What We Don’t Know About Oxygen Can Kill.

Introduction

Picture this: a 67-year-old lady with severe COPD is wheeled into casualty, breathless and frightened. The nurse snaps on a simple face mask and cranks the oxygen to 10 L/min. Twenty minutes later she’s drowsy, her CO₂ has rocketed and we’re racing to the ICU. Was the nurse careless? No. She simply didn’t know that oxygen is a drug—one that can save or suffocate depending on how you use it. That scene played out so often during my specialist-registrar days at Whiston Hospital (part of St. Helens & Knowsley NHS Trust) that we decided to test how much frontline staff actually knew about oxygen delivery. Spoiler alert: the answer was “not enough.” What we found changed the way I teach, and it still keeps me awake at night now that I practise in India where the gaps are even wider.

Why we did the study

I’d just finished a night shift and was doing the oxygen-rounds with a mug of lukewarm coffee when I noticed something odd. Every patient—whether they’d had a heart attack, a broken rib or a flare-up of COPD—was on the same green plastic mask at 4 L/min. It was like watching every restaurant dish being seasoned with the same random pinch of salt. Oxygen is the commonest “prescription” in hospital, yet it’s the least understood. British Thoracic Society (BTS) had released crystal-clear guidelines in 2008, but nobody seemed to follow them. So, we devised a 10-scenario questionnaire mirroring real-life decisions:

• Cardiac arrest: bag-and-mask or non-rebreather?
• Pneumothorax: how high is “high-flow”?
• COPD exacerbation: Venturi or nasal cannula?
• Target saturations: 88–92 % vs 94–98 %? We handed it to 139 staff—41 % junior doctors, 59 % nurses—on medical wards, in corridors, even in the lift. The deal was: tick what you would do right now, no textbook peeking.

What we discovered (and the red faces)

The results were equal parts funny and terrifying.

1. Cardiac arrest: Over half chose a non-rebreather. I get it—there’s something heroic about that tight reservoir bag inflating, but it won’t deliver the 100 % oxygen and tight seal you need. Think of trying to inflate a party balloon through a straw.
2. Pneumothorax: Only 1 in 4 knew conservative management needs 10 L/min through a non-rebreather to speed re-absorption. The rest stuck with 2 L via nasal specs “so the patient doesn’t feel dry.” That’s like drying your hair with a desk fan.
3. COPD: Two-thirds missed the Venturi mask. Several wrote “nasal cannula at 1 L/min—low dose is safe.” True, but it’s also like watering a garden with an eye-dropper during a drought; you’ll still kill the roses.
4. Target saturations: The answers ranged from “above 80 % is fine” to “100 % for everyone.” Only 30 % could state the 88–92 % range for COPD. One doctor summed it up: “I just aim for 100 % minus a bit.” We presented these findings at the European Respiratory Society Congress in Barcelona 2010, and later that year at the BTS Winter Meeting. Each time the auditorium gave that collective guilty laugh—everyone recognised their own blind spots.

Why getting it wrong hurts people

Oxygen toxicity in COPD can tip a patient into hypercapnic coma. On the flip side, under-oxygenation during sepsis or major trauma fuels multi-organ failure. We calculated that inappropriate device selection occurred in roughly 1 in 3 prescriptions on our wards. Extrapolate that to a 600-bed trust and you’re looking at 200 chances every single day for preventable harm. That’s a jumbo-jet-load of risk.

Culture shock—moving to India

Fast-forward a couple of years; I’m now heading respiratory services in a busy Indian metropolis. On my first ward round I asked for a 28 % Venturi for a CO₂-retaining patient. The nurse blinked, rummaged through a drawer and produced a blue adapter I hadn’t seen since medical-school exams. “Sir, we just use this or a plain mask.” Nasal cannulas adorn almost every patient, oxygen cylinders are green instead of white, and the word “prescription” is alien. Target saturations? Blank looks. We pride ourselves on cheap, high-volume care, yet the humble oxygen device is stuck in 1970s mode.

Oxygen as a drug—my favourite analogy

Imagine oxygen is fire. Too little and you’re in the dark, too much and you burn the house down. The trick is to dial the exact flame you need:

• Room air = candle
• Nasal cannula 1–4 L = gas stove on simmer
• Simple mask 5–10 L = barbecue grill
• Venturi 24–60 % = Bunsen burner with precise collar
• Non-rebreather 10–15 L = blow-torch
• Bag-valve-mask with reservoir = industrial furnace You wouldn’t flambé a crème brûlée with a furnace, so don’t give 100 % oxygen to every breathless patient.

Practical take-aways for today’s clinician

1. Prescribe oxygen, don’t just “put it on.” Note the device, flow, and target saturation—exactly like you would for amoxicillin 500 mg TDS.
2. Know your two big numbers by heart:
   • 88–92 % for COPD, obesity-hypoventilation, severe kyphoscoliosis.
   • 94–98 % for everything else unless told otherwise.
3. Master four devices; everything else is window-dressing:
   • Nasal cannula (1–4 L)
   • Venturi mask (24, 28, 35, 40, 60 %)
   • Non-rebreather (70–90 %)
   • Bag-valve-mask (nearly 100 %)
4. Re-check in 5–10 minutes. Oxygen is dynamic; saturations after a cup of tea can drift. Document again.
5. Teach one colleague every week. Knowledge multiplies faster than any aerosol.

A blueprint for hospitals—UK, India or anywhere

• Wall charts above every oxygen port: colour-coded device + indication.
• Mandatory e-learning module (15 min) with certificate renewal yearly.
• Drug-chart redesign: oxygen section alongside antibiotics, insulin.
• Ward “oxygen champions” who audit monthly and give rapid feedback.
• Grand-round case presentations where junior doctors defend their oxygen plan.

My pledge (and a challenge for you)

I’ve started rolling out “Oxygen is a Drug” workshops in Indian nursing schools; the first batch of 120 students cut device-selection errors by half within two weeks. My goal is zero preventable hypercapnic comas in my unit this year. If you’re reading this—doctor, nurse, medical student, policymaker—join me. Print the BTS chart, stick it on the ward, quiz your team at coffee. Tag me on social media with your oxygen selfie; I’ll send you a virtual high-five and a pocket guide.

Conclusion

Oxygen is the only drug we give without thinking, yet it can maim when we forget the basics. Our UK study proved that even in a guideline-rich environment knowledge gaps are common; in India those gaps are chasms. The good news? The fix is cheap—education, checklists, a splash of humility and a willingness to treat every hissing cylinder with the same respect we give a vial of insulin. So next time you reach for that mask, ask yourself: “What flame do I need?” Because somewhere, a 67-year-old lady is counting on you to keep the lights on without burning the house down.

Remember: give oxygen like you’d season food—measure, taste, adjust, and never stop tasting.