Oxygen therapy nursing: devices, flow rates, and clinical decision-making

Oxygen therapy is among the most commonly administered treatments in acute care, yet it is also among the most frequently misapplied. Nurses administer supplemental oxygen to correct hypoxemia — an insufficient partial pressure of oxygen in arterial blood — and to protect tissues from the downstream consequences of oxygen deprivation: organ dysfunction, cardiac dysrhythmias, altered mental status, and death. The decision of when to start oxygen, which device to use, how much to give, and when to wean it requires clinical reasoning, not habit. A nasal cannula delivering 2 L/min is appropriate for a patient with mild community-acquired pneumonia; the same device is wholly inadequate for a patient in anaphylactic shock. Getting this wrong exposes the patient to either unresolved hypoxemia or the toxicity of excess oxygen. Both are dangerous.

This guide covers oxygen delivery systems, flow rate selection, SpO2 targets by condition, the hypoxic drive controversy, pulse oximetry limitations, oxygen toxicity, and the NCLEX-focused pearls most likely to appear on licensing exams.

Oxygen delivery devices: flow rates and FiO2

FiO2 — the fraction of inspired oxygen — is what actually matters clinically. Room air is 21% oxygen (FiO2 0.21). Supplemental oxygen raises FiO2 above this baseline. The device selected determines both the achievable FiO2 range and the precision of delivery.

FiO2 estimation for nasal cannula: The rough formula is FiO2 (%) ≈ 20 + (4 × L/min). At 1 L/min ≈ 24%, 2 L/min ≈ 28%, 4 L/min ≈ 36%, 6 L/min ≈ 44%. These are estimates — patient respiratory rate, tidal volume, and leak all affect delivered FiO2 in practice.

Oxygen administration: step-by-step procedure

Step 1: assess the patient

Before selecting a device, establish the clinical picture. Obtain SpO2, respiratory rate, work of breathing (use of accessory muscles, nasal flaring, paradoxical breathing), and mental status. Know the baseline — a patient with COPD who normally runs SpO2 88–90% is different from a post-surgical patient whose baseline is 98%. Note whether the patient is mouth breathing, as this significantly reduces nasal cannula effectiveness.

Review the physician or provider order. In most settings, nurses can initiate emergency oxygen at their clinical judgment, but standing orders or protocols should guide routine administration.

Step 2: select the device

Match the device to the clinical need:

• Mild hypoxemia (SpO2 90–94%, no distress) — nasal cannula at 2–4 L/min
• Moderate hypoxemia or respiratory distress — simple face mask or PNRM
• Severe hypoxemia, high oxygen requirement, or pre-intubation bridge — NRM or HFNC
• COPD or hypercapnic respiratory failure requiring precise FiO2 — Venturi mask
• Patient requiring FiO2 >60% without immediate intubation — HFNC

Step 3: apply and initiate

Confirm correct flow rate is set before applying the device to the patient. For a non-rebreather mask, inflate the reservoir bag before placing it on the patient’s face. Ensure proper fit — mask should cover nose and mouth fully; nasal cannula prongs should curve into the nares, not press against the philtrum.

Step 4: titrate to target

Oxygen is a drug — titrate it. Reassess SpO2 at 5 minutes after initiation. Adjust flow to meet the target SpO2 for this patient’s condition (see Clinical decision-making section below). The goal is the lowest FiO2 that achieves the target SpO2, not the highest FiO2 available.

Step 5: monitor continuously

Reassess work of breathing, mental status, respiratory rate, and SpO2 at regular intervals. Worsening despite supplemental oxygen is an escalation signal — not a sign to increase flow on the same device. A patient on a non-rebreather mask whose SpO2 remains 84% needs escalation to HFNC, BiPAP, or intubation, not 16 L/min on the NRM.

Step 6: wean when appropriate

Wean oxygen proactively once the underlying cause is resolving and the patient maintains target SpO2 on a lower FiO2. Wean the flow rate incrementally — reduce by 1–2 L/min, reassess SpO2 at 15–30 minutes, continue if SpO2 holds. Step down devices (NRM → PNRM → simple mask → NC) as tolerated. Do not wait for a physician order to wean if you have a wean order or standing protocol — prolonged high-flow oxygen carries its own risks.

Clinical decision-making: SpO2 targets by condition

SpO2 targets are not universal. The British Thoracic Society (BTS) guidelines and supporting evidence establish condition-specific ranges:

General inpatient (most conditions): SpO2 94–98%. Most post-surgical patients, pneumonia, pulmonary embolism, sepsis, and similar conditions target this range. Avoid routinely targeting SpO2 >98% — hyperoxia offers no benefit and carries risk.

