Medical Tests

Arterial blood gases

Arterial blood gases

Last literature review version 19.3: Fri Sep 30 00:00:00 GMT 2011 | This topic last updated: Mon Jul 12 00:00:00 GMT 2010 (More)
Arthur C Theodore, MD
Section Editor
Scott Manaker, MD, PhD
Deputy Editor
Kevin C Wilson, MD

INTRODUCTION — An arterial blood gas (ABG) is a test that measures the arterial oxygen tension (PaO2), carbon dioxide tension (PaCO2), and acidity (pH). In addition, arterial oxyhemoglobin saturation (SaO2) can be determined. Such information is vital when caring for patients with critical illness or respiratory disease. As a result, the ABG is one of the most common tests performed on patients in intensive care units (ICUs).

The sites, technique, and complications of arterial sampling are discussed here. Transport and analysis of the arterial blood are also reviewed. Finally, pulse oximetry plus transcutaneous carbon dioxide measurement (an alternative method of obtaining similar information) is discussed.

ARTERIAL SAMPLING — Arterial blood is required for an ABG. It can be obtained by percutaneous needle puncture or from an indwelling arterial catheter.

Needle puncture — Percutaneous needle puncture refers to the withdrawal of arterial blood via a needle stick. It needs to be repeated every time an ABG is performed, since an indwelling catheter is not inserted.

Site selection — The initial step in percutaneous needle puncture is locating a palpable artery. Common sites include the radial, femoral, brachial, dorsalis pedis, or axillary artery. There is no evidence that any site is superior to the others. However, the radial artery is used most often because it is accessible, easily positioned, and more comfortable for the patient than the alternative sites.

The radial artery is best palpated between the distal radius and the tendon of the flexor carpi radialis when the wrist is extended (figure 1 and figure 2). To get the wrist into this position, the arm should be positioned on an armboard with the palm facing upward and a large roll of gauze should be placed between the wrist and the armboard in a position that extends the wrist. Taping the forearm and palm to the armboard helps maintain the position.

The brachial artery is best palpated medial to the biceps tendon in the antecubital fossa, when the arm is extended and the palm is facing up (figure 3). The needle should be inserted just above the elbow crease (figure 4).

The femoral artery is best palpated just below the midpoint of the inguinal ligament, when the lower extremity is extended (figure 5). The needle should be inserted at a 90 degree angle just below the inguinal ligament (figure 6).

The dorsalis pedis artery is best palpated lateral to the extensor hallucis longus tendon (figure 7). It receives collateral flow from the lateral plantar artery through an arch similar to that in the hand, which should be checked prior to percutaneous needle puncture. This is done by occluding the dorsalis pedis artery and compressing the nailbed of the great toe. Color should return to the nailbed rapidly when pressure on the toe is released (figure 8).

The axillary artery is best palpated in the axilla, when the arm is abducted and externally rotated (figure 9). There is good collateral flow to the arm through the thyrocervical trunk and subscapular artery; thus, the risk of ischemic complications to the arm is low. The needle should be inserted as high into the axilla as possible (figure 10).

Collateral circulation — Patients undergoing radial or dorsalis pedis artery puncture should have the collateral flow to those vessels evaluated prior to puncture.

The Allen test or modified Allen test can be performed in patients undergoing radial artery puncture. These are bedside tests that demonstrate collateral flow through the superficial palmar arch [1]. Their purpose is to identify patients who have impaired collateral circulation in the hand and, therefore, may be at increased risk of an ischemic complication.

To perform the modified Allen test, the patient’s hand is initially held high with the fist clenched and both the radial and ulnar arteries compressed (figure 11). This allows the blood to drain from the hand. The hand is then lowered, the fist is opened, and pressure is released from the ulnar artery. Color should return to the hand within six seconds, indicating that the ulnar artery is patent and the superficial palmar arch is intact. The test is considered abnormal if ten seconds or more elapses before color returns to the hand (figure 12).

The Allen test (from which the modified Allen test evolved) is performed identically, except these steps are executed twice: once with release of pressure from the ulnar artery and once with release of pressure from the radial artery.

For patients undergoing dorsalis pedis artery puncture, the dorsalis pedis artery can be occluded, followed by compression of the nailbed of the great toe and assessment of the rapidity with which color returns to the nailbed after pressure is released from the great toe.

Technique — Once a palpable artery has been located, blood is withdrawn using the following steps.

