Bedside Snapshot
- Normal (Adult)
- 95–100% on room air (typical; lab-specific). Co-oximetry SaO₂ on an arterial sample; SpO₂ is an estimate.
- Critical Values
- No universal critical SaO₂. Act on SaO₂ below the patient's target (below 94% in most acutely ill adults; below 88% when hypercapnic risk), treat SaO₂ <90% (PaO₂ about 60 mmHg) as urgent, and act on any SpO₂ and co-oximetry mismatch.
- High Suggests
- Supplemental oxygen above need; SaO₂ above target gives no added benefit
- Low Suggests
- V/Q mismatch: pneumonia, PE, asthma, COPD
- Shunt: ARDS, pulmonary edema, atelectasis
- Hypoventilation: opioids, CNS depression
- Dyshemoglobin: carbon monoxide, methemoglobin
- Act Now If
- SaO₂ or SpO₂ below target, or SaO₂ <90%: oxygen to target, assess airway and breathing, ABG with co-oximetry if SpO₂ may be unreliable, treat the cause; escalate per protocol
- Watch Out For
- A normal SpO₂ misses CO poisoning and does not exclude hypoxemia in shock or darker skin; methemoglobinemia can hold SpO₂ near 85%.
Reference & Critical Ranges
Lab-specific ranges: Every lab sets its own range. The analyzer and population change it too. Read the range on your lab report first. The values below are typical.
Reference ranges
| Population | Conventional | SI | Notes |
|---|---|---|---|
| Adult, arterial, room air, sea level | 95–100% | 0.95–1.00 (fraction) | Co-oximetry on an arterial sample. Lower at altitude and with low inspired O₂. |
| Children | Same as adult | Same as adult | Interpret with PaO₂ and FiO₂. Neonatal targets differ; see Special Populations. |
Severity and oxygen targets
| Level | Conventional | SI | What it means |
|---|---|---|---|
| Normal | 95–100% | 0.95–1.00 | Adequate oxygenation for most adults on room air. |
| Target, most acutely ill adults | 94–98% | 0.94–0.98 | BTS 2017 target when the patient is not at risk of hypercapnic failure. |
| Target, hypercapnic risk (e.g., COPD) | 88–92% | 0.88–0.92 | Use until blood gas results are known. Protects against CO₂ retention. |
| Hypoxemia | <90% | <0.90 | PaO₂ about 60 mmHg. On the steep part of the curve, where small drops in PaO₂ cause large drops in SaO₂. |
| Critical | No universal value | — | No laboratory-wide critical SaO₂ is defined. Treat SaO₂ below the patient's target range as requiring action; below 90% is urgent hypoxemia. |
Critical values: There is no universal critical SaO₂ value. Act when SaO₂ falls below the patient's target range: below 94% in most acutely ill adults, and below 88% when there is hypercapnic risk. SaO₂ below 90% is urgent hypoxemia. It is a clinical threshold, not a lab panic value. Act on any mismatch between SpO₂ and co-oximetry.
Conversion: Nothing to convert. Saturation is a percentage in both systems. PaO₂ in kPa = mmHg ÷ 7.5.
Alveolar-arterial (A-a) gradient
PAO₂ = FiO₂ × (760 − 47) − PaCO₂ ÷ 0.8; A-a = PAO₂ − PaO₂
Normal A-a gradient rises with age. A widened gradient points to intrinsic lung pathology. A normal gradient with low PaO₂ points to hypoventilation or low inspired oxygen.
Causes of Abnormal Results
Low (hypoxemia, SaO₂ below 90%)
| Mechanism | Common causes | Bedside clue |
|---|---|---|
| V/Q mismatch | Pneumonia, asthma, COPD, pulmonary embolism | Partial response to supplemental oxygen; work of breathing, wheeze or crackles |
| Shunt | Pulmonary edema, ARDS, atelectasis, right-to-left cardiac shunt | Poor response to oxygen; bilateral infiltrates; cyanotic congenital heart disease |
| Diffusion impairment | Interstitial lung disease, pulmonary fibrosis | Desaturation with exertion more than at rest |
| Hypoventilation | Opioids, sedatives, neuromuscular disease, obesity hypoventilation | Rising PaCO₂, slow respiratory rate or depressed mental status |
| Low inspired PO₂ | Altitude; low FiO₂ delivery | Normal A-a gradient; history of altitude exposure or a low-flow device |
| Dyshemoglobin | Carbon monoxide, methemoglobin | SpO₂ does not match co-oximetry; chocolate-brown blood |
Oxygen delivery drops with low cardiac output or anemia. SaO₂ does not. Check lactate and hemoglobin in those patients.
