Oxygen levels, hypoxemia & O2 terminology

Oxygen is an essential medicine that is needed treat (e.g. pneumonia) or safely manage (e.g. safe surgery and anesthesia) patients with a wide range of conditions. Nonetheless, oxygen is not universally accessible in all health systems and for all patients. In this article we discuss some of the fundamental terminology related to oxygen.

Low blood oxygen levels is termed hypoxemia. Acute hypoxemia can cause tissue hypoxia (low oxygen levels at the cellular/tissue level) which can lead to organ dysfunction and death.

While there is no universal definition for hypoxemia, the most frequently used definition is a functional oxygen saturation (sO2) of <90% or an arterial partial pressure of oxygen of <60 mmHg. Below we define these terms.

Functional Oxygen Saturation (sO2) is the percent of hemoglobin which is capable of transporting oxygen Use the content level slider at the top left to view intermediate and advanced versions of this article for illustrations of functional and fractional saturation.. Functional oxygen saturation can be measured non-invasive by a pulse oximeter and is termed SpO2, or can be measured with a blood gas analysis by a co-oximeter and is termed SaO2 for arterial functional saturation, SvO2 for venous saturation, ScvO2 for central venous saturation an so on.

Check out more information on the clinical use of SpO2 and SpO2 clinical targets.

Fractional saturation (FO2Hb) measured both functional and non functional hemoglobin. Most commonly this refers to the fraction of total hemoglobin that is either bound by oxygen or carbon monoxide, or is oxidized to methemoglobin. Fractional saturation is measured by multi-wavelength co-oximetry which can be done non-invasively by specialized pulse co-oximeters or by blood gas analysis via co-oximeters (see figure below).

The arterial partial pressure of oxygen is the pressure of oxygen in arterial blood (PaO2). This is measured by arterial blood gas.

Oxygen content of the blood (CaO2) is determined by not only the functional oxygen saturation but also by hemoglobin concentration and to a much lesser extent, by PaO2. As illustrated by the formula below, at normal barometric pressures, PaO2 contributes relatively little to oxygen content as compared to oxygen bound by hemoglobin.

CaO2 = [Hb(1.34 ml O2/gm Hb)(SaO2)]+PaO2 x 0.003 ml O2/mmHg/dl)

Fraction of inspired oxygen (FiO2) is the proportion of inhaled gas that is oxygen. This is generally expressed as a decimal. For example, room air is 0.21 FiO2, whereas pure oxygen (i.e. no nitrogen) is 1.0 FiO2.

Read more about titration of FiO2

Atmospheric air (commonly referred to in clinical settings as ‘room air’) is composed of approximately 78.08% nitrogen, 20.95% oxygen, 0.93% argon, 0.04% carbon dioxide, variable amounts of water and negligible amounts of other gases. The majority of clinical focus is exclusively on the oxygen composition of the atmosphere, however, in select circumstances the other gases on the oxygen may have physiologic relevance Learn more by using the 'content level' slider at the top right to view the more advanced version of this article.

The optimal target for oxygen saturation (SpO2) in patients with acute hypoxemic respiratory failure is unknown. Hypoxemia causes pulmonary vasoconstriction and pulmonary hypertension in its chronic form, and death when it is acute and severe. Hyperoxemia also causes physiologic disturbances, through toxic reactive oxygen species and absorption atelectasis (Angus et al).

The World Health Organization (WHO) interim guidance for patients with hypoxemic respiratory failure due to COVID-19 suggests an initial SpO2 target >94% for stabilization, then >90% for non-pregnant patients and 92-95% for pregnant patients, once stable (WHO SARI Toolkit). The summary of evidence below suggests another reasonable target might be SpO2 90-96%, and perhaps 92-96% in settings with only intermittent pulse oximetry monitoring, or in patients with darker skin pigmentation.

The evidence

Several recent studies have examined conservative versus liberal oxygenation targets, with mixed results (See Table). In 2016, a before and after stepwise implementation study found a trend toward improved clinical outcomes with SpO2 target of 92-95% and PaO2 target of 55-86 mmHg as compared with earlier higher targets in ICU patients (Helmerhorst et al). Also in 2016, a single-center, open-label, randomized clinical trial compared targets of SpO2 94-98%/paO2 70-100 mmHg versus SpO2 97-100%/paO2 up to 150 mmHg in ICU patients, and found a mortality benefit with the lower target; however, the study was stopped early due to poor enrollment after a natural disaster, so its results cannot be considered definitive (Girardis et al).

In 2020, two multicenter randomized trials published together in the New England Journal of Medicine produced different results. One compared targets of SpO2 88-92%/paO2 55-70 mmHg with SpO2>=96%/PaO2 90-105 mmHg in ARDS patients, and was stopped early for suggestion of harm in the lower-SpO2 target arm (Barrot et al). The other compared one arm with a target SpO2<97% with an arm with no maximum target SpO2 limit (lower limit 90% for both arms) in mechanically ventilated patients, and found no difference in ventilator-free days or 180-day mortality (Mackle et al).

