CONTENTS
- Assessment of oxygenation:
- Assessment of ventilation:
- Additional topics related to pulse oximetry:
- Additional topics related to ABG/VBG
Pulse oximetry is utilized far more broadly for assessment of oxygenation than ABG (for every ABG drawn, there are thousands of oxygen saturation values measured). This is because pulse oximetry has numerous advantages over ABG for the assessment of oxygenation. Generally, pulse oximetry is the default tool for assessment of oxygenation.
pulse oximetry is a better measurement of oxygen delivery to the tissues than PaO2
(Systemic O2 delivery) = 13.4(cardiac output)(hemoglobin)(O2 saturation)
- The oxygen content of the blood depends most directly on oxygen saturation – not on PaO2. When considering whether there is an adequate amount of oxygen being provided to tissues, the oxygen saturation is therefore more important than the PaO2.
- Portable ABG meters often measure pO2 and subsequently estimate the oxygen saturation. Thus, the portable ABG meter may provide an inferior evaluation of systemic oxygen delivery as compared to pulse oximetry (which is directly measuring the oxygen saturation).
ABGs occasionally reflect venous samples
- ~10-15% of ABGs may actually reflect a venous sample (if the ABG isn't obtained via an arterial line).
- The oxygen extraction of the hand isn't very high, so a venous sample will often have only a slightly lower oxygen level than the arterial blood (so it often won't be obvious that it's a venous sample).
ABG is logistically problematic (invasive, painful, time-consuming, expensive)
- The total cost of drawing and analyzing an ABG is nearly ~$200. One ABG frequently leads to a follow-up ABG or a whole series of ABGs, so the costs can rapidly add up.
- Obtaining an ABG may delay care, as practitioners postpone decisions pending ABG results. In many cases, clinical decisions could be made more rapidly and accurately based simply on immediately available clinical data (e.g., decisions regarding intubation).
- If an arterial line is inserted to facilitate serial ABGs, this may lead to significant blood loss from phlebotomy.(2356936)
ABGs are a poor monitoring tool
- ABGs offer only intermittent snapshots of oxygenation, as compared to pulse oximetry (which can literally provide thousands of data points within the same time frame).
- ABG values often don't track with the patient's overall course; for example:
- (1) Patients with COPD or asthma who are treated with BiPAP will often experience immediate improvement in their work of breathing, with relatively stable ABG values. The patient is working less hard to achieve the same PaCO2 – which is a big clinical improvement. ABG values will eventually improve, but these may lag behind for several hours.
- (2) There is a substantial degree of random variation between sequential ABGs (discussed further below: 📖). Any intermittent monitoring tool with random variation will tend to be misleading about the patient's condition.
PaO2 values often induce panic
- PaO2 values are invariably much lower than oxygen saturation values:
- A saturation of 90% correlates with a PaO2 value of ~60 mm.
- A saturation of 88% may occur with a PaO2 value of ~55 mm.
- PaO2 values in the 50s often lead to excessive anxiety, despite the fact that they correlate with a reasonable oxygen saturation (especially among patients with chronic hypoxemia). This may lead to needless escalations in care and overly aggressive oxygen administration.
poor arterial waveform
- Patients with poor systemic perfusion occasionally will have an unreliable pulse oximetry waveform. Such patients obviously require ABG analysis to evaluate their oxygenation.
- (Please note, however – the absence of a reliable pulse oximetry waveform suggests severe hypoperfusion. These patients often have more of a perfusion problem than an oxygenation problem. So obtaining an ABG to evaluate oxygenation is great, but the primary focus of management should often be on restoration of systemic perfusion.)
dyshemoglobinemia
- Pulse oximetry will fail in the presence of dyshemoglobinemia (e.g., methemoglobinemia, sulfhemoglobinemia).
- Further discussion of methemoglobinemia here: 📖
PaO2/FiO2 ratio (aka, P/F ratio)
- The P/F ratio is utilized to assess oxygen among ventilated patients. It's not a terrific index (because it doesn't take PEEP into account). Nonetheless, this is often used as an indication for proning (based on its historical use in prior RCTs).
- Further discussion of proning in ARDS is here: 📖
A-a gradient
what is the A-a gradient?
