You're staring at a wall of green tubing, a humidifier bottle bubbling away, and a flow meter with that little ball bouncing somewhere between 2 and 6 liters. The patient's saturations are holding at 92%. Which device is on their face right now? The doctor wants them at 94%. You reach for the dial — but wait. Because that changes everything.
I've watched new nurses and respiratory therapists make the same mistake a hundred times. Now, they treat flow rate like a universal language. Here's the thing — it's not. A nasal cannula at 4 L/min delivers something completely different than a simple mask at 4 L/min. And a non-rebreather at 4 L/min? That's practically useless.
Let's clear this up once and for all.
What Is Oxygen Delivery
Oxygen delivery devices are the interface between a gas source and a patient's airway. That's the technical definition. In practice, they're the tools that decide how much oxygen actually reaches the alveoli — not how much leaves the wall outlet The details matter here..
The main keyword here is FiO2 — fraction of inspired oxygen. Room air is 21%. A non-rebreather can push that above 90%. Worth adding: a nasal cannula at 6 L/min lands somewhere around 44%. But those numbers assume perfect fit, normal breathing patterns, and no leaks. That's why real patients breathe through their mouths. But they talk. They pull the cannula off to eat pudding.
Every device has a maximum FiO2 it can theoretically deliver. Think about it: every device has a flow rate range where it performs as designed. And every device has a sweet spot — the flow range where you get predictable, reliable oxygenation without wasting gas or drying out mucosa.
Why It Matters
Here's what happens when you mismatch device and flow: your COPD patient retains CO2 because you put them on a non-rebreather at 15 L/min "just to be safe.On top of that, " Your pneumonia patient stays hypoxemic because you left them on a nasal cannula at 3 L/min when they needed a Venturi mask at 40%. Your pediatric patient gets gastric distension from a simple mask running at 10 L/min because nobody checked the minimum flow requirement.
These aren't theoretical scenarios. They happen every shift in every hospital.
The device is the prescription. If you don't know what each device actually does at each flow setting, you're guessing. Flow rate is just the dial. And in respiratory care, guessing gets people intubated No workaround needed..
How It Works — Device by Device
Nasal Cannula
The workhorse. Low-flow. Plus, cheap. Comfortable enough for long-term use. But it's a variable performance device — meaning the FiO2 changes with the patient's inspiratory flow rate, tidal volume, and respiratory rate.
At 1 L/min: ~24% FiO2
At 2 L/min: ~28%
At 3 L/min: ~32%
At 4 L/min: ~36%
At 5 L/min: ~40%
At 6 L/min: ~44%
That's the textbook math. Practically speaking, the rule of thumb: 4% per liter above 20%. But that assumes a normal adult breathing quietly through their nose. That said, mouth breather? Cut those numbers by 5-10%. Tachypneic? Same problem — they're entraining more room air.
Max flow is 6 L/min. Above that, you're just drying the nares, causing epistaxis, and wasting oxygen. The gas exits the prongs at high velocity and creates turbulence, not higher FiO2. I've seen orders for "NC at 8 L/min.Here's the thing — " That's not a thing. Stop writing it.
Pro tip: Humidify at 4 L/min and above. Below that, the nose does a decent job on its own. But if the patient's on it for days? Humidify anyway. Dry mucosa bleeds. Bleeding obscures assessment. It's a cascade That's the part that actually makes a difference..
Simple Face Mask
Plastic mask. Practically speaking, flow range: 5–10 L/min. Worth adding: elastic strap. On the flip side, holes on the sides for exhalation and room air entrainment. FiO2: 35–55%.
Here's the catch — you cannot run this below 5 L/min. That's why at 4 L/min, the patient rebreathes their own CO2. Check the flow. Patient gets sleepy, confused, saturations look fine but pCO2 is climbing. The exhalation ports are too small. I've seen this missed on night shift more times than I can count. Every time.
At 10 L/min, you're maxing out the entrainment capacity. FiO2 tops out around 55-60%. Beyond that, you need a different device.
Fit matters. So naturally, they'll rip it off. Practically speaking, there's no perfect answer here. A loose mask entrains more room air — FiO2 drops. A tight mask on a claustrophobic patient? That's why we have options.
Partial Rebreather Mask
Looks like a simple mask but with a reservoir bag attached. Plus, no one-way valves. The bag fills during exhalation. During inspiration, the patient draws from the bag and entrains room air through the side ports That's the part that actually makes a difference..