COPD and hypercapnic respiratory failure: SpO2 88–92%. Evidence from multiple trials, including the DECAF study and Bardsley et al. (2018), confirms this narrower target reduces risk of hypercapnic respiratory failure from oxygen-driven loss of hypoxic ventilatory drive (see The hypoxic drive question below). Use a Venturi mask to hit the target precisely.

Acute MI / STEMI without hypoxemia: SpO2 ≥94% only if hypoxemic. The AVOID and DETO2X trials showed that high-flow oxygen in normoxic STEMI patients was associated with worse outcomes — increased myocardial infarct size. Do not apply supplemental oxygen to a patient with acute MI if SpO2 is already ≥94%.

Neonates and preterm infants: Targets vary by gestational age and condition. Preterm infants generally target SpO2 91–95% to balance retinopathy of prematurity risk with hypoxemia risk. Neonatal resuscitation begins with room air in term infants. This is a specialized area — defer to NICU protocols.

Carbon monoxide poisoning: 100% FiO2 via NRM regardless of SpO2 reading. SpO2 is unreliable in CO poisoning (see Pulse oximetry limitations). High FiO2 accelerates CO displacement from hemoglobin.

When to escalate device

Escalate when the patient fails to achieve target SpO2 on the current device despite adequate flow, or when clinical status is deteriorating despite supplemental oxygen. Escalation sequence:

NC → Simple mask → PNRM → NRM → HFNC or BiPAP/CPAP → intubation and mechanical ventilation

Call the provider when: SpO2 fails to reach target on NRM at 15 L/min, respiratory rate >30 and rising, increasing work of breathing, altered mental status, or the patient looks like they are tiring. Do not wait for the SpO2 to crash — respiratory arrest is preceded by a period of compensated distress that is the intervention window. For a deeper look at escalation decision-making in deteriorating patients, see acute respiratory failure nursing.

The hypoxic drive question

The hypoxic drive myth — that giving supplemental oxygen to a patient with COPD will stop them breathing — is one of the most persistently misapplied concepts in clinical practice. The reality is more nuanced.

What is real: Patients with chronic hypercapnia (chronically elevated PaCO2, common in severe COPD) can have a blunted central chemoreceptor response to CO2. In these patients, hypoxemia becomes a relatively more important driver of ventilation — hence the term hypoxic drive. Giving high-flow uncontrolled oxygen can increase PaO2, reduce this hypoxic stimulus, and worsen hypoventilation and CO2 retention.

What is overstated: The blanket fear of oxygen in any patient with COPD is dangerous. A hypoxic patient with COPD who is not breathing adequately needs oxygen. Withholding O2 from a patient with SpO2 78% because they have COPD is not cautious — it is harmful. The Ronning and Guldvog trial and subsequent data confirmed that uncontrolled high-flow oxygen in acute COPD exacerbation increases mortality compared to titrated oxygen. The mechanism is partly hypoxic drive suppression and partly the Haldane effect (CO2 displacement from hemoglobin by oxygen, increasing PaCO2 without true hypoventilation).

Clinical application: Titrate oxygen in COPD to SpO2 88–92%. Use a Venturi mask for precision. Obtain an arterial blood gas if the patient is known hypercapnic or deteriorating. The BTS 2017 guideline is explicit: target 88–92% in known or risk-of-hypercapnic respiratory failure, and escalate to NIV if pH falls below 7.35 despite controlled oxygen. The point is controlled titration, not oxygen avoidance.

For COPD-specific nursing considerations, see COPD nursing. For ARDS and higher-acuity oxygen failure, see ARDS nursing.

Pulse oximetry limitations

SpO2 is not the same as SaO2 (arterial oxygen saturation). Pulse oximetry estimates saturation by comparing absorption of red and infrared light — it assumes hemoglobin is either fully oxygenated or fully deoxygenated. This assumption fails in several clinically important scenarios:

Carbon monoxide poisoning: Carboxyhemoglobin (COHb) absorbs red light similarly to oxyhemoglobin, causing the oximeter to read COHb as oxygenated hemoglobin. A patient with 40% carboxyhemoglobin may show SpO2 of 98% while being severely hypoxic at the tissue level. SpO2 is unreliable in CO poisoning — use co-oximetry (arterial blood gas with co-oximeter).