- The A-a gradient is the the difference between inspired PO2 and the patient's PaO2. It is a measurement of lung function that is elevated by V/Q mismatch and/or shunting.
- To accurately measure the A-a gradient, an ABG is required in a patient who is breathing room air.
when is the A-a gradient useful?
- A normal A-a gradient may occasionally be useful to differentiate anxiety, vocal cord dysfunction, or severe asthma:
- The A-a gradient will be normal in anxiety or vocal cord dysfunction.
- The A-a gradient will be elevated in severe asthma.
- A normal A-a gradient may occasionally be useful to diagnose a patient with mild hypoxemia that is being caused by occult hypoventilation (e.g., 1-2 liter oxygen requirement).
A-a gradient measurement is generally not useful in critical care
- Any patient with critical hypoxemia (e.g., requiring >1-2 liters of oxygen) will have an elevated A-a gradient. In fact, most patients with critical hypoxemia are receiving too much oxygen to even measure an A-a gradient accurately (this requires leaving the patient on room air for 10-15 minutes to obtain an ABG on room air).
patients with dark skin
pulse oximetry over-estimates oxygen saturation in Black patients
- Pulse oximeters were optimized for use in patients with lighter skin.
- Numerous studies have recently demonstrated that among patients with dark skin, pulse oximetry has a reduced precision plus an average bias of ~+2% (so pulse oximetry will under-diagnose hypoxemia in Black patients).(33326721, 34592317, 35100193, 35231085, 35639368)
- More research is needed to compare different pulse oximeters, but this problem seems to be widespread across various brands.
clinical implications?
- There should be a reduced threshold to obtain an ABG among patients with dark skin if this will affect clinical management (e.g., in the management of COVID pneumonia, hypoxemia is an indication for hospital admission, oxygen administration, dexamethasone, and baricitinib).
- Oxygen saturation may remain a useful tool to trend oxygenation.
- Targeting an oxygen saturation goal of ~2% higher than usual might be reasonable. (However, since there is reduced precision among Black patients, simply subtracting 2% from pulse oxygenation values cannot fully resolve the problem. 🌊)
other (rare) causes of incorrect pulse oximetry results
other limitations of pulse oximetry:
- Nail polish.
- Intravenous dye administration (e.g., IV methylene blue).(26179876)
- Motion artefact (especially a tremor).
- Venous pulsation (e.g., severe tricuspid regurgitation combined with systolic arterial hypotension may cause the predominant pulsation detected to be venous rather than arterial).
Peripheral VBG is a more humane and convenient test:
- VBG may be measured along with routine venous phlebotomy for other labs.
- Peripheral IV lines can often be used to draw back a small sample of venous blood for analysis (again, without requiring a separate puncture).
VBG cannot be used to assess oxygenation, but it's generally adequate to assess pH and ventilation (pCO2 and pH).
The difference between ABG and VBG values is proportional to the difference in the oxygen saturation between arterial and venous blood:
This physiological truism has been shown experimentally using a variety of data sets. 🌊 The bottom line is that if the VBG oxygen saturation is reasonably high, it will be very close to the ABG values (figure below). If prolonged tourniquet time or delayed processing is avoided, this will usually allow a VBG to be used in place of an ABG.
- The etCO2 is always less than the arterial CO2 (PaCO2). The size of this gap depends on how well the lung is functioning.
- For patients without known lung disease, the PaCO2 is generally ~5-15 mm higher than the etCO2. Therefore, if you target an etCO2 of 30mm, the patient's PaCO2 will end up in a safe place (35-45 mm).
- EtCO2 measurement may allow for elimination of most ABG/VBG measurements among ventilated patients (discussed further here: 📖).
- Vasoconstriction causes a broad, short waveform whereas vasodilation causes a tall, narrow waveform.
- This is the electronic equivalent of a weak pulse (vasoconstriction) versus a “bounding” pulse (vasodilation).
- With progressive vasodilation, the dicrotic notch moves further away from the peak amplitude.
basics
- Perfusion index is the ratio of pulsatile blood volume as compared to the static blood volume. This is largely a reflection of the interplay between stroke volume and vascular tone. (34660615)
- Perfusion index may range roughly from 0.02% to 20%.
