂ rather than increasing tidal volume or airway pressure to normalise it. It is tolerated in acute respiratory distress syndrome and severe asthma because lung-protective ventilation matters more than a normal number, and because hypercapnic acidosis appears to be anti-inflammatory. Avoid it where raised intracranial pressure makes cerebral vasodilatation dangerous, and where pulmonary hypertension or right ventricular failure makes a further rise in pulmonary vascular resistance unsafe.

Gas analysis and capnography →