Key Facts

Oxygen therapy is one of the most commonly administered treatments in emergency and hospital medicine. Despite its ubiquity, inappropriate oxygen use can cause harm — both under-oxygenation (hypoxic organ damage) and over-oxygenation (oxygen toxicity, absorption atelectasis, and worsened hypercapnia in susceptible patients).

Epidemiology

Oxygen is administered to approximately 34% of emergency ambulance patients and >25% of hospital inpatients. LTOT is prescribed to approximately 85,000 patients in England. COPD is the indication for LTOT in approximately 70% of patients.

Aetiology

Oxygen therapy is indicated for hypoxaemia of any cause:

• Acute: pneumonia, asthma, COPD exacerbation, PE, heart failure, trauma, sepsis, anaphylaxis
• Chronic: COPD, ILD, pulmonary hypertension, cystic fibrosis, sleep-disordered breathing with daytime hypoxaemia

Pathophysiology

Oxygen toxicity: high FiO₂ generates reactive oxygen species (superoxide, hydroxyl radicals) causing oxidative damage to alveolar epithelium. Absorption atelectasis occurs when high FiO₂ washes out alveolar nitrogen, causing alveolar collapse in poorly ventilated units.

Hypercapnia risk: In susceptible patients (especially COPD), excessive oxygen administration can worsen CO₂ retention via three mechanisms:

1. Haldane effect: oxyhaemoglobin binds less CO₂ → releases CO₂ into blood
2. V/Q mismatch worsening: relief of hypoxic pulmonary vasoconstriction redirects blood to poorly ventilated areas
3. Reduced hypoxic drive: minor contributor (historically overemphasised)

Indications for Acute Oxygen Therapy

• SpO₂ <94% (or <88% in hypercapnic risk patients) in acutely unwell patients
• Cardiac arrest, major trauma, anaphylaxis, carbon monoxide poisoning (give 100% regardless of SpO₂)
• Peri-operative and critical care settings

Signs of Hypoxaemia

• Tachypnoea, dyspnoea, use of accessory muscles
• Tachycardia, confusion, agitation
• Central cyanosis (SpO₂ typically <85%)

Signs of Oxygen Toxicity/Hypercapnia from Over-oxygenation

• Drowsiness, confusion, flapping tremor
• Headache, warm peripheries, bounding pulse
• Rising PaCO₂ on ABG

Red Flags

• SpO₂ <85% or rapidly falling despite oxygen
• CO₂ narcosis (drowsy on oxygen, rising PaCO₂)
• Carbon monoxide poisoning (SpO₂ may read falsely normal)
• New oxygen requirement in previously well patient (investigate cause)

Bedside

• SpO₂: continuous pulse oximetry — note limitations (CO poisoning, methaemoglobinaemia, poor perfusion, nail varnish)
• ABG: gold standard — PaO₂, PaCO₂, pH, HCO₃⁻, lactate, carboxyHb, metHb
• Respiratory rate, GCS: clinical assessment

Bloods

• FBC: polycythaemia (chronic hypoxia)
• U&Es, BNP: underlying cause assessment

Imaging

• CXR: underlying cause of hypoxaemia

Special Tests (for LTOT assessment)

• ABG on air (×2, at least 3 weeks apart, when stable): required for LTOT prescription
• Overnight oximetry: if nocturnal desaturation suspected
• 6-minute walk test with oximetry: for ambulatory oxygen assessment
• Spirometry: COPD severity assessment

Non-pharmacological

• Emergency oxygen: give 15L/min via non-rebreathe mask for cardiac arrest, major trauma, anaphylaxis, CO poisoning
• Titrate to target: 94-98% (most patients); 88-92% (hypercapnia risk — use 24% or 28% Venturi mask initially)
• Delivery devices:
  • Nasal cannulae: 1-6 L/min (FiO₂ ~24-44%)
  • Simple face mask: 5-10 L/min (FiO₂ ~40-60%)
  • Venturi mask: precise FiO₂ (24%, 28%, 35%, 40%, 60%)
  • Non-rebreathe mask: 15 L/min (FiO₂ ~85%)
  • High-flow nasal oxygen (HFNO): up to 60 L/min, FiO₂ 21-100% — humidified, generates low-level PEEP

Pharmacological

• LTOT (≥15 hours/day): for COPD patients meeting criteria (PaO₂ ≤7.3 kPa stable, or 7.3-8 kPa with complications) — MRC trial (1981) and NOTT trial (1980) demonstrated mortality benefit
• Flow rate typically 1-4 L/min via concentrator, titrated to PaO₂ >8 kPa without rising PaCO₂ >1 kPa
• Ambulatory oxygen: portable cylinder/liquid O₂ for patients who desaturate on exertion and show ≥1 kPa PaO₂ improvement or subjective benefit on formal walking test
• Short-burst oxygen: NOT recommended for breathlessness without hypoxaemia — LOTT trial showed no benefit
• Palliative oxygen: for comfort in end-of-life care; handheld fan may be equally effective for dyspnoea relief

Surgical/Interventional

• Not applicable as primary intervention

Referral Criteria

• Home oxygen assessment service for LTOT/ambulatory oxygen assessment
• Respiratory specialist if new/unexplained oxygen requirement
• Smoking cessation service (mandatory before LTOT — fire risk)

LTOT improves survival in COPD with chronic hypoxaemia: the MRC trial showed 5-year survival improvement from 25% to 42%. LTOT must be used ≥15 hours/day for mortality benefit. Survival benefit has NOT been demonstrated for moderate resting or exertional hypoxaemia (LOTT trial). Patients on LTOT who continue to smoke have significantly increased fire risk. Mean survival on LTOT for COPD is approximately 3-5 years, depending on severity of underlying disease.

Oxygen Delivery Devices Summary

LTOT Prescribing Criteria (BTS Guideline)