Oxygen Is a Drug: The Dangers of the Reflex Non-Rebreather

The patient is breathing hard and you reach for the oxygen. It is the most reflexive move in EMS — hypoxia is the enemy, oxygen is the antidote, so more must be better. A non-rebreather at 15 litres, and you have done your job.

Except oxygen is a drug. It has a dose, a therapeutic window, and a real capacity to harm when you overshoot. And “slap on a non-rebreather” is exactly how you overshoot.

Oxygen Has a Target, Not a Maximum

The goal is not the highest saturation you can achieve. It is a normal one. The British Thoracic Society guideline — the reference standard for this — sets two working targets (O’Driscoll et al., Thorax 2017):

• 94–98% for most acutely unwell adults.
• 88–92% for patients at risk of hypercapnic (carbon-dioxide-retaining) respiratory failure — COPD, obesity-hypoventilation, neuromuscular disease, chest-wall deformity, cystic fibrosis.

Note what this does not say: it does not say withhold oxygen from the critically hypoxic patient. A saturation of 80% needs oxygen, now. The instruction is to titrate to a target — deliver enough to reach it and no more — rather than flood every patient with high-flow oxygen by reflex.

Why High-Flow Oxygen Hurts the Retainer

In a patient who retains CO₂, pouring on oxygen can drive the carbon dioxide up and tip them into a dangerous respiratory acidosis. Three mechanisms, and the old “you knocked out their hypoxic drive” explanation is the least important of them:

• V/Q mismatch (the big one). Diseased lung protects gas exchange by shunting blood away from poorly ventilated alveoli — hypoxic pulmonary vasoconstriction. Flood the alveoli with oxygen and that protective vasoconstriction relaxes, redirecting blood to areas that cannot ventilate. Dead space grows, and CO₂ that should have been exhaled is not.
• The Haldane effect. Deoxygenated hemoglobin carries CO₂ well; oxygenated hemoglobin does not. Saturate the blood with oxygen and hemoglobin releases its CO₂ load into the plasma, pushing the arterial CO₂ higher.
• Reduced respiratory drive. Real, but smaller and slower than the textbooks long implied. In a chronic retainer, removing the hypoxic stimulus can blunt ventilation — but the mismatch and Haldane effects do most of the damage.

The result is a rising CO₂, a falling pH, drowsiness, narcosis, and — if unchecked — respiratory arrest. This is iatrogenic acidosis: harm you created with a therapy meant to help.

It Is Not Only the Lungs

Hyperoxia has effects well beyond the airway. High arterial oxygen causes coronary and systemic vasoconstriction and reduces cardiac output; it drives absorption atelectasis and oxidative stress. This is not theoretical:

• In the prehospital AVOID trial, giving routine high-flow oxygen to normoxic STEMI patients was associated with larger infarcts, not smaller ones (Stub et al., Circulation 2015).
• The IOTA meta-analysis of more than 16,000 acutely ill adults found that liberal oxygen therapy increased mortality, with harm appearing above roughly 94–96% saturation (Chu et al., Lancet 2018).
• Most striking for us: a prehospital randomized trial in suspected COPD found that titrated oxygen roughly halved mortality compared with routine high-flow oxygen — the intervention was simply aiming for a target saturation instead of cranking the flow (Austin et al., BMJ 2010).

How to Give It Well

• Pick the target, then titrate to it with the lowest flow and concentration that holds it. Do not auto-apply a 15 litre non-rebreather — match the device to the need.
• Monitor continuously. Oxygen is a titrated drug; a saturation you set once and walk away from is a dose you are no longer controlling.
• In the retainer, get a blood gas where you can, and watch for the early signs of CO₂ narcosis — drowsiness, confusion, a patient who is “settling” a little too much.

The Pearl

Oxygen is a drug with a target, not a maximum. Aim for 94–98% in most adults and 88–92% in CO₂ retainers, and titrate with the lowest flow that gets you there. High-flow oxygen given by reflex drives up CO₂ through V/Q mismatch and the Haldane effect, causes vasoconstriction, and — in COPD, STEMI, and the acutely ill broadly — measurably increases harm. Treat hypoxia, not a number below 100%. Do not withhold oxygen from the truly hypoxic patient; just stop flooding everyone else.

---

References

• O’Driscoll BR, et al. BTS guideline for oxygen use in adults in healthcare and emergency settings. Thorax. 2017;72(Suppl 1):ii1–ii90. PubMed
• Austin MA, et al. Effect of high flow oxygen on mortality in chronic obstructive pulmonary disease patients in prehospital setting: randomised controlled trial. BMJ. 2010;341:c5462. PubMed
• Chu DK, et al. Mortality and morbidity in acutely ill adults treated with liberal versus conservative oxygen therapy (IOTA): a systematic review and meta-analysis. Lancet. 2018;391(10131):1693–1705. PubMed
• Stub D, et al. Air Versus Oxygen in ST-Segment-Elevation Myocardial Infarction (AVOID). Circulation. 2015;131(24):2143–2150. PubMed

This clinical pearl is for educational discussion only. Always follow your local protocols and medical direction. Your protocols exist for good reasons — this content is meant to enhance your clinical thinking, not replace your guidelines.