Finally, in a 2021 multicenter randomized trial also in the New England Journal of Medicine, investigators randomized almost 3,000 ICU patients with acute hypoxemic respiratory failure to receive PaO2 targets of either 60 mmHg or 90 mmHg (Schjorring et al). The median SpO2 in the higher-target group was 96% (IQR 95-97%), and in the lower-target group was 93% (IQR 92-94%). The study found no difference in 90-day mortality.

The evidence taken together suggests that a target range that avoids both hypoxemia and hyperoxemia may be beneficial. A reasonable example target range based on the above evidence is SpO2 of 90-96%.

Other considerations

Other factors may need to be considered when deciding on a Spo2 target range:

1) Resource-variable settings:

• Lower target saturations can conserve scarce oxygen resources, making more oxygen available for more patients who need it; a hospital Emergency Department in Rwanda found that a target of 90-95% resulted in better oxygen supply reserves for the hospital, as compared with previous higher SpO2 targets (Sutherland et al).
• Intermittent pulse oximetry monitoring (versus continuous) could increase the risk for periods of undetected hypoxemia.1 This could be an argument for a slightly higher target, for example 92-96%.
• The accuracy of inexpensive pulse oximeters without regulatory approval is variable (Lipnick et al). Using validated oximeters and following trends in SpO2 rather than individual measurements, and/or checking saturations with more than one device, may help mitigate this concern.

2) Race and skin pigmentation: Data suggest that pulse oximetry more frequently under reports hypoxemia in patients who self-identify as Black, as compared with patients who self-identify as White (Sjoding et al). While more work needs to be done with specific devices and with documentation of a validated range of skin pigmentations, it may be reasonable currently to target higher ranges in patients with darker skin tones, for instance 93-96%.

3) Altitude: Geographic variation in elevation above sea level may necessitate adjustment in SpO2 targets. Baseline SpO2 values in healthy people will be lower at higher elevation due to decreases in the partial pressure of oxygen.

4) Other conditions: The recommendations here are related to acute hypoxemic respiratory failure. Other conditions may necessitate higher or lower targets. For example, a COPD exacerbation in a patient with chronic hypercarbic respiratory failure should likely have a lower target of 88% to avoid worsening of hypercarbia. Pregnant patients generally have higher SpO2 goals (WHO SARI Toolkit).

According to the WHO, medicinal oxygen contains >82% pure oxygen, is free from contamination and must come from an oil-free compressor. Oxygen was added to the World Health Organization (WHO) list of essential medicines (EML) in 1979 for use in anesthesia, and was amended in 2017 for treating hypoxemia. There is ongoing debate and lack of consensus around the minimum required oxygen concentration for clinical/medical grade oxygen sources.

While the European Pharmacopoeia and the US Pharmacopeia include “Oxygen 93%,” the International Pharmacopeia 10th Edition 2020 contains only “Oxygen 99%.” The 11th edition is currently being revised with consideration to also add 93% by WHO. Of note, there is the International Pharmacopoeia, regional pharmacopoeias (e.g. European), and national pharmacopoeias (e.g. United States Pharmacopeia), each with defined values for medical grade oxygen concentration and purity.

It is widely considered that medicinal oxygen supply should contain >93% oxygen purity from a certified medical oxygen supplier or source. Of note, it is common that oxygen concentration from various sources (e.g. PSA plants, cylinders, portable concentrators) can vary widely and facility level quality control with oxygen analyzers is always recommended.

(Read more on oxygen supply types)

Depending on the source of oxygen (e.g. liquid oxygen, pressure swing adsorption plant, portable bedside oxygen concentrator etc) and the quality of that source, the concentration of oxygen produced and delivered to the patient’s bedside can vary dramatically.

• Liquid oxygen produces the highest purity of medical oxygen supply at >99.5% purity. “Oxygen 99.5%” should have <67 ppm water, <300ppm CO2, <5ppm CO
• PSA and VSA plants produce oxygen at approximately 93% ± 3% purity. “Oxygen 93%” should contain >90% O2, <300ppm CO2, <5ppm CO, NO and NO2 <2ppm, SO2 <1ppm, oil <1 mh/m3, water <67 ppm
• Portable PSA devices produce oxygen at approximately 93% ± 3 purity though this can vary with flow and device quality. According to the WHO, such devices should have fail alarms if concentration drops <82%.

