Arterial vs. Venous vs. Capillary Blood Gas Analysis

Author: Andrew Bland, MD, MBA, MS
Medical Associates Department of Nephrology | University of Illinois College of Medicine at Peoria | University of Dubuque Physician Assistant Program

Executive Summary

1. Introduction

The arterial blood gas (ABG) has historically been the reference standard for acid-base and respiratory assessment, but its routine use has been challenged on the basis of procedural pain, arterial injury risk, and a growing body of evidence supporting venous blood gas (VBG) accuracy for many clinical endpoints. Selecting the appropriate modality requires understanding not only the correlation statistics but also the specific physiologic information each sampling site provides and the clinical circumstances in which each is appropriate.

This review provides a structured comparison of ABG and VBG across the domains of pH, PCO2, PO2, bicarbonate, lactate, and electrolytes; discusses the pain-hyperventilation artifact unique to ABG sampling; and addresses the specific use of capillary blood gas in neonates.

2. Physiologic Basis of Arteriovenous Differences

Blood gas values differ between arterial and venous compartments because peripheral tissues extract oxygen and release carbon dioxide during metabolic activity. The magnitude of these differences depends on cardiac output (low-flow states exaggerate arteriovenous differences), peripheral oxygen extraction fraction, and metabolic rate.

Under normal hemodynamic conditions:

3. pH: The Strongest Case for VBG

Agreement with arterial pH is the most robustly validated VBG parameter. The landmark 2014 systematic review and meta-analysis by Byrne et al. (Respirology, PMID 24383789) pooled 22 studies across emergency and critical care populations and reported a mean bias of −0.035 (venous lower than arterial) with 95% limits of agreement of approximately −0.12 to +0.05. The correlation coefficient between venous and arterial pH was 0.97.

From a clinical decision standpoint:

• A venous pH >7.35 reliably excludes clinically significant acidemia
• A venous pH <7.30 is a reliable indicator of arterial acidemia and warrants further assessment
• The uncertain zone — venous pH 7.30–7.35 — is where ABG confirmation adds most value

4. PCO2: The Critical Limitation of VBG for Respiratory Assessment

While pH agreement between arterial and venous blood is clinically acceptable, PCO2 agreement is substantially worse. The Byrne meta-analysis reported a mean bias of +5.5 mmHg (venous higher) with 95% limits of agreement of approximately −4 to +15 mmHg. Kelly et al. (Emergency Medicine Journal 2002; PMID 12153457) reported similar findings with limits of agreement spanning nearly 20 mmHg in some populations.

The clinical implication: a venous PCO2 of 48 mmHg is compatible with an arterial PCO2 anywhere from approximately 38–58 mmHg — a range spanning normal ventilation to moderate hypoventilation. This uncertainty is clinically unacceptable when the question is:

• Is this COPD exacerbation causing hypercapnic respiratory failure?
• Is this post-procedure patient retaining CO2?
• Is NIV or intubation indicated based on ventilatory status?

5. The Pain and Hyperventilation Artifact in ABG Sampling

This is among the most underappreciated sources of error in clinical blood gas interpretation. The radial artery puncture is a painful procedure. Pain and anticipatory anxiety trigger activation of the sympathetic nervous system, producing acute hyperventilation in a substantial proportion of patients.

The physiologic consequence is a procedurally induced reduction in PCO2 of 3–8 mmHg and a corresponding rise in pH of 0.02–0.06 units. The effect occurs within seconds to minutes of needle insertion.

Clinical Scenarios Where This Artifact Matters Most

1. Evaluating early respiratory acidosis: A patient with early COPD exacerbation with arterial PCO2 of 50 mmHg may hyperventilate during the procedure and generate a sample with PCO2 of 44 mmHg and pH of 7.40 — appearing normal. The underlying ventilatory impairment is masked.
2. Assessing compensatory response: If a metabolic acidosis patient is hyperventilating both physiologically and procedurally, the measured PCO2 will appear lower than the true value, suggesting a more vigorous compensatory response than actually exists.
3. Diagnosing primary respiratory alkalosis: Genuine primary respiratory alkalosis cannot be reliably distinguished from artifactual hyperventilation during sampling without clinical correlation.

Mitigation Strategies

• Adequate local anesthesia with buffered lidocaine
• Patient reassurance
• Use of pre-placed arterial catheters when available
• Awareness that a "normal" ABG in a patient expected to have respiratory compromise may warrant repeat sampling under better conditions

6. PO2 and Oxygenation Assessment

PO2 cannot be reliably derived from VBG. The venous PO2 reflects peripheral tissue oxygen extraction and is influenced by cardiac output, hemoglobin concentration, oxygen consumption, and regional perfusion heterogeneity. A venous PO2 of 35 mmHg is normal and does not imply hypoxemia; a venous PO2 of 60 mmHg may indicate inadequate peripheral oxygen extraction in sepsis rather than adequate alveolar oxygenation.

