Respiratory physiology,
from the alveolus to the tissue.

Because tension and content are different quantities. PaO₂ is generated by the oxygen dissolved in plasma, which follows Henry’s law and depends on the alveolar pressure and the state of the lung — neither of which anaemia alters. Saturation is equally normal, because it reports the percentage of the binding sites that are occupied, not how many sites exist.

Content is what falls. CaO₂ = (1.34 × Hb × SaO₂) + (0.0031 × PaO₂), and only the haemoglobin term is affected. At Hb 6 g/dL with SaO₂ 100% and PaO₂ 100 mmHg, CaO₂ is about 8.4 mL/dL against a normal 20 mL/dL — less than half, with an entirely reassuring blood gas and a perfect saturation.

The clinical point: a pulse oximeter cannot detect anaemia, and neither can a PaO₂. The same trap operates in carbon monoxide poisoning and methaemoglobinaemia.

Oxygen transport →

Because of the shape of the oxyhaemoglobin dissociation curve. Blood leaving a high-V/Q unit already sits on the flat plateau: raising its PO₂ from 100 to 130 mmHg adds almost no oxygen content, because the haemoglobin is already close to fully saturated and only the tiny dissolved fraction increases.

Blood from a low-V/Q unit sits on the steep part, where content is genuinely reduced. When the two mix, the surplus tension from the good units cannot make good the content deficit from the bad ones, so the pooled content — and therefore PaO₂ — falls.

Carbon dioxide behaves differently because its dissociation curve is steep and near-linear over the physiological range, so over-ventilated units really can eliminate proportionally more. Add the chemoreceptor response to any rise in PaCO₂ and the characteristic picture is hypoxaemia with a normal or low PaCO₂.

Oxygen transport →

They are two halves of one reciprocal arrangement. Keep them apart by asking which gas is being affected.

Bohr effect — carbon dioxide and H⁺ acting on oxygen carriage. A rise in CO₂ or H⁺ reduces haemoglobin’s affinity for oxygen and shifts the curve right. Protonation of histidine residues stabilises the T-state salt bridges, and carbamino formation does the same. P50 rises about 2–3 mmHg for each 0.1 fall in pH. It operates in the tissues, where local acidosis promotes oxygen release exactly where metabolism is greatest.

Haldane effect — oxygen acting on carbon dioxide carriage. Deoxygenated haemoglobin carries more carbon dioxide, because it is a better proton buffer and forms carbamino compounds more readily. It accounts for roughly a quarter to a third of arteriovenous CO₂ transport. It operates at both ends: in the tissues, releasing oxygen enhances CO₂ uptake; in the lung, taking up oxygen drives CO₂ off.

The two are complementary — whichever gas is being loaded assists the unloading of the other.

Oxygen transport →

A single transfer of carbon dioxide across the placenta shifts both dissociation curves in the helpful direction at the same time. Carbon dioxide passes from fetus to mother, so maternal blood becomes more acidic and its curve shifts right, releasing oxygen; simultaneously the fetal blood loses carbon dioxide, becomes more alkaline, and its curve shifts further left, taking oxygen up. Hence double.

The fetus needs it because the umbilical venous PO₂ is only about 30–35 mmHg — a value that would be alarming in an adult. Three other factors work alongside it: HbF (α₂γ₂), whose γ chains bind 2,3-DPG poorly and so give a P50 of about 19 mmHg; a fetal haemoglobin concentration around 17 g/dL; and a high umbilical blood flow.

The mirror image is the double Haldane effect for carbon dioxide: fetal haemoglobin becoming oxygenated releases CO₂, while maternal haemoglobin becoming deoxygenated takes it up.

Oxygen transport →

Normally it is not. A red cell spends about 0.75 s in the pulmonary capillary and equilibrates with alveolar gas in roughly 0.25 s — a threefold reserve — so oxygen transfer is perfusion-limited: more oxygen requires more blood flow, not more time.

It becomes diffusion-limited when that reserve is consumed:

• Exercise, which shortens transit time towards the equilibration time — the commonest single factor.
• Altitude, where the reduced alveolar PO₂ lowers the driving gradient so equilibration takes longer.
• A thickened blood–gas barrier — pulmonary fibrosis, interstitial oedema.
• Reduced surface area — emphysema, lobectomy, consolidation.

Usually it takes a combination: the classic case is the fibrotic patient who is normoxic at rest and desaturates on exertion. Two useful contrasts: nitrous oxide is the extreme perfusion-limited gas, because it is poorly bound and equilibrates almost at once; carbon monoxide is always diffusion-limited, because it binds haemoglobin so avidly that its partial pressure barely rises and the gradient never falls — which is why it is used to measure diffusing capacity.

Oxygen transport →

Saturation reports only one term of oxygen delivery. ḋO₂ = cardiac output × CaO₂, and CaO₂ itself depends on haemoglobin as well as saturation, so a normal SaO₂ is compatible with grossly inadequate delivery. Work through the four types of tissue hypoxia:

• Anaemic. Content is low despite full saturation. Check the haemoglobin — and remember that carbon monoxide and methaemoglobinaemia both reduce carrying capacity while a conventional oximeter reads normally or high. Co-oximetry is the test.
• Stagnant. Content is normal but flow is not — cardiogenic shock, hypovolaemia, an aortic cross-clamp, or regional occlusion such as mesenteric ischaemia. Expect a widened arteriovenous difference and a low SvO₂.
• Histotoxic. Delivery is entirely normal but the mitochondria cannot use the oxygen — cyanide inhibiting cytochrome oxidase. The signature is a high SvO₂ with a lactic acidosis.
• Hypoxic is the one you have already excluded, since the saturation is 100%.

Two non-hypoxic causes deserve a sentence: impaired lactate clearance in hepatic failure, and type B lactic acidosis from metformin, adrenaline infusion or thiamine deficiency. So: send a co-oximetry sample and a mixed venous gas, and assess the cardiac output. The mixed venous saturation is the single most discriminating measurement — low points to flow, high points to utilisation.

Oxygen transport →

V/Q = 0 is shunt; V/Q approaching infinity is alveolar dead space.

Physiology SAQ Paper 1 · Q2 →

Pulmonary vasodilators and some anaesthetic effects can inhibit HPV and increase perfusion of poorly ventilated lung.

Physiology SAQ Paper 1 · Q2 →

It estimates the percentage saturation of pulsatile arterial haemoglobin, not haemoglobin concentration or oxygen content.

Physiology SAQ Paper 1 · Q5 →

Higher CO, increased extraction, redistribution of flow and increased 2,3-DPG.

Physiology SAQ Paper 1 · Q5 →