High (SaO₂ above target)
Usually this is supplemental oxygen above what the patient needs. No added benefit is expected above the target range. Titrate down.
Dyshemoglobins and measured SaO₂
Co-oximetry reports oxyhemoglobin as a fraction of total hemoglobin. The more carboxyhemoglobin or methemoglobin, the lower the reported SaO₂. Pulse oximetry can still read normal, or near 85%.
Carbon monoxide: Whenever CO exposure is possible, measure COHb on co-oximetry. Your SpO₂ can read normal or high even while the co-oximetry SaO₂ is low.
Interpretation & Bedside Approach
Step-by-step approach
- Is it real? Compare SpO₂ with the co-oximetry SaO₂. Check the sample (air bubbles, venous contamination, delay to analysis) and the sensor (perfusion, nail polish, motion).
- How sick is the patient? Work of breathing, mental status, hemodynamics and the response to oxygen.
- Acute or chronic? Compare with prior values. A chronic lung patient may have a valid lower baseline and a deliberate lower target.
- Calculate derived values: A-a gradient and PaO₂/FiO₂ ratio where respiratory failure is suspected.
- Classify against the target: normal, most-acutely-ill target (94–98%) or hypercapnic-risk target (88–92%) per BTS 2017.
- Escalate or dispose: titrate oxygen, support ventilation and treat the cause per protocol; involve the physician early when SaO₂ stays below target.
Pattern and what it suggests
| Pattern | Suggests |
|---|---|
| Low SaO₂, low PaO₂, widened A-a gradient | Intrinsic lung disease: V/Q mismatch, shunt or diffusion impairment |
| Low SaO₂, high PaCO₂, normal A-a gradient | Hypoventilation |
| Low SaO₂, normal PaCO₂, normal A-a gradient, altitude | Low inspired PO₂ |
| SpO₂ normal, co-oximetry SaO₂ low | Carbon monoxide, or pulse-oximetry error |
| SpO₂ near 85%, high MetHb on co-oximetry | Methemoglobinemia |
| SpO₂ low, co-oximetry SaO₂ normal | Pulse-oximetry artifact: motion, poor perfusion or sensor position |
One value tells you less than the trend. A falling SaO₂ at a fixed FiO₂ warrants reassessment, even while it stays above 90%.
Management Priorities
Urgent: SaO₂ below target, or below 90%
- Restore oxygenation: titrate oxygen to the target range by the device your protocol specifies. Exception: suspected carbon monoxide poisoning gets 100% oxygen regardless of SpO₂.
- Protect ventilation: assess airway and breathing; escalate support per protocol if ventilation is inadequate.
- Confirm the measurement: obtain an arterial blood gas with co-oximetry when SpO₂ may be unreliable.
- Treat the cause: dyshemoglobin, pneumonia, PE, edema, opioid toxicity and other causes in the table above.
- Reassess: recheck SpO₂ and SaO₂ after each change in oxygen, position or ventilation.
Medications
| Drug | Adult dose | Pediatric dose (max) | Onset / key notes |
|---|---|---|---|
| Oxygen | Titrate to target: 94–98%; 88–92% if at risk of hypercapnic failure. Exception: suspected carbon monoxide poisoning gets 100% oxygen regardless of SpO₂ until COHb has fallen; do not titrate to SpO₂ | After ROSC: wean to SpO₂ 94–99% (2025 AHA Part 8, COR 2b); otherwise titrate to the condition-specific target. No fixed pediatric dose. | Device (nasal cannula, non-rebreather, high-flow or NIV) per protocol. Avoid unnecessary high FiO₂ except in suspected CO poisoning. |
| Methylene blue (methemoglobinemia) | 1 mg/kg IV over 5–30 min; repeat 1 mg/kg one hour later if MetHb stays above 30% or symptoms persist; if no response after 2 doses, consider alternatives. Moderate or severe renal impairment (eGFR 15–59 mL/min/1.73 m²): single 1 mg/kg dose only; if MetHb stays above 30% or symptoms persist at 1 h, consider alternative interventions | 1 mg/kg IV over 5–30 min; same dose as adults. Pediatric use is supported by limited case-series data (2 ProvayBlue patients, 12 on another methylene blue product) | Contraindicated in G6PD deficiency and in severe hypersensitivity to methylene blue or other thiazine dyes. May dilute in 50 mL D5W to reduce local pain; avoid saline diluent. Boxed warning: serotonin syndrome with SSRIs, SNRIs, MAOIs, opioids; avoid serotonergic drugs for 72 h after the last dose. Causes falsely low SpO₂. |
Methylene blue: Confirm G6PD status, renal function (eGFR) and the serotonergic drug list before dosing. In G6PD deficiency, or if methemoglobinemia does not resolve after the labeled doses, contact toxicology or poison control (1-800-222-1222 in the US). The label advises alternative interventions.