  • The planned puncture site should be sterilely prepped.
  • Local analgesia prior to arterial puncture should be considered, since it appears to prevent pain without adversely impacting the success of the procedure [2]. This was illustrated by a trial that randomly assigned 101 patients undergoing arterial puncture to receive 2 percent lidocaine, normal saline, or no agent prior to the procedure [3]. The lidocaine decreased pain without increasing the difficulty of the procedure (ie, the number of attempts), compared to the other groups.
  • The seal of a heparinized syringe should be broken by pulling its plunger. The plunger can then be pushed back into the syringe, leaving a small empty volume (eg, less than 1 mL) in the syringe. A small needle (eg, 22 to 25 gauge) should then be attached to the syringe.
  • Using one hand to gently palpate the artery and the other to manipulate the syringe and needle, the artery should be punctured with the needle at a 30 to 45 degree angle relative to the skin. The syringe will fill on its own (ie, pulling the plunger is unnecessary). Approximately 2 to 3 mL of blood should be removed.
  • After withdrawing a sufficient volume of blood, the needle should be removed and pressure applied to the puncture site for five to ten minutes to achieve hemostasis.
SEE MORE:  Is the Biopsy Still the Gold Standard for Diagnosing Celiac Disease?

Complications — Complications due to percutaneous needle puncture are rare. They include persistent bleeding, bruising, and injury to the blood vessel. Circulation distal to the puncture site may also be impaired following percutaneous needle puncture, presumably due to thrombosis at the puncture site.

Indwelling catheters — Arterial blood can also be obtained via an indwelling arterial catheter. Indwelling catheters provide continuous access to arterial blood, which is helpful when frequent blood gases are needed (eg, respiratory failure).

Indwelling arterial catheters should be preferentially placed in an artery that has collateral flow and allows easy maintenance of aseptic care. Options include the radial, dorsalis pedis, femoral, axillary, or brachial artery. Insertion of an arterial catheter is described separately. (See “Arterial catheterization techniques for invasive monitoring”.)

Complications of indwelling arterial catheters include local and systemic infection, bleeding, hematoma, bruising, and vascular complications such as blood vessel injury, pseudoaneurysm, thromboembolism, and vasospasm. The frequency of these complications is related to the insertion technique, duration of catheterization, and site.

SPECIMEN CARE — Regardless of the method used to withdraw the arterial blood, several things should be considered prior to sending the specimen to the laboratory:

  • Gas diffusion through the plastic syringe is a potential source of error. However, it appears that the clinical significance of the error is minimal if the sample is placed on ice and analyzed within 15 minutes [4-7]. Using a glass syringe will also prevent this error.
  • The heparin that is added to the syringe as an anticoagulant can decrease in the pH if acidic heparin is used. It can also dilute the PaCO2, resulting in a falsely low value [4,8]. Thus, the amount of heparin solution should be minimized and at least 2 mL of blood should be obtained.
  • Air bubbles that exceed 1 to 2 percent of the blood volume can cause a falsely high PaO2 and a falsely low PaC02 [9]. The magnitude of this error depends upon the difference in gas tensions between blood and air, the exposure surface area (which is increased by agitation), and the time from specimen collection to analysis. The clinical significance of this error can be decreased by gently removing the bubbles without agitation and analyzing the sample as soon as possible [5,10].

TRANSPORT — The arterial blood should be placed on ice during transport to the lab and then analyzed as quickly as possible. This reduces oxygen consumption by leukocytes, which can cause a factitiously low PaO2 [11]. This effect is most pronounced in patients whose leukocytosis is profound. In addition, it reduces the likelihood that error due to gas diffusion through the plastic syringe or air bubbles will be clinically significant.

ANALYSIS — Analysis of arterial blood is usually performed by automated blood gas analyzers, which automatically transport the specimen to electrochemical sensors to measure pH, PaCO2, and PaO2:

  • The PaCO2 is measured using a chemical reaction that consumes CO2 and produces a hydrogen ion, which is sensed as a change in pH [9]
  • The PaO2 is measured using oxidation-reduction reactions that generate measurable electric currents [9]

In addition, automated blood gas analyzers rinse the system, calibrate the sensors, and report the results. Rigorous quality control by the laboratory is essential for accurate results.

Arterial blood gas measurements are effected by temperature. Specifically, pH increases and both PaO2 and PaCO2 decrease as temperature declines (table 1) [12,13]. Modern automated blood gas analyzers can report the pH, PaO2, and PaCO2 at either 37ºC (the temperature at which the values are measured by the blood gas analyzer) or at the patient’s body temperature. Most centers report the values of pH, PCO2, and PO2 at 37ºC, even if the patient’s body temperature is different. However, this practice is controversial [12-15].

