Why oxygenation fails on one lung

Four mechanisms, one failure. Why supplemental oxygen fixes dead space and not shunt, why the dependent lung is usually the problem rather than the collapsed one, and why the same reflex that saves you segmentally harms you globally.

Reviewed and updated September 2026 · Written by Anastasia Jackson, DNAP, CRNA

Every clinical claim on this page is cited to its source below. How we source clinical content — and the corrections we have made, with what each one was and why it changed.

Turning the oxygen up corrects one of these mechanisms completely and another not at all. The four questions below cover the physiology that decides which situation you are in.

Why does hypoxic pulmonary vasoconstriction happen, and what blunts it?

It is a rapid, reversible rise in pulmonary vascular resistance when alveolar oxygen tension falls below a threshold. The response is intrinsic to the lung, occurring in small pulmonary arteries, and it diverts blood away from poorly ventilated regions toward better-oxygenated ones — matching ventilation to perfusion and raising systemic oxygen delivery.^1,2,3

The mechanism

The sensor is mitochondrial. It sits in the electron transport chain of the pulmonary artery smooth muscle cell, including NDUFS2 in Complex I.³ Those mitochondria vary their production of reactive oxygen species and hydrogen peroxide in proportion to alveolar oxygen tension. Hypoxia reduces that signal and shifts the cell toward a reduced state, which inhibits oxygen-sensitive voltage-gated potassium channels including Kv1.5 and Kv2.1. The membrane depolarizes, L-type calcium channels open, cytosolic calcium rises, and the contractile apparatus activates.^2,3 Sustained hypoxia recruits Rho kinase, which reinforces the constriction through calcium sensitization.²

What blunts it

Onset occurs around a PaO₂ of 85–90 mmHg with maximum response near 65–70 mmHg.⁴

Shunt and dead space: why does oxygen fix one and not the other?

These are the two ends of ventilation–perfusion mismatch. Shunt is perfusion without ventilation — blood reaches the arterial side without contacting a ventilated alveolus, a ventilation–perfusion ratio of zero. Dead space is ventilation without perfusion — gas moves through units with little or no blood flow, a ratio approaching infinity.^5,6

Why oxygen fails against shunt

Blood leaving a shunt unit stays at mixed venous oxygen content, then mixes with well-oxygenated blood from the rest of the lung. Because the oxyhemoglobin dissociation curve is flat at its top, the normal units are already near-saturated and cannot compensate by carrying more. Raising the inspired oxygen fraction adds almost nothing to those units and nothing at all to blood that never sees an alveolus.⁵

Dead space behaves oppositely. It wastes ventilation and impairs carbon dioxide clearance, and it is corrected by raising minute ventilation rather than oxygen.⁶

Cardiac output moves shunt in both directions and is worth understanding as a trap: lowering it reduces the shunt fraction but also lowers mixed venous saturation, and the net effect on PaO₂ is a fall.⁴

Why does one-lung ventilation cause hypoxemia?

With both lungs perfused and only one ventilated, theoretical shunt approaches half the cardiac output. Observed shunt is much lower because two mechanisms redistribute flow — but the residual is still enough to cause hypoxemia in roughly 5–10% of cases.⁷

What limits the shunt

Gravity in the lateral decubitus position preferentially perfuses the dependent, ventilated lung. Hypoxic pulmonary vasoconstriction in the collapsed lung can reduce shunt flow through the non-dependent lung by about 50%.⁸ Total shunt is therefore residual non-dependent flow plus the dependent lung’s roughly 5% baseline.⁸

Why the ventilated lung is the problem

The dependent lung is doing all the gas exchange while being actively disadvantaged. It lies under the mediastinum and abdominal contents, its resting volume is reduced, and it collapses easily. It is both the only functioning lung and a compromised one — which is why the useful intervention is usually directed there rather than at the collapsed lung.

The escalation sequence follows the mechanism: recruit and apply positive end-expiratory pressure to the ventilated lung first, then continuous positive airway pressure to the operative lung, then intermittent two-lung ventilation. Position and agent are decisions made before the desaturation, not after it.

Why does functional residual capacity fall after induction, and how does atelectasis follow?

Functional residual capacity is the resting lung volume at end-expiration — the reservoir that keeps small airways open and continues oxygenating during apnea. Anesthesia reduces it, and the reduction produces atelectasis in roughly 90% of patients, typically 15–20% of the lung at the bases, before surgery begins.^10,11

Two losses that stack

Once resting volume falls below closing capacity, dependent airways close. Three mechanisms then produce collapse: compression from the weight of lung and abdominal contents, which is fast; absorption, where gas behind a closed airway is taken up faster than it is replaced, which is slow and strongly oxygen-dependent; and surfactant impairment.^10,11

Obesity and pregnancy amplify the volume loss through the same abdominal mechanism.¹² Pneumoperitoneum and Trendelenburg add compression. Continuous positive airway pressure during induction and positive end-expiratory pressure afterward prevent much of it, and recruitment maneuvers reverse established collapse.⁶

Atelectasis is the principal cause of impaired oxygenation under anesthesia, and it converts into shunt — which closes the loop with the second question on this page. That is why the fix is recruitment and positive end-expiratory pressure rather than more oxygen, and why the collapse can persist for days and contribute to postoperative pulmonary complications.