Definitive care and disposition
- Escalation: noninvasive ventilation, high-flow oxygen or intubation per local protocol when SaO₂ stays below target.
- After cardiac arrest (ROSC): give 100% oxygen until SpO₂ or PaO₂ can be measured reliably (2025 AHA Part 11, rec 1, COR 1). Then titrate to SpO₂ 90–98% (PaO₂ 60–105 mmHg) to avoid both hypoxemia and hyperoxemia (Part 11, rec 3, COR 2a). In infants and children, wean oxygen to SpO₂ 94–99% (2025 AHA Part 8, rec 2, COR 2b). Pulse oximetry can hide hypoxemia in darker skin (Part 11, rec 4, COR 2b).
- Carbon monoxide: give 100% oxygen regardless of SpO₂ until COHb has been measured and has fallen; do not titrate to SpO₂. Hyperbaric and toxicology referral per the carbon-monoxide reference page and local protocol.
- Disposition: ICU or step-down for persistent hypoxemia or a rising oxygen requirement.
Doses are typical references. Follow local protocols, medical direction and institutional policy.
Field & Transport Considerations
Point-of-care testing
In the field, pulse oximetry (SpO₂) is the usual measure. Arterial co-oximetry is the reference for SaO₂. Cartridge blood-gas analyzers report SaO₂, but only inside the temperature and hematocrit limits the manufacturer sets. Follow your device instructions.
Before and during interfacility transport
- Record SpO₂, device, FiO₂ or flow, and the time at departure and arrival.
- Recheck SpO₂ after every move, suction event and change in position. Even a sudden fall that stays inside the target range deserves a fuller assessment.
- Titrate oxygen up when below target and down when above target, especially in patients at risk of CO₂ retention.
- Use an arterial sample when SpO₂ is unreliable (shock, poor perfusion, dark nail polish, suspected CO or methemoglobinemia).
Altitude and flight
Inspired oxygen falls as altitude rises, and SaO₂ falls with it. Cabin pressure and flight altitude change which target is achievable at a given FiO₂. Follow the transport service's protocol.
Special Populations
| Population | What changes | Clinical implication |
|---|---|---|
| Neonates and infants | Preductal SpO₂ target rises over the first minutes after birth: 65–70% at 2 min, 70–75% at 3 min, 75–80% at 4 min, 80–85% at 5 min and 85–95% at 10 min (2025 AHA Part 5, Targeted Oxygen Saturation Table) | Place a pulse oximeter as soon as possible (COR 1, rec 2); the sensor goes preductal, on the right hand or wrist. Titrate oxygen to the target (COR 1, rec 1). Infants 35 weeks or more may begin with 21% oxygen, 32 to 34+6 weeks with 21% to 30%, and under 32 weeks with 30% to 100% (2025 AHA Part 5, recs 3 to 5, COR 2b). Avoid hyperoxia as well as hypoxia once the target is reached. |
| Children and adolescents | Normal SaO₂ is the same as adult | After ROSC, wean oxygen to SpO₂ 94–99% (2025 AHA Part 8, COR 2b); otherwise titrate to the condition-specific target. Use the same pulse-oximetry error checks as adults. |
| Pregnancy | Maternal SaO₂ and PaO₂ targets follow the adult range | Treat maternal hypoxemia promptly; fetal oxygenation is not measured by maternal SaO₂. |
| Older adults | No clinically meaningful difference in the normal range is established here | Interpret with PaO₂ and chronic lung disease, not the number alone. |
| Chronic lung disease (COPD) | Hypercapnic-risk target 88–92% until blood gas results are known | Excess oxygen can worsen CO₂ retention (V/Q mismatch, Haldane effect, reduced ventilatory drive). |
| Darker skin pigmentation | Pulse oximetry can overestimate SaO₂ | Hidden hypoxemia is possible. Confirm with co-oximetry in critically ill patients. |
| High altitude | Lower inspired PO₂ lowers the normal SaO₂ for a given setting | Interpret with the altitude and FiO₂ at the time of measurement. |