INTERPRETATION — ABGs provide information about oxygenation, ventilation, and acid-base balance. The interpretation of ABG results is reviewed separately. (See “Oxygenation and mechanisms of hypoxemia” and “Oxygen delivery and consumption” and “Simple and mixed acid-base disorders”.)

ALTERNATIVE APPROACHES — ABGs are performed frequently in critically ill patients because they are the gold standard measure of a patient’s oxygenation (measured as PaO2) and ventilation (measured as PaCO2 and pH). However, they are invasive and require repeated collection of arterial blood. Thus, alternative methods to monitor oxygenation and ventilation that do not require arterial blood are desirable, especially if they are noninvasive or minimally invasive.

Assess oxygenation — Pulse oximetry reliably evaluates oxygenation noninvasively, but does not provide information about ventilation. It is discussed in detail separately. (See “Pulse oximetry”.)

Assess ventilation and acid-base status — Venous blood gases (VBGs) refer to blood gases measured using venous blood instead of arterial blood:

  • When drawn from a central venous catheter, VBGs can be useful for evaluating a patient’s ventilation and acid-base status because they estimate the arterial pH and arterial carbon dioxide tension (PaCO2). The central venous pH is usually 0.03 to 0.05 less than the arterial pH, while the central venous carbon dioxide tension is usually 4 to 5 mmHg higher than the PaCO2 [16,17]. In addition, the serum bicarbonate level can be measured from the venous blood.
  • When drawn from a peripheral vein, VBGs have been shown to correlate with ABGs in patients with uremic acidosis [18], diabetic ketoacidosis [18,19], COPD exacerbations [20], and general acute illness [21,22]. While most of the evidence is from patients in an emergency department setting, correlation has also been demonstrated in adult intensive care unit patients with respiratory failure on mechanical ventilation [23]. However, the correlation between peripheral VBGs and ABGs is not universal, as one study found poor correlation among hypotensive patients in a pediatric ICU [24].
SEE MORE:  Arterial catheterization techniques for invasive monitoring

The major advantage of VBGs is convenience, since most patients in an ICU have a central venous catheter from which venous blood can be drawn and patients in an emergency department can be spared the discomfort of an arterial puncture. However, VBGs have important limitations. First, they do not provide information about a patient’s oxygenation. Second, care must be taken to only use VBGs in patients who have conditions in which VBGs reliably estimate ABGs. Third, central venous and arterial pH and carbon dioxide tension may not correlate in the most severely ill patients, particularly those with cardiogenic shock [25,26] or in conditions with hypotension [24]. For this reason, periodic correlation between a VBG and an ABG is necessary for each patient in whom the venous blood gas is being used to monitor ventilatory status. Studies comparing venous to arterial blood gasses in critically ill patients are ongoing.

Measurement of end tidal CO2 (PetCO2) is another way of non-invasively measuring the PaCO2. This technique requires a closed system of gas collection, either with a tight fitting mask or a ventilator circuit. A sample of expired gas is analyzed by infrared or mass spectrometry, and then displayed as a numerical value or a graph. The PetCO2 is usually within 1 mm of the PaCO2 in healthy adults, but is far less accurate in critically ill adults [27]. Thus, routine use of PetCO2 exists primarily in newborn ICUs, operating rooms, and emergency departments, to provide early warning of tube complications. PetCO2 has not been validated by clinical trials in adult ICUs.

Assess oxygenation, ventilation, and acid-base status — The combination of pulse oximetry and repeated VBGs is a reasonable alternative to ABGs for monitoring both oxygenation and ventilation, if the VBGs and ABGs have been shown to correlate in the specific patient.

Systems that combine pulse oximetry and transcutaneous pCO2 (ptcCO2) can evaluate both oxygenation and evaluation and, therefore, have been studied as a potential substitute for conventional ABGs. Such combination systems usually have a heating element that raises the skin temperature to 42 to 45ºC to increase local perfusion, an electrode to measure ptcCO2, and a light emitter and sensor to measure arterial oxyhemoglobin saturation [28].

Older studies suggested that ptcCO2 measurements are accurate in neonates, but not critically ill adults because of poor peripheral perfusion (peripheral artery disease, hypotension, vasopressors). Devices have since improved and several more recent observational studies suggest that such systems are accurate in most critically ill patients, although accuracy diminishes when the PaCO2 is greater than 56 mmHg [29-33]. Despite the promising observational data, clinical trials are necessary before combined pulse oximetry and ptcCO2 monitoring can be recommended as routine care.

