The clinical question
Oxygen treats hypoxemia in pneumothorax. Whether it should also be given to a patient with normal oxygenation solely to accelerate pleural-air absorption is a separate question. The physiological rationale is plausible, but clinical benefit remains uncertain. Suspected tension pneumothorax with instability requires immediate decompression; oxygen cannot relieve the obstructive physiology.
What nitrogen washout can and cannot do
A higher inspired oxygen concentration reduces alveolar and blood nitrogen partial pressure, increasing the gradient for nitrogen to leave the pleural space. This may speed absorption after the leak has sealed. It does not seal an ongoing air leak or reliably prevent an enlarging pneumothorax.
Northfield’s 1971 study compared 12 patients breathing air with 10 patients alternating air and high-concentration oxygen. Radiographic absorption increased about fourfold during oxygen periods. This small study of spontaneous pneumothorax does not establish a universal absorption rate or prove fewer procedures, shorter admissions or better patient comfort.
What the guidance actually says
BTS oxygen guidance (2017) recommends a reservoir mask at 15 L/min for hospital observation without drainage unless hypercapnic respiratory failure is a risk (grade D). The French multisociety PSP guideline (2023) recommends against systematic oxygen therapy. These positions differ.
The BTS 2023 pleural guideline and ERS/EACTS/ESTS 2024 guideline focus on pleural-management decisions and do not provide an equivalent explicit recommendation for routine nitrogen washout. An omitted recommendation is not a recommendation against treatment. Identify the local pathway rather than presenting a single international oxygen rule.
Apply the evidence to the pneumothorax type
Primary spontaneous pneumothorax
PSP occurs without clinically apparent underlying lung disease; a tall, thin body habitus is not a diagnostic requirement.
Park et al. (2017) retrospectively compared hospitalized patients receiving nasal oxygen at 2–4 L/min with outpatients breathing air. Radiographic resolution was faster with oxygen, but admission, activity, baseline size and imaging intervals differed. Fourteen episodes with enlargement, all in the oxygen group, were excluded. The result is hypothesis-supporting, not proof that routine oxygen improves recovery.
For selected minimally symptomatic, stable adults, conservative care can be considered regardless of size. Adequate observation, support and early follow-up are essential. If intervention is needed, discuss aspiration, ambulatory treatment or drainage through the respiratory pathway. ERS/EACTS/ESTS 2024 favors aspiration over a chest tube for initial PSP treatment.
Secondary spontaneous pneumothorax
Underlying lung disease reduces reserve: even a relatively small pneumothorax may cause substantial compromise. Correct hypoxemia and obtain prompt respiratory assessment, usually in hospital. Oxygen is supportive care, not a replacement for indicated pleural intervention.
The 2024 ERS/EACTS/ESTS panel could not recommend for or against conservative SSP management because evidence was insufficient. Any observation plan needs specialist selection, close reassessment and access to rescue drainage. Uncontrolled reports of oxygen-only treatment cannot define a safe size threshold.
Iatrogenic pneumothorax
After a procedure, establish the mechanism. A puncture-related leak, ventilator-associated injury and post-thoracentesis non-expandable lung are different problems. The BTS pleural-procedures statement distinguishes visceral pleural injury, air entrainment and pneumothorax ex vacuo.
Correct hypoxemia; reassess symptoms, physiology and imaging through the procedure-specific pathway. A stable patient may be observed, but progression or an ongoing leak may require intervention. Do not assume that another drain will re-expand a non-expandable lung, or apply an undefined percentage cutoff to every post-procedure pneumothorax.
Traumatic pneumothorax
Treat the patient’s resuscitation needs and associated injuries. There is no established patient-important benefit from oxygen solely to accelerate traumatic pleural-air absorption.
The WTA 2022 consensus algorithm permits observation of selected stable patients below its imaging thresholds: 35 mm on CT or about 20% on chest radiography. Selected patients receiving positive-pressure ventilation may also be observed. This requires close monitoring, repeat imaging, typically a chest radiograph within 6 hours, and prompt rescue drainage. Physiological compromise attributable to the pneumothorax, progression or inadequate monitoring changes the decision.
Trauma guidance is not uniform: WSES-AAST 2025 more strongly favors drainage during positive-pressure ventilation. Its conservative option for small, stable pneumothoraces includes at least 24 hours of observation. Do not combine thresholds and observation schedules from different algorithms; use the responsible trauma team’s pathway.
Prescribe oxygen to a target
For most acutely ill adults requiring oxygen, BTS uses SpO₂ 94–98%; for patients at risk of hypercapnic respiratory failure, use 88–92% pending blood gases and reassessment. SSP does not automatically mean hypercapnia risk, and a normal SpO₂ does not exclude CO₂ retention.
During critical-illness resuscitation, give high-concentration oxygen initially, then titrate using reliable oximetry and blood gases. Reassess increasing oxygen requirements promptly. These are BTS oxygen targets, not a mandate to keep every stable normoxemic patient on oxygen.
Do not confuse oxygen concentration with pressure support
A reservoir mask delivering high-concentration oxygen is not high-flow nasal cannula (HFNC) or non-invasive ventilation (NIV). Do not start HFNC or NIV solely for nitrogen washout.
BTS/ICS guidance advises draining an acute pneumothorax before NIV. If too small for safe drainage or suspected to be chronic, carefully monitored NIV may sometimes proceed under expert supervision. HFNC needs a separate clinical assessment; neither its safety nor its efficacy follows from reservoir-mask oxygen studies.
Reassessment and discharge
An improving saturation does not prove that the pneumothorax is resolving. Follow pain, breathlessness, respiratory effort, circulation and serial imaging as appropriate. Provide explicit return precautions for worsening breathlessness, chest pain or faintness, and ensure the responsible team has arranged follow-up.
Do not fly with unresolved pneumothorax. BTS advises waiting at least 7 days after radiographic confirmation of complete resolution; individual circumstances may require longer. BTS pleural guidance.
Clinical takeaway
Treat hypoxemia, recognize deterioration and choose pleural management from the patient’s physiology and pneumothorax type. Faster absorption on an image is not enough to justify routine prolonged oxygen in every normoxemic patient.
Sources and scope
The linked guidelines and primary studies support the distinctions above. This is an adult clinical overview, not a neonatal oxygen protocol. Guidance differs by setting; follow the applicable pleural and oxygen pathway.