Interfering Factors & Pitfalls
Collection and handling
| Factor | Effect | Why | Avoid or fix |
|---|---|---|---|
| Air bubbles in syringe | ↑ PaO₂ (SaO₂ may rise) | Air diffuses into the sample | Expel bubbles immediately and cap the syringe |
| Venous contamination | False − PaO₂ and SaO₂ | Mixed venous blood is less saturated | Confirm bright red, pulsatile arterial flow before drawing |
| Delayed analysis | ↓ PaO₂ | Cells keep consuming oxygen | Analyze promptly or keep on ice per lab policy |
| Excess liquid anticoagulant | Dilutes sample | Alters the measured values | Use the correct heparin quantity per lab policy |
Pulse oximetry (SpO₂) accuracy
| Factor | Effect | Why | Avoid or fix |
|---|---|---|---|
| Poor perfusion, shock, hypothermia | Unreliable or absent signal | Weak pulsatile signal | Use co-oximetry or move the sensor to a warmer site |
| Motion, shivering, seizures | Artifact | Movement corrupts the signal | Stabilize the limb or use a different site |
| Dark nail polish, artificial nails | False − or no reading | Light absorption by pigment | Remove the polish or use another digit |
| Darker skin pigmentation | False + (overestimates SaO₂) | Pigment affects light absorption | Confirm with co-oximetry in critically ill patients |
| Bright ambient light | Unreliable reading | Light interference | Shield the sensor |
Dyshemoglobins and dyes
| Substance | Effect on SpO₂ | Effect on co-oximetry SaO₂ | Avoid or fix |
|---|---|---|---|
| Carbon monoxide (COHb) | False + (reads normal or high) | Low: O₂Hb falls | Measure COHb on co-oximetry |
| Methemoglobin | Tends toward about 85% at high MetHb | Reports O₂Hb fraction correctly | Co-oximetry for MetHb and SaO₂ |
| Methylene blue | False − (underestimates) | Reports correctly | Use co-oximetry while dye is present |
| Indocyanine green | False − (dye interference) | Reports correctly | Use co-oximetry for accuracy |
Medications that cause methemoglobinemia
- Topical anesthetics: benzocaine (topical airway or mucosal anesthetic sprays); prilocaine; lidocaine at high dose
- Antibiotics and antimalarials: dapsone, sulfonamides, primaquine
- Other drugs: nitrites and nitrates, metoclopramide, phenazopyridine; inhaled nitric oxide
- Chemical exposures: aniline dyes, nitrobenzene
Common misinterpretations
- A normal SpO₂ can still hide hypoxemia in shock, CO poisoning or dark-pigmented skin.
- The pair of 90% and 60 mmHg sits on the standard curve. Right or left shifts (pH, temperature, 2,3-DPG) move that pair.
- A low PaO₂ with a normal SaO₂ is no contradiction. SaO₂ comes from PaO₂ on the curve.
- Oxygen above the target range adds nothing. It can do harm.
Physiology & Test Basics
What it measures
SaO₂ is the share of arterial hemoglobin carrying oxygen. Co-oximetry measures it on an arterial blood sample. SpO₂ is a non-invasive estimate from pulse oximetry, built on light absorption. The two agree in most patients. They can split apart in dyshemoglobinemia, shock, poor perfusion and darker skin pigmentation.
SpO₂ vs SaO₂: SaO₂ from co-oximetry is the measured arterial value. SpO₂ from pulse oximetry is an estimate. Device testing validates it across only a limited range of saturations.
Physiology and regulation
- Oxygen-hemoglobin dissociation curve: The curve is S-shaped. SaO₂ holds above 90% while PaO₂ slides to about 60 mmHg, then it drops steeply.
- Right shift: Acidosis, fever, higher 2,3-DPG and higher PaCO₂ all lower affinity. Hemoglobin lets go and hands the oxygen to tissues.
- Left shift: Alkalosis, low PaCO₂, hypothermia, lower 2,3-DPG (for example, stored blood), fetal hemoglobin and carboxyhemoglobin all raise affinity. Hemoglobin holds on tighter.