The alveolar-arterial (A-a) gradient — the fundamental metric for distinguishing intrinsic pulmonary from extrapulmonary hypoxemia — cannot be calculated without arterial PO2. For any clinical question involving oxygenation assessment, pulmonary embolism evaluation, diffusion impairment, or shunt quantification, ABG is mandatory.

Pulse oximetry provides continuous, non-invasive SpO2 monitoring and is appropriate for oxygenation surveillance in most stable patients, though it cannot substitute for ABG in quantifying ventilation (PCO2) or acid-base status.

7. Bicarbonate, Lactate, and Electrolytes

Bicarbonate (HCO3⁻)

Venous bicarbonate is typically 1–2 mEq/L higher than arterial bicarbonate, reflecting CO2 hydration in peripheral venous blood. This difference is sufficiently small that venous HCO3⁻ is clinically equivalent to arterial HCO3⁻ for diagnosing metabolic alkalosis, chronic respiratory compensation, and mixed disorders.

Lactate

Peripheral venous lactate is clinically acceptable for screening, with multiple studies demonstrating that a venous lactate <2.0 mmol/L reliably excludes clinically significant hyperlactatemia in the hemodynamically stable patient. When lactate is elevated on VBG, arterial measurement may be needed for precise characterization, though the absolute management threshold is typically the venous level in most sepsis protocols.

Electrolytes

Potassium, sodium, chloride, and ionized calcium obtained from VBG are reliable for clinical decision-making. Ionized calcium is particularly subject to alteration by pH and PCO2 perturbation, so ABG-level precision may be needed when precise calcium correction is required.

8. Site-Specific VBG Considerations

The venous site of sampling affects VBG reliability. Peripheral venous blood (antecubital vein) is the standard site for VBG and the population studied in most correlation analyses. Central venous blood from the superior vena cava or pulmonary artery provides a mixed venous sample (ScvO2 or SvO2) with different clinical utility — it reflects global venous oxygen saturation and is used for hemodynamic monitoring in sepsis, not for acid-base assessment.

9. Capillary Blood Gas in Neonates

In neonatal practice, arterial blood gas sampling from peripheral arteries or umbilical artery catheters is technically challenging, painful, and carries risks of arterial spasm, thrombosis, and infection. Capillary blood gas (CBG) obtained from a warmed heel via lancet provides an arterialized capillary sample that approximates arterial blood gas values when sampling technique is optimized.

Correlation with Arterial Values

Neonatal guidelines from the American Academy of Pediatrics specify that CBG PO2 should not be used to guide oxygen supplementation; SpO2 monitoring and, when necessary, arterial sampling via umbilical artery catheter remain the standards for neonatal oxygenation assessment.

Sampling Technique Determinants of CBG Quality

• Heel warming to 40–42°C for at least 3 minutes promotes arterialization
• The first drop of blood is discarded to avoid tissue fluid dilution
• Free-flowing blood collection; squeezing the heel introduces venous admixture
• Immediate capping and analysis to prevent CO2 loss

10. Clinical Decision Framework: When to Use Each Modality

11. Summary

VBG has appropriately expanded its role in acute care medicine as the evidence base for pH and bicarbonate correlation has matured. The single most important limitation of VBG is PCO2 precision — a ±10 mmHg uncertainty is clinically tolerable for metabolic acid-base assessment but unacceptable when ventilatory management decisions depend on accurate CO2 quantification. The pain-hyperventilation artifact is an underappreciated source of false reassurance in ABG interpretation that clinicians should explicitly consider when ABG results conflict with the clinical presentation. In neonatal practice, CBG provides a practical arterialized sample for pH and CO2 assessment, with the important caveat that PO2 remains unreliable.

References

1. Byrne AL, Bennett M, Chatterji R, Symons R, Pace NL, Thomas PS. Peripheral venous and arterial blood gas analysis in adults: are they comparable? A systematic review and meta-analysis. Respirology. 2014;19(2):168–175. PMID: 24383789
2. Malinoski DJ, Todd SR, Slone S, Mullins RJ, Schreiber MA. Correlation of central venous and arterial blood gas measurements in mechanically ventilated trauma patients. Arch Surg. 2005;140(11):1122–1125. PMID: 16301441
3. Kelly AM, McAlpine R, Kyle E. Agreement between bicarbonate measured on arterial and venous blood gases. Emerg Med Australas. 2004;16(5-6):407–409. PMID: 15537406
4. Aaron SD, Vandemheen KL, Naftel SA, Lewis MJ, Rodger MA. Topical tetracaine prior to arterial puncture: a randomized, placebo-controlled clinical trial. Respir Med. 2003;97(11):1195–1199. PMID: 14620571
5. American Academy of Pediatrics Committee on Fetus and Newborn. Levels of neonatal care. Pediatrics. 2012;130(3):587–597. PMID: 22926180
6. Treger R, Pirouz S, Kamangar N, Corry D. Agreement between central venous and arterial blood gas measurements in the intensive care unit. Clin J Am Soc Nephrol. 2010;5(3):390–394. PMID: 20089492
7. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801–810. PMID: 26903338

© 2025 Andrew Bland, MD, MBA, MS — Urine Nephrology Now