- Interpretation requires some nuance:
- Clinical significance depends greatly on the clinical context.
- Changes in perfusion index may be more significant than absolute values.
- Perfusion index is best poised to detect pathophysiologic states which have opposite effects on stroke volume and vascular tone, for example:
- Cardiogenic or hypovolemic shock tend to decrease stroke volume and increase vascular tone – so they cause a reduction in the perfusion index.
- Septic shock may decrease the stroke volume and decrease the vascular tone – which could have unpredictable effects on the perfusion index.
potential clinical applications
- ⚠️ The applications listed below have not been validated broadly across different patient populations and different devices.
- Accuracy of pulse oximetry: Perfusion index below ~0.3 suggests that the pulse oximetry value may be less reliable (although this may vary between different devices).(Lee 2020)
- Early detection of hypovolemia following trauma: Perfusion index <1 may be an early indicator of hemorrhage.(32439257)
- Prediction of intra-dialytic hypotension: A pre-dialysis perfusion index <1.8 may predict hypotension during intermittent hemodialysis.(31881971)
- Prediction of fluid responsiveness: Increase in the perfusion index by >10% during a passive leg raise may predict fluid responsiveness.(30658663)
- Predictor of successful weaning from ventilation: An increase in the perfusion index by >40% during a spontaneous breathing trial may predict successful extubation.(32036499)
limitations
- Local vasoconstriction may reflect hypothermia or cold environmental temperature (rather than systemic vasoconstriction).
- Local venous or arterial obstruction may affect the perfusion index.
- Digital compression by the pulse oximeter itself may decrease the pulsatility index.
- Venous pulsations due to tricuspid regurgitation might theoretically affect the perfusion index.
- Exact values may vary between different devices, so cutoff values may not necessarily be reproducible across different centers.(34687923)
ABG/VBG to evaluate the cause of respiratory distress
- ABG/VBG is often utilized as a diagnostic tool to determine the etiology of respiratory distress. Unfortunately, ABG/VBG has horrible performance for differentiating different causes of respiratory distress.
- For a patient in respiratory distress due to pulmonary dysfunction, ABG/VBG values are usually similar regardless of etiology (typically demonstrating a mild respiratory alkalosis; figure below). The only disorder which can be identified based on unique ABG/VBG values is anxiety-induced hyperventilation (since these patients can achieve extremely low pCO2 values).(21663600)
- Please note that a normal A-a gradient doesn't exclude pulmonary embolism.(7632205)
ABG/VBG to evaluate for hypercapnia
- ABG/VBG is a useful tool to evaluate for hypercapnia.
- For example:
- In a patient with COPD and encephalopathy, ABG/VBG would be a rational tool to evaluate for the presence of hypercapnic encephalopathy.
- In a patient with asthma or COPD who is sedated in order to tolerate BiPAP, periodic ABG/VBG may be needed to differentiate between medication-induced sedation versus hypercapnic encephalopathy (further discussion of this here: 📖).
A-a gradient evaluation
- Evaluation of the A-a gradient may occasionally be useful to diagnose occult hypercapnia, anxiety-related hyperventilation, or vocal cord dysfunction.
- These are niche applications that are discussed above. 📖
Be very careful about over-interpreting differences in serial ABG values.
random variation in PaO2 over time
- Among a group of 129 stable ICU patients, back-to-back ABGs drawn from arterial catheters demonstrated large differences in PaO2 values (figure below).(25621691) Other studies have demonstrated similar variability.(6407807, 8020270)
- PaO2 differences in serial ABGs will often merely reflect random variation (rather than reflecting true changes in the patient's trajectory).
random variation in PaCO2 and pH over time
- Among clinically stable patients, pH and pCO2 demonstrate random variation with the following 95% confidence intervals:(6407807, 8020270)
- pH +/- 0.03
- pCO2 +/- 5 mm
- It's unclear exactly how large differences in pH or pCO2 should be, in order to reflect a physiologically relevant change in the patient's condition. However, for any difference to be meaningful it should certainly be well above the random baseline variation in the test.
- Pregnancy induces mild, chronic hyperventilation. As pregnancy advances, the normal values may become roughly:
- PaCO2 of ~28-32 mm.