How it is measured
The sample goes into a heparinized syringe with no air, then gets analyzed promptly or kept on ice. Co-oximetry pulls oxyhemoglobin apart from COHb and MetHb. Pulse oximeters shine red and infrared light through tissue. They estimate saturation from the absorbance ratio, using the pulsatile signal.
Why it matters
SaO₂ and PaO₂ steer oxygen therapy, ventilator changes and escalation. Why can patients slide so fast once SaO₂ falls below 90%? The steep part of the curve.
Clinical Pearls
90-60 rule: On the standard curve, 90% SaO₂ sits next to a PaO₂ of about 60 mmHg. Past that point, a small drop in PaO₂ makes a large drop in SaO₂.
Titrate both ways: Below target, raise oxygen. Above target, lower it. In patients at risk of CO₂ retention, the lower target is 88–92%.
Transport recheck: Move the patient, suction them or reposition them, then recheck SpO₂. A sudden fall, even inside the target range, warrants reassessment.
Dark skin pigmentation: Pigmentation can throw off pulse oximeter accuracy, and the FDA draft guidance from January 2025 says so. Hidden hypoxemia is possible. Confirm critically ill patients with co-oximetry.
Methemoglobin: Think methemoglobinemia when SpO₂ sits low, tends toward about 85% at high MetHb, and refuses to rise with oxygen. Chocolate-brown blood adds to the picture. Confirm with co-oximetry. Methylene blue dosing is in Management Priorities.
Carbon monoxide: In CO poisoning, the SpO₂ can read normal. The patient may look well and still be severely hypoxic at the tissue level. Co-oximetry shows the low SaO₂.
Avoid excess FiO₂: Too much FiO₂ risks absorption atelectasis. Aim for the lowest FiO₂ that holds SaO₂ inside the target range.
References
- O'Driscoll BR, Howard LS, Earis J, Mak V, on behalf of the BTS Emergency Oxygen Guideline Development Group. (2017). BTS guideline for oxygen use in adults in healthcare and emergency settings. Thorax, 72(Suppl 1), ii1–ii90. https://www.brit-thoracic.org.uk/quality-improvement/guidelines/emergency-oxygen
- U.S. Food and Drug Administration. (2025). Pulse oximeters for medical purposes: non-clinical and clinical performance testing, labeling, and premarket submission recommendations. Draft guidance, docket FDA-2023-N-4976. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/pulse-oximeters-medical-purposes-non-clinical-and-clinical-performance-testing-labeling-and
- Sjoding MW, Dickson RP, Iwashyna TJ, Gay SE, Valley TS. (2020). Racial bias in pulse oximetry measurement. New England Journal of Medicine, 383(25), 2477–2478. https://doi.org/10.1056/NEJMc2029240
- ProvayBlue (methylene blue) injection, prescribing information. American Regent. U.S. Food and Drug Administration label, revised 2024. https://www.accessdata.fda.gov/drugsatfda_docs/label/2024/204630Orig1s023correctedlbl.pdf
- Lee HC, Strand ML, Finan E, et al. (2025). Part 5: Neonatal resuscitation: 2025 American Heart Association and American Academy of Pediatrics Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 152(suppl 2), S385–S423. doi:10.1161/CIR.0000000000001367. https://cpr.heart.org/en/resuscitation-science/cpr-and-ecc-guidelines/neonatal-resuscitation
- Lasa JJ, Dhillon GS, Duff JP, et al. (2025). Part 8: Pediatric advanced life support: 2025 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 152(suppl 2), S479–S537. doi:10.1161/CIR.0000000000001368. https://cpr.heart.org/en/resuscitation-science/cpr-and-ecc-guidelines/pediatric-advanced-life-support
- Hirsch KG, Amorim E, Coppler PJ, et al. (2025). Part 11: Post-cardiac arrest care: 2025 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 152(suppl 2), S673–S718. doi:10.1161/CIR.0000000000001375. https://cpr.heart.org/en/resuscitation-science/cpr-and-ecc-guidelines/post-cardiac-arrest-care
Medical Disclaimer
- For Educational Purposes Only: This content is intended for educational reference and should not be used for clinical decision-making.
- Not a Substitute for Professional Judgment: Always consult your local protocols, institutional guidelines, and supervising physicians.
- Verify Before Acting: Users are responsible for verifying information through authoritative sources before any clinical application.
AI Assistance Notice
AI was used to assist in organizing, paraphrasing, and formatting this information.