Methemoglobinemia: Methemoglobin (MetHb) causes SpO2 to trend toward 85% regardless of actual saturation — it does not reflect the true oxygen-carrying capacity. Suspect in patients exposed to nitrites, dapsone, or local anesthetics (benzocaine) who appear cyanotic with a “normal” or fixed SpO2. Treat with methylene blue.

Peripheral hypoperfusion: Poor peripheral perfusion — from shock, vasoconstriction, hypothermia, or peripheral vascular disease — reduces the pulsatile signal that the oximeter relies on. Readings become unreliable, intermittent, or unobtainable. Move the probe to the forehead or earlobe if finger readings are poor.

Nail polish and artificial nails: Dark nail polish, particularly blue and black, can falsely lower SpO2 readings by blocking light transmission. Remove polish or reposition to a bare digit, earlobe, or forehead. Acrylic nails can also interfere.

Anemia: Severe anemia does not directly cause false SpO2 readings — what hemoglobin remains is still measured. However, SpO2 96% with a hemoglobin of 5 g/dL represents severely impaired oxygen delivery (DO2 = CO × CaO2) despite a reassuring saturation number. Do not rely on SpO2 alone in anemic patients.

Skin tone and SpO2 accuracy: Multiple studies — including the 2020 analysis by Sjoding et al. in the New England Journal of Medicine — demonstrated that pulse oximeters overestimate true arterial oxygen saturation in patients with darker skin tones. The disparity is clinically meaningful: Black patients were approximately three times more likely than white patients to have occult hypoxemia (SaO2 <88% on blood gas) with a SpO2 reading of 92–96%. This is an unresolved design limitation of current oximetry technology. In patients with darker skin tones or in any patient where clinical status does not match the SpO2 reading, obtain an arterial blood gas for definitive measurement.

Motion artifact: Movement causes false low readings or signal loss. Ensure the patient is still during the reading; some devices have motion-filtering algorithms but accuracy is reduced.

Oxygen toxicity

Supplemental oxygen is not benign at high concentrations or prolonged exposure. Two main mechanisms drive oxygen toxicity:

Absorption atelectasis: When breathing high FiO2, nitrogen — the dominant gas in room air that normally splints alveoli open — is washed out and replaced by oxygen. Oxygen is rapidly absorbed into pulmonary capillaries, collapsing alveoli and creating atelectasis. This paradoxically worsens oxygenation and can impair lung mechanics. This effect is seen even at FiO2 of 0.5–0.6 over time.

Free radical (reactive oxygen species) damage: High PaO2 generates superoxide radicals and hydrogen peroxide within lung tissue. Prolonged exposure — generally FiO2 >0.6 for more than 24–48 hours — causes oxidative injury to alveolar epithelium, type II pneumocytes, and pulmonary endothelium, resulting in diffuse alveolar damage that is histologically indistinguishable from ARDS. This is sometimes called pulmonary oxygen toxicity or hyperoxic lung injury.

Clinical threshold: Avoid FiO2 >60% for extended periods unless there is no alternative. In mechanically ventilated patients, the goal is to reduce FiO2 to ≤0.6 as soon as PEEP and other strategies permit. In spontaneously breathing patients on NRM, transition to HFNC or NIV if FiO2 requirements remain high. For the mechanics of ventilator management in these patients, see mechanical ventilation nursing.

Retinopathy of prematurity: In neonates, hyperoxia causes abnormal retinal vascularization and can lead to blindness. This is a distinct mechanism from adult oxygen toxicity and underscores why SpO2 targets in preterm infants are intentionally lower than in adults.

Common mistakes

Using nasal cannula when flow requirements exceed device capability. NC maxes out at 6 L/min (~44% FiO2). A patient with SpO2 of 82% on 6 L/min NC needs device escalation — not 8 or 10 L/min on a device not designed for it. NCLEX scenarios frequently test this: recognizing when to change devices, not just increase flow.

Not titrating — setting and forgetting. Oxygen is prescribed for a target, not a flow rate. If SpO2 rises to 100% on 6 L/min NC in a patient with a COPD exacerbation, the correct response is to reduce flow to the target range (88–92%), not to leave it unchanged.

Running a simple face mask below 6 L/min. Below 6 L/min, the mask accumulates exhaled CO2 and the patient rebreathes it. The minimum safe flow for any face mask is 6 L/min.

Assuming a normal SpO2 rules out a problem. In CO poisoning, methemoglobinemia, or severe anemia, SpO2 can be falsely reassuring. A patient who looks sick despite SpO2 of 96% warrants investigation — not dismissal.