Monitoring purity as well as impurities is essential for ensuring quality of care. Pharmacopoeias provide guidance (though sometimes varied) about how testing for purities and impurities must be done (e.g. high-pressure liquid chromatography or gas chromatography by national or 3rd party lab testing sites). Little guidance exists to dictate the frequency/intervals for such such checks or how to conduct onsite, self testing.

• Winter, Bruce, Bach, Jay, Eger. “The anesthetic effect of air at atmospheric pressure.” Anesthesiology. 1975
• Parker C, Snowdon S. “Predicted and measured oxygen concentrations in the circle system using low fresh gas flows with oxygen supplied by an oxygen concentrator,” British Journal of Anaesthesia, 1988 Oct;61(4):397-402
• Angus DC. Oxygen Therapy for the Critically Ill. N Engl J Med 2020;382:1054-6.
• World Health Organization. Clinical management of severe acute respiratory infection (SARI) when COVID-19 disease is suspected: Interim guidance. 2020 March 13.
• Helmerhorst HJ, Schultz MJ, van der Voort PH, Bosman RJ, Juffermans NP, de Wilde RB, van den Akker-van Marle ME, van Bodegom-Vos L, de Vries M, Eslami S, de Keizer NF, Abu-Hanna A, van Westerloo DJ, de Jonge E. Effectiveness and Clinical Outcomes of a Two-Step Implementation of Conservative Oxygenation Targets in Critically Ill Patients: A Before and After Trial. Crit Care Med 2016;44:554-63.
• Girardis M, Busani S, Damiani E, Donati A, Rinaldi L, Marudi A, Morelli A, Antonelli M, Singer M. Effect of Conservative vs Conventional Oxygen Therapy on Mortality Among Patients in an Intensive Care Unit: The Oxygen-ICU Randomized Clinical Trial. Jama 2016;316:1583-9.
• Barrot L, Asfar P, Mauny F, Winiszewski H, Montini F, Badie J, Quenot JP, Pili-Floury S, Bouhemad B, Louis G, Souweine B, Collange O, Pottecher J, Levy B, Puyraveau M, Vettoretti L, Constantin JM, Capellier G. Liberal or Conservative Oxygen Therapy for Acute Respiratory Distress Syndrome. N Engl J Med 2020;382:999-1008.
• Mackle D, Bellomo R, Bailey M, Beasley R, Deane A, Eastwood G, Finfer S, Freebairn R, King V, Linke N, Litton E, McArthur C, McGuinness S, Panwar R, Young P. Conservative Oxygen Therapy during Mechanical Ventilation in the ICU. N Engl J Med 2020;382:989-98.
• Schjørring OL, Klitgaard TL, Perner A, Wetterslev J, Lange T, Siegemund M, Bäcklund M, Keus F, Laake JH, Morgan M, Thormar KM, Rosborg SA, Bisgaard J, Erntgaard AES, Lynnerup AH, Pedersen RL, Crescioli E, Gielstrup TC, Behzadi MT, Poulsen LM, Estrup S, Laigaard JP, Andersen C, Mortensen CB, Brand BA, White J, Jarnvig IL, Møller MH, Quist L, Bestle MH, Schønemann-Lund M, Kamper MK, Hindborg M, Hollinger A, Gebhard CE, Zellweger N, Meyhoff CS, Hjort M, Bech LK, Grøfte T, Bundgaard H, Østergaard LHM, Thyø MA, Hildebrandt T, Uslu B, Sølling CG, Møller-Nielsen N, Brøchner AC, Borup M, Okkonen M, Dieperink W, Pedersen UG, Andreasen AS, Buus L, Aslam TN, Winding RR, Schefold JC, Thorup SB, Iversen SA, Engstrøm J, Kjær MN, Rasmussen BS. Lower or Higher Oxygenation Targets for Acute Hypoxemic Respiratory Failure. N Engl J Med 2021.
• Sutherland T, Moriau V, Niyonzima JM, Mueller A, Kabeja L, Twagirumugabe T, Rosenberg N, Umuhire OF, Talmor DS, Riviello ED. The “Just Right” Amount of Oxygen. Improving Oxygen Use in a Rwandan Emergency Department. Ann Am Thorac Soc 2019;16:1138-42.
• Lipnick MS, Feiner JR, Au P, Bernstein M, Bickler PE. The Accuracy of 6 Inexpensive Pulse Oximeters Not Cleared by the Food and Drug Administration: The Possible Global Public Health Implications. Anesth Analg 2016;123:338-45.
• Sjoding MW, Dickson RP, Iwashyna TJ, Gay SE, Valley TS. Racial Bias in Pulse Oximetry Measurement. N Engl J Med 2020;383:2477-8.
• WHO Good manufacturing practices for medicinal gases. WHO. 2021
• Medicinal oxygen working document. WHO. 2022
• International Pharmacopoeia, 10th Edition, 2020
• WHO Model Lists of Essential Medicine, 2022