Such combination systems have limitations. They may be difficult to keep calibrated, may be difficult to mount in a way that prevents air trapping, and may take up to an hour to sufficiently warm the skin [33]. In addition, the devices must be attached to an ear, which may be difficult in agitated patients or in those who had neurosurgery. Given the limitations of noninvasive monitoring, any persistent or unexpected change in SaO2 or ptcCO2 should be confirmed with an ABG.


  • An arterial blood gas (ABG) is a test that measures the arterial oxygen tension (PaO2), carbon dioxide tension (PaCO2), and acidity (pH). In addition, arterial oxyhemoglobin saturation (SaO2) can be determined. (See ‘Introduction’ above.)
  • Percutaneous needle puncture is one method of obtaining the arterial blood necessary for an ABG (see ‘Arterial sampling’ above):
  • Common sites include the radial, femoral, brachial, dorsalis pedis, or axillary artery. There is no evidence that any site is superior to the others. However, the radial artery is used most often because it is accessible, easily positioned, and more comfortable for the patient than the alternative sites. (See ‘Site selection’ above.)
  • For patients undergoing radial or dorsalis pedis artery puncture, we suggest evaluating the collateral flow to those vessels prior to puncture (Grade 2C). (See ‘Collateral circulation’ above.)
  • Once the target artery has been identified, the planned puncture site should be sterilely prepped. The artery should be punctured with a small needle and syringe, 2 to 3 mL of blood should be withdrawn, and then needle should be removed. Finally, pressure should be applied to the puncture site for five to ten minutes. (See ‘Technique’ above.)
  • We recommend the administration of local analgesia prior to arterial puncture (Grade 1B). Local analgesia prevents pain without adversely impacting the success of the procedure. (See ‘Technique’ above.)
  • Complications due to percutaneous needle puncture are rare, but include persistent bleeding, bruising, and injury to the blood vessel. Circulation distal to the puncture site may also be impaired. (See ‘Complications’ above.)
  • Alternatively, an indwelling arterial catheter can be used to obtain arterial blood for an ABG. An indwelling catheter provides continuous access to arterial blood. Arterial catheterization is discussed in greater detail separately. (See ‘Indwelling catheters’ above and “Arterial catheterization techniques for invasive monitoring”.)
  • Regardless of the method used to withdraw the arterial blood, the amount of heparin solution should be minimized, at least 2 mL of blood should be obtained, air bubbles should be removed, and the specimen should immediately be placed on ice and analyzed as quickly as possible. (See ‘Specimen care’ above and ‘Transport’ above.)
  • Alternative approaches to assessing oxygenation or ventilation include pulse oximetry, venous blood gases, end tidal carbon dioxide, and percutaneous carbon dioxide, alone or in combination. (See ‘Alternative approaches’ above.)
SEE MORE:  Abdominal Ultrasound
Use of UpToDate is subject to the Subscription and License Agreement.