Applying supplemental oxygen to a normoxic MI patient. The reflex to “put everyone on oxygen” in acute chest pain is harmful in patients with SpO2 ≥94%. Current evidence supports oxygen only when the patient is hypoxemic.

Neglecting to inflate the reservoir bag before placing an NRM. If the bag is not pre-inflated, the patient’s first breath collapses it and they inhale against resistance. Pre-inflate before application.

Not reassessing after device escalation. A patient placed on NRM at 15 L/min with SpO2 of 86% who does not improve to ≥90% within 10–15 minutes needs the next escalation step, not more time on the same device.

NCLEX tips

• The minimum flow rate for any face mask (simple, PNRM, NRM) is 6 L/min — this prevents CO2 rebreathing.
• Nasal cannula delivers approximately 4% FiO2 per liter: 1 L/min = 24%, 2 L/min = 28%, 3 L/min = 32%, 4 L/min = 36%, 5 L/min = 40%, 6 L/min = 44%.
• The non-rebreather mask provides the highest FiO2 of all low-flow devices — up to 90% with a tight seal. It is the first choice for acute severe hypoxemia before intubation.
• COPD patients target SpO2 88–92%, not 94–98%. The Venturi mask is the preferred device for controlled, precise oxygen delivery in COPD.
• The Venturi mask is the only low-flow system that delivers a reliably precise FiO2 — critical for any patient where controlled oxygen matters.
• SpO2 is falsely normal in carbon monoxide poisoning — always check carboxyhemoglobin via co-oximetry, and give 100% oxygen via NRM regardless of SpO2.
• SpO2 trends toward 85% in methemoglobinemia regardless of actual hemoglobin saturation.
• Oxygen toxicity risk increases with FiO2 >60% maintained for extended periods — mechanisms include absorption atelectasis and free radical lung damage.
• High-flow nasal cannula provides flow rates up to 60 L/min, heated humidification, and titrable FiO2 from 21–100%. It generates a small amount of positive airway pressure and is evidence-based for hypoxemic respiratory failure.
• Never withhold oxygen from an acutely hypoxic patient regardless of diagnosis. The concern in COPD is excess oxygen causing hypoventilation — not oxygen itself. Titrate; do not avoid.
• Pulse oximetry overestimates true saturation in patients with darker skin tones — a clinically significant disparity documented in peer-reviewed literature. Obtain an ABG when clinical assessment doesn’t match the SpO2 reading.
• Device escalation sequence: nasal cannula → simple face mask → partial non-rebreather → non-rebreather → HFNC/NIV → mechanical ventilation.
• Reservoir bag on an NRM must never fully collapse — if it does, the flow rate is too low and the patient is rebreathing exhaled gas.
• Oxygen is titrated to effect — the target is the lowest FiO2 that achieves the condition-appropriate SpO2 range. Hyperoxia (SpO2 consistently >98% on supplemental O2) is not a clinical goal.

Sources

• O’Driscoll BR, Howard LS, Earis J, Mak V; British Thoracic Society Emergency Oxygen Guideline Group. BTS guideline for oxygen use in adults in healthcare and emergency settings. Thorax. 2017;72(Suppl 1):ii1–ii90.
• Sjoding MW, Dickson RP, Iwashyna TJ, Gay SE, Valley TS. Racial bias in pulse oximetry measurement. N Engl J Med. 2020;383(25):2477–2478.
• Stub D, Smith K, Bernard S, et al.; AVOID Investigators. Air versus oxygen in ST-segment–elevation myocardial infarction. Circulation. 2015;131(24):2143–2150.
• Hofmann R, James SK, Jernberg T, et al.; DETO2X–SWEDEHEART Investigators. Oxygen therapy in suspected acute myocardial infarction. N Engl J Med. 2017;377(13):1240–1249.
• Frat JP, Thille AW, Mercat A, et al.; FLORALI Study Group. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N Engl J Med. 2015;372(23):2185–2196.
• Nolan JP, Soar J, Cariou A, et al. European Resuscitation Council and European Society of Intensive Care Medicine guidelines for post-resuscitation care 2015. Resuscitation. 2015;95:202–222.
• Bardsley G, Pilcher J, McKinstry S, et al. Oxygen versus air-driven nebulisers for exacerbations of chronic obstructive pulmonary disease. BMC Pulm Med. 2018;18(1):157.

For related content, see asthma nursing and ARDS nursing.