Flow range: 6–10 L/min. FiO2: 40–70%.
The bag should stay inflated about 2/3 full during inspiration. Here's the thing — if it collapses completely, flow is too low. If it stays full and the patient's still desaturating, they need more FiO2 than this device can give — move up.
Biggest mistake: not checking the bag. I've walked into rooms where the bag was flat as a pancake because the flow was at 4 L/min. That's not a partial rebreather anymore. That's a CO2 trap Most people skip this — try not to..
Non-Rebreather Mask (NRB)
The heavy hitter. Reservoir bag plus one-way valves on the exhalation ports and between the bag and mask. Goal: prevent room air entrainment. Deliver the highest FiO2 possible without intubation And that's really what it comes down to. Less friction, more output..
Flow range: 10–15 L/min. FiO2: 60–90%+.
Critical: The bag must be pre-inflated before you put it on the patient. If you slap it on empty, the first few breaths pull from a flat bag — they get room air. Pre-fill it. Squeeze the bag. Watch it expand. Then apply.
At 15 L/min with a good seal, you can hit 90%+. But "good seal" is the operative phrase. Think about it: edentulous patients. Agitated patients pulling the mask away. Beards. Every leak drops FiO2 fast That alone is useful..
And please — don't leave someone on an NRB for hours without reassessing. That's a bridge. Still, either they improve and step down, or they deteriorate and need more support. An NRB is not a destination.
Venturi Mask
The precision instrument. Now, color-coded adapters. Each color = fixed FiO2 regardless of flow (within range). Uses the Venturi effect — high-velocity oxygen jet entrains a precise amount of room air through calibrated ports Less friction, more output..
Colors and specs:
-
Blue: 24% at 2–4 L/min
-
White: 28% at 4–6 L/min
-
Yellow
-
Yellow: 31% at 6–8 L/min
-
Orange: 37% at 8–12 L/min
-
Red: 40% at 10–15 L/min
-
Green: 50% at 12–15 L/min
The beauty? That's why a patient removes half the mask, and you're still delivering 31% on yellow. If the seal breaks, FiO2 doesn't skyrocket — it stays predictable. That's reliability you don't get with NRB.
The catch? Plus, flow must stay within range. Plus, drop below 2 L/min on blue, and FiO2 plummets. Go above 15 L/min, and you're just wasting oxygen. These masks demand respect for their numbers.
High-Flow Nasal Cannula (HFNC)
The new kid on the block, but already indispensable. Plus, delivers heated, humidified oxygen at flows up to 60 L/min. The high flow creates a positive airway pressure of about 2–5 cmH2O — enough to stent open the airways and washout CO2.
FiO2 ranges from 21–100%, controlled precisely by blending oxygen with room air. In practice, the nasal prongs sit comfortably in the nostrils, and patients can talk, eat, and swallow secretions normally. No mask anxiety.
But it's not magic. Requires specialized equipment and trained staff. Not available everywhere. And while it can reduce work of breathing and improve oxygenation, it's not for everyone — severe hypoxemia or respiratory failure may need escalation.
Tracheostomy Mask
For the trached patient. So similar to NRB but designed for the opening. High FiO2, high flow. That's why fits over the stoma site. Watch for granulation tissue and ensure proper sizing.
Choosing Your Weapon
Start low and go slow. Try 24% with a Venturi blue. Because of that, orange or red Venturi, or NRB at 10–12 L/min. Practically speaking, a patient in mild respiratory distress? Now, moderate distress? Severe? HFNC or NRB, reassess every 15 minutes.
Document FiO2, flow rate, and patient response. On top of that, if they're getting better, step down. If they're not improving in 30–60 minutes, escalate. This isn't set-it-and-forget-it medicine.
Common Pitfalls
- Assuming higher flow = better oxygenation (it doesn't, beyond entrainment limits)
- Forgetting to check reservoir bags
- Ignoring mask fit and leaks
- Not reassessing
- Mixing up adapters or colors
Bottom Line
Oxygen delivery isn't one-size-fits-all. Check them. Day to day, know them. But each device has its place, limits, and failure modes. Change course when needed.
The patient's saturation is just the beginning. Watch their mental status, respiratory rate, and work of breathing. Oxygen is a tool, not a cure. Use it wisely.