- Bicarbonate of ~18-21 mEq/L.
- pH of ~7.40-7.47 (Murray 2022)
- Oxygenation:
- A-a gradient increases 5-10 mm above baseline, especially in a supine position.
- Supine A-a gradient reaches ~20 mm at term.
- PaO2 >70 mm is often targeted to establish adequate fetal oxygenation (in the absence of any rigorous evidence).
- ABG in asthma:
- PaCO2 >35 mm may indicate respiratory fatigue.
- PaCO2 >42 may be an indication to consider intubation.
- Diabetic ketoacidosis is a metabolic process. Thus, a chemistry panel alone provides all the information needed to treat DKA: 🌊
- Anion gap metabolic acidosis (AGMA) should be treated with insulin and fluid resuscitation.
- Non-anion gap metabolic acidosis (NAGMA) should be treated with exogenous bicarbonate.
- There is no reason to specifically obtain either an ABG or a VBG in DKA.(12896883) Checking a blood gas analysis is often nonproductive, since this has a tendency to lead to inappropriate bicarbonate administration (e.g., if the pH is <6.9, some providers feel compelled to administer IV bicarbonate).
- (Many hospitals utilize a portable blood gas meter to check chemistry panels. This makes sense, as a strategy to obtain expedited information about the patient's electrolytes.)
- Further discussion of the disutility of blood gas analysis in DKA: 🌊
Follow us on iTunes
To keep this page small and fast, questions & discussion about this post can be found on another page here.
- The practice of broadly obtaining ABGs in critically ill patients provides lots of noise, without much useful information.
- ABGs rarely provide useful diagnostic information regarding the cause of the patient's respiratory failure (useful primarily for panic disorder, vocal cord dysfunction, and hypercapnic coma).
Guide to emoji hyperlinks
- = Link to online calculator.
- = Link to Medscape monograph about a drug.
- = Link to IBCC section about a drug.
- = Link to IBCC section covering that topic.
- = Link to FOAMed site with related information.
- 📄 = Link to open-access journal article.
- = Link to supplemental media.
References
- 02356936 Tarpey J, Lawler PG. Iatrogenic anaemia? A survey of venesection in patients in the intensive therapy unit. Anaesthesia. 1990 May;45(5):396-8. doi: 10.1111/j.1365-2044.1990.tb14785.x [PubMed]
- 06407807 Thorson SH, Marini JJ, Pierson DJ, Hudson LD. Variability of arterial blood gas values in stable patients in the ICU. Chest. 1983 Jul;84(1):14-8. doi: 10.1378/chest.84.1.14 [PubMed]
- 07632205 Stein PD, Goldhaber SZ, Henry JW. Alveolar-arterial oxygen gradient in the assessment of acute pulmonary embolism. Chest. 1995 Jan;107(1):139-43. doi: 10.1378/chest.107.1.139 [PubMed]
- 08020270 Sasse SA, Chen PA, Mahutte CK. Variability of arterial blood gas values over time in stable medical ICU patients. Chest. 1994 Jul;106(1):187-93. doi: 10.1378/chest.106.1.187 [PubMed]
- 12896883 Ma OJ, Rush MD, Godfrey MM, Gaddis G. Arterial blood gas results rarely influence emergency physician management of patients with suspected diabetic ketoacidosis. Acad Emerg Med. 2003 Aug;10(8):836-41. doi: 10.1111/j.1553-2712.2003.tb00625.x. PMID: 12896883. [PubMed]
- 21663600 Burri E, Potocki M, Drexler B, Schuetz P, Mebazaa A, Ahlfeld U, Balmelli C, Heinisch C, Noveanu M, Breidthardt T, Schaub N, Reichlin T, Mueller C. Value of arterial blood gas analysis in patients with acute dyspnea: an observational study. Crit Care. 2011;15(3):R145. doi: 10.1186/cc10268 [PubMed]