  1. Kaye W. Invasive monitoring techniques. In: Textbook of Advanced Cardiac Life Support, American Heart Association, Dallas.
  2. Guidelines for the measurement of respiratory function. Recommendations of the British Thoracic Society and the Association of Respiratory Technicians and Physiologists. Respir Med 1994; 88:165.
  3. Lightowler JV, Elliott MW. Local anaesthetic infiltration prior to arterial puncture for blood gas analysis: a survey of current practice and a randomised double blind placebo controlled trial. J R Coll Physicians Lond 1997; 31:645.
  4. Bageant, RA. Variations in arterial blood gas measurements due to sampling techniques. Respir Care 1975; 20:565.
  5. Harsten A, Berg B, Inerot S, Muth L. Importance of correct handling of samples for the results of blood gas analysis. Acta Anaesthesiol Scand 1988; 32:365.
  6. Evers W, Racz GB, Levy AA. A comparative study of plastic (polypropylene) and glass syringes in blood-gas analysis. Anesth Analg 1972; 51:92.
  7. Smeenk FW, Janssen JD, Arends BJ, et al. Effects of four different methods of sampling arterial blood and storage time on gas tensions and shunt calculation in the 100% oxygen test. Eur Respir J 1997; 10:910.
  8. Hansen JE, Simmons DH. A systematic error in the determination of blood PCO2. Am Rev Respir Dis 1977; 115:1061.
  9. Williams AJ. ABC of oxygen: assessing and interpreting arterial blood gases and acid-base balance. BMJ 1998; 317:1213.
  10. Mueller RG, Lang GE, Beam JM. Bubbles in samples for blood gas determinations. A potential source of error. Am J Clin Pathol 1976; 65:242.
  11. Hess CE, Nichols AB, Hunt WB, Suratt PM. Pseudohypoxemia secondary to leukemia and thrombocytosis. N Engl J Med 1979; 301:361.
  12. Shapiro BA. Temperature correction of blood gas values. Respir Care Clin N Am 1995; 1:69.
  13. Hansen JE. Arterial blood gases. Clin Chest Med 1989; 10:227.
  14. Bacher A. Effects of body temperature on blood gases. Intensive Care Med 2005; 31:24.
  15. Ream AK, Reitz BA, Silverberg G. Temperature correction of PCO2 and pH in estimating acid-base status: an example of the emperor’s new clothes? Anesthesiology 1982; 56:41.
  16. Malinoski DJ, Todd SR, Slone S, et al. Correlation of central venous and arterial blood gas measurements in mechanically ventilated trauma patients. Arch Surg 2005; 140:1122.
  17. Walkey AJ, Farber HW, O’Donnell C, et al. The accuracy of the central venous blood gas for acid-base monitoring. J Intensive Care Med 2010; 25:104.
  18. Gokel Y, Paydas S, Koseoglu Z, et al. Comparison of blood gas and acid-base measurements in arterial and venous blood samples in patients with uremic acidosis and diabetic ketoacidosis in the emergency room. Am J Nephrol 2000; 20:319.
  19. Brandenburg MA, Dire DJ. Comparison of arterial and venous blood gas values in the initial emergency department evaluation of patients with diabetic ketoacidosis. Ann Emerg Med 1998; 31:459.
  20. Ak A, Ogun CO, Bayir A, et al. Prediction of arterial blood gas values from venous blood gas values in patients with acute exacerbation of chronic obstructive pulmonary disease. Tohoku J Exp Med 2006; 210:285.
  21. Gennis PR, Skovron ML, Aronson ST, Gallagher EJ. The usefulness of peripheral venous blood in estimating acid-base status in acutely ill patients. Ann Emerg Med 1985; 14:845.
  22. Malatesha G, Singh NK, Bharija A, et al. Comparison of arterial and venous pH, bicarbonate, PCO2 and PO2 in initial emergency department assessment. Emerg Med J 2007; 24:569.
  23. Chu YC, Chen CZ, Lee CH, et al. Prediction of arterial blood gas values from venous blood gas values in patients with acute respiratory failure receiving mechanical ventilation. J Formos Med Assoc 2003; 102:539.
  24. Yildizdaş D, Yapicioğlu H, Yilmaz HL, Sertdemir Y. Correlation of simultaneously obtained capillary, venous, and arterial blood gases of patients in a paediatric intensive care unit. Arch Dis Child 2004; 89:176.
  25. Adrogué HJ, Rashad MN, Gorin AB, et al. Assessing acid-base status in circulatory failure. Differences between arterial and central venous blood. N Engl J Med 1989; 320:1312.
  26. Weil MH, Rackow EC, Trevino R, et al. Difference in acid-base state between venous and arterial blood during cardiopulmonary resuscitation. N Engl J Med 1986; 315:153.
  27. Tobin MJ. Respiratory monitoring in the intensive care unit. Am Rev Respir Dis 1988; 138:1625.
  28. Severinghaus JW, Astrup P, Murray JF. Blood gas analysis and critical care medicine. Am J Respir Crit Care Med 1998; 157:S114.
  29. Bendjelid K, Schütz N, Stotz M, et al. Transcutaneous PCO2 monitoring in critically ill adults: clinical evaluation of a new sensor. Crit Care Med 2005; 33:2203.
  30. Senn O, Clarenbach CF, Kaplan V, et al. Monitoring carbon dioxide tension and arterial oxygen saturation by a single earlobe sensor in patients with critical illness or sleep apnea. Chest 2005; 128:1291.
  31. Vivien B, Marmion F, Roche S, et al. An evaluation of transcutaneous carbon dioxide partial pressure monitoring during apnea testing in brain-dead patients. Anesthesiology 2006; 104:701.
  32. Hurley RA, Fisher R, Taber KH. Sudden onset panic: epileptic aura or panic disorder? J Neuropsychiatry Clin Neurosci 2006; 18:436.
  33. Cuvelier A, Grigoriu B, Molano LC, Muir JF. Limitations of transcutaneous carbon dioxide measurements for assessing long-term mechanical ventilation. Chest 2005; 127:1744. INT

March 2016
« Feb   Apr »

tittygram INT