Frequently asked questions

Should I use propofol instead of a volatile agent for one-lung ventilation?

The mechanistic argument favors it, since volatile anesthetics attenuate hypoxic pulmonary vasoconstriction dose-dependently by opening potassium channels and inhibiting calcium influx, while propofol does not appear to interfere with the oxygen-sensing pathway.² Whether that translates into a clinically meaningful difference in a given case depends on how much of the shunt reduction is coming from the reflex versus from gravity, and on the other reasons an agent gets chosen.

My patient desaturated on one lung — what do I do first?

Address the ventilated lung before escalating oxygen. Desaturation during one-lung ventilation is more often dependent-lung derecruitment than shunt through the collapsed lung, because that lung sits under the mediastinum and abdominal contents with a reduced resting volume.⁸ Recruitment and positive end-expiratory pressure to the ventilated lung come first, then continuous positive airway pressure to the operative lung, then intermittent two-lung ventilation.

Why does lowering cardiac output not improve oxygenation even though it reduces shunt?

Because it moves two variables in opposite directions. A fall in cardiac output does reduce the shunt fraction, but it also lowers mixed venous oxygen saturation, and the net effect on PaO₂ is a fall.⁴ This is the reason hemodynamic management and oxygenation cannot be treated as separate problems during thoracic surgery.

Does preoxygenating with 100% oxygen cause harm?

It causes absorption atelectasis, and the magnitude is large — the critical ventilation–perfusion ratio at which alveoli collapse is roughly 0.001 breathing air versus 0.07 breathing 100% oxygen.¹⁰ That is a real cost weighed against a real benefit, since the same 100% oxygen extends safe apnea time. Continuous positive airway pressure during induction mitigates much of the collapse without giving up the reservoir.^10,11

References

1. Archer S, Michelakis E. The mechanism(s) of hypoxic pulmonary vasoconstriction: potassium channels, redox O₂ sensors, and controversies. News Physiol Sci. 2002. doi:10.1152/nips.01388.2002
2. Dunham-Snary KJ, Wu D, Sykes EA, Thakrar A, Parlow LRG, Mewburn JD, Parlow JL, Archer SL. Hypoxic pulmonary vasoconstriction: from molecular mechanisms to medicine. Chest. 2017;151(1):181–192. doi:10.1016/j.chest.2016.09.001. PMID 27645688
3. Archer SL, Dunham-Snary KJ, Bentley RET, Alizadeh E, Weir EK. Hypoxic pulmonary vasoconstriction: an important component of the homeostatic oxygen sensing system. Physiol Res. 2024;73(S2):S493–S510. doi:10.33549/physiolres.935431. PMID 39589299
4. Shum S, et al. Hypoxaemia during one lung ventilation. BJA Educ. 2023;23(9):328–336. doi:10.1016/j.bjae.2023.05.006. Source of the threshold PaO₂ of 85–90 mmHg with maximum response near 65–70 mmHg, and of the cardiac output and mixed venous saturation relationship.
5. Petersson J, Glenny RW. Gas exchange and ventilation–perfusion relationships in the lung. Eur Respir J. 2014;44(4):1023–1041. PMID 25063240
6. Raimondi Cominesi D, et al. Pulmonary shunt in critical care: a practical approach with clinical scenarios. J Anesth Analg Crit Care. 2024. doi:10.1186/s44158-024-00147-5
7. Lee K, et al. Effects of iloprost on oxygenation during one-lung ventilation in patients with low diffusing capacity for carbon monoxide: a randomized controlled study. J Clin Med. 2022;11(6). PMID 35329869. Source of the 5–10% incidence of hypoxemia.
8. Slinger P, Campos JH. Physiology of the lateral decubitus position, open chest and one-lung ventilation. In: Principles and Practice of Anesthesia for Thoracic Surgery. Figures modified from Benumof, Elsevier 1995. Edition and page not yet confirmed.
9. Li X, et al. The effects of thoracic epidural analgesia on oxygenation and pulmonary shunt fraction during one-lung ventilation: a meta-analysis. BMC Anesthesiol. 2015. doi:10.1186/s12871-015-0142-5
10. Hedenstierna G, Edmark L. Mechanisms of atelectasis in the perioperative period. Best Pract Res Clin Anaesthesiol. 2010;24(2):157–169. PMID 20608554. Source of the FRC figures, the 90% incidence, the 15–20% basal collapse and the critical ventilation–perfusion values.
11. Effects of anesthesia on the respiratory system. OpenAnesthesia. Corroborates the FRC figures and the three atelectasis mechanisms; no stable identifier or version.
12. Aretha D, et al. Safety and effectiveness of alveolar recruitment maneuvers and positive end-expiratory pressure during general anesthesia for cesarean section: a prospective, randomized trial. Int J Obstet Anesth. 2017;30:30–38. PMID 28108076