- 25621691 Mallat J, Lazkani A, Lemyze M, Pepy F, Meddour M, Gasan G, Temime J, Vangrunderbeeck N, Tronchon L, Thevenin D. Repeatability of blood gas parameters, PCO2 gap, and PCO2 gap to arterial-to-venous oxygen content difference in critically ill adult patients. Medicine (Baltimore). 2015 Jan;94(3):e415. doi: 10.1097/MD.0000000000000415 [PubMed]
- 26179876 Jubran A. Pulse oximetry. Crit Care. 2015 Jul 16;19(1):272. doi: 10.1186/s13054-015-0984-8 [PubMed]
- 27183375 Tusman G, Bohm SH, Suarez-Sipmann F. Advanced Uses of Pulse Oximetry for Monitoring Mechanically Ventilated Patients. Anesth Analg. 2017 Jan;124(1):62-71. doi: 10.1213/ANE.0000000000001283 [PubMed]
- 33326721 Sjoding MW, Dickson RP, Iwashyna TJ, Gay SE, Valley TS. Racial Bias in Pulse Oximetry Measurement. N Engl J Med. 2020 Dec 17;383(25):2477-2478. doi: 10.1056/NEJMc2029240 [PubMed]
- 34592317 Valbuena VSM, Barbaro RP, Claar D, Valley TS, Dickson RP, Gay SE, Sjoding MW, Iwashyna TJ. Racial Bias in Pulse Oximetry Measurement Among Patients About to Undergo Extracorporeal Membrane Oxygenation in 2019-2020: A Retrospective Cohort Study. Chest. 2022 Apr;161(4):971-978. doi: 10.1016/j.chest.2021.09.025 [PubMed]
- 34660615 Elshal MM, Hasanin AM, Mostafa M, Gamal RM. Plethysmographic Peripheral Perfusion Index: Could It Be a New Vital Sign? Front Med (Lausanne). 2021 Oct 1;8:651909. doi: 10.3389/fmed.2021.651909 [PubMed]
- 34687923 Coutrot M, Dudoignon E, Joachim J, Gayat E, Vallée F, Dépret F. Perfusion index: Physical principles, physiological meanings and clinical implications in anaesthesia and critical care. Anaesth Crit Care Pain Med. 2021 Dec;40(6):100964. doi: 10.1016/j.accpm.2021.100964 [PubMed]
- 34730820 Wong AI, Charpignon M, Kim H, Josef C, de Hond AAH, Fojas JJ, Tabaie A, Liu X, Mireles-Cabodevila E, Carvalho L, Kamaleswaran R, Madushani RWMA, Adhikari L, Holder AL, Steyerberg EW, Buchman TG, Lough ME, Celi LA. Analysis of Discrepancies Between Pulse Oximetry and Arterial Oxygen Saturation Measurements by Race and Ethnicity and Association With Organ Dysfunction and Mortality. JAMA Netw Open. 2021 Nov 1;4(11):e2131674. doi: 10.1001/jamanetworkopen.2021.31674 [PubMed]
- 35100193 Henry NR, Hanson AC, Schulte PJ, Warner NS, Manento MN, Weister TJ, Warner MA. Disparities in Hypoxemia Detection by Pulse Oximetry Across Self-Identified Racial Groups and Associations With Clinical Outcomes. Crit Care Med. 2022 Feb 1;50(2):204-211. doi: 10.1097/CCM.0000000000005394 [PubMed]
- 35231085 Burnett GW, Stannard B, Wax DB, Lin HM, Pyram-Vincent C, DeMaria S, Levin MA. Self-reported Race/Ethnicity and Intraoperative Occult Hypoxemia: A Retrospective Cohort Study. Anesthesiology. 2022 May 1;136(5):688-696. doi: 10.1097/ALN.0000000000004153 [PubMed]
- 35639368 Fawzy A, Wu TD, Wang K, Robinson ML, Farha J, Bradke A, Golden SH, Xu Y, Garibaldi BT. Racial and Ethnic Discrepancy in Pulse Oximetry and Delayed Identification of Treatment Eligibility Among Patients With COVID-19. JAMA Intern Med. 2022 Jul 1;182(7):730-738. doi: 10.1001/jamainternmed.2022.1906 [PubMed]
- 36049490 Wick KD, Matthay MA, Ware LB. Pulse oximetry for the diagnosis and management of acute respiratory distress syndrome. Lancet Respir Med. 2022 Nov;10(11):1086-1098. doi: 10.1016/S2213-2600(22)00058-3 [PubMed]