How do changes in arterial carbon dioxide tension (PaCO₂) or the venous‑arterial CO₂ gap (Pv‑aCO₂) predict whether a patient is fluid‑responsive with respect to cardiac output after a fluid challenge?

Medical Advisory BoardAll articles are reviewed for accuracy by our Medical Advisory Board
Educational purpose only • Exercise caution as content is pending human review
Article Review Status
Submitted
Under Review
Approved

Last updated: September 15, 2026 • View editorial policy

Have a follow-up question?

Our Medical A.I. is used by practicing medical doctors at top research institutions around the world. Ask any follow up question and get world-class guideline-backed answers instantly.

Ask Question

Changes in CO₂-Derived Variables to Predict Fluid Responsiveness

Changes in CO₂-derived variables—specifically a decrease in the venous-arterial CO₂ gap (Pv-aCO₂) and an increase in central venous oxygen saturation (ScvO₂)—can serve as surrogate markers to predict fluid responsiveness when direct cardiac output measurement is unavailable, though they should be used alongside other tissue perfusion markers rather than in isolation. 1

How to Use CO₂-Derived Variables

The 2025 ESICM guidelines explicitly recommend that the effectiveness of a fluid bolus in improving tissue perfusion should be evaluated by considering changes in variables such as capillary refill time, skin mottling, oxygen saturation, pCO₂-derived variables, and lactate 1. This positions CO₂ variables as part of a multiparameter assessment rather than standalone predictors.

Key Physiologic Principle

When cardiac output increases after fluid administration:

  • Pv-aCO₂ decreases (improved tissue perfusion washes out accumulated CO₂)
  • ScvO₂ increases (better oxygen delivery relative to consumption)
  • These changes reflect improved tissue perfusion and oxygen metabolism 2, 3

Evidence-Based Thresholds and Performance

For Fluid Challenge Assessment (After Fluid Given)

ΔPv-aCO₂ (change in CO₂ gap):

  • Cutoff: ≤ -2 mmHg decrease predicts fluid responsiveness
  • Diagnostic accuracy: AUC 0.83 (95% CI 0.77-0.88) 2
  • Positive predictive value: 90% 2
  • Negative predictive value: 58% 2
  • Gray zone: -2 to 0 mmHg (includes 22% of patients) 2

ΔScvO₂ (change in central venous saturation):

  • Cutoff: ≥ 3.4-5% increase predicts fluid responsiveness
  • Diagnostic accuracy: AUC 0.76-0.80 2, 3
  • Positive predictive value: 86-88% 2, 3
  • Gray zone: -3 to 4.6% (includes 51.7% of patients) 3

For Passive Leg Raising Test (Before Fluid Given)

Recent 2026 data shows that PLR-induced changes in Pv-aCO₂ can predict preload responsiveness without actually giving fluid:

  • ΔPv-aCO₂ during PLR: AUC 0.89 (95% CI 0.74-1.0) 4
  • ΔPv-aCO₂/C(a-cv)O₂ ratio during PLR: AUC 0.85 (95% CI 0.70-1.0) 4
  • PLR-induced decrease in Pv-aCO₂ was independently associated with preload responsiveness (OR 0.48) 4

Critical Clinical Context: When These Variables Work Best

Baseline Values Matter Significantly

High baseline Pv-aCO₂ (>6 mmHg):

  • Indicates inadequate cardiac output for tissue perfusion 5, 6
  • In this group, both baseline Pv-aCO₂ and ΔPv-aCO₂ have excellent predictive value (AUC 0.84-0.85) 5
  • 68.8% of these patients respond to fluid challenge 5
  • Positive predictive value approaches 100% when baseline >8 mmHg 5

Low baseline Pv-aCO₂ (≤6 mmHg):

  • Neither baseline nor change in Pv-aCO₂ reliably predicts fluid responsiveness 5
  • Only 45.8% respond to fluid challenge 5
  • These patients likely have adequate cardiac output already

Low baseline ScvO₂ (<70%):

  • ΔScvO₂ shows better accuracy (AUC 0.84) in this subgroup 3
  • Suggests oxygen delivery-consumption mismatch that may improve with fluids

VO₂ Responsiveness Context

In patients who are NOT VO₂ responsive (oxygen consumption doesn't increase with fluid):

  • ΔScvO₂ performs better (AUC 0.91) 3
  • ΔPv-aCO₂ also improves (AUC 0.78) 3
  • This suggests these markers detect cardiac output changes even when tissue oxygen extraction is impaired (common in sepsis)

Practical Algorithm for Clinical Use

Step 1: Assess Baseline Values

  • Measure baseline Pv-aCO₂ and ScvO₂ before fluid challenge
  • If Pv-aCO₂ >6 mmHg: High likelihood of fluid responsiveness; proceed with confidence
  • If Pv-aCO₂ ≤6 mmHg: Use dynamic predictors (PPV, SVV, PLR) instead 1

Step 2: Perform Fluid Challenge or PLR

  • Give 200-500 mL over 5-10 minutes 1
  • OR perform PLR test if avoiding actual fluid administration 4

Step 3: Reassess CO₂ Variables

  • Responders show:
    • Pv-aCO₂ decrease ≥2 mmHg
    • ScvO₂ increase ≥3.4-5%
    • Improved clinical perfusion markers
  • Non-responders show:
    • Minimal or no change in CO₂ variables
    • No improvement in tissue perfusion

Step 4: Integrate with Direct CO Measurement When Available

  • CO₂ variables correlate inversely with cardiac output (r = -0.63 for Pv-aCO₂) 6
  • Pv-aCO₂/C(a-cv)O₂ ratio correlates with DO₂/VO₂ (r = -0.71) 6
  • However, pulse pressure changes correlate poorly with CO changes and should not be used alone 1
  • Direct CO measurement remains the gold standard when available 1

Critical Pitfalls and Limitations

Major Caveats

  1. Not a substitute for direct CO monitoring: The guidelines emphasize that CO₂ variables are surrogate markers to use when CO measurement is unavailable 1

  2. Large gray zones: 22-51% of patients fall into indeterminate zones where these variables cannot reliably classify fluid responsiveness 2, 3

  3. Baseline dependency: Predictive value is substantially better when baseline Pv-aCO₂ is elevated (>6 mmHg) 5

  4. Microcirculatory dysfunction: In sepsis, even when CO increases, oxygen consumption may not improve due to microcirculatory abnormalities 1. CO₂ variables may not capture this dissociation.

  5. Not for predicting responsiveness: These variables assess the effect after fluid is given, not predict responsiveness beforehand (except with PLR) 2, 3. For prediction, use dynamic variables like PPV or SVV in appropriate patients 1

When NOT to Rely on CO₂ Variables Alone

  • Baseline Pv-aCO₂ ≤6 mmHg 5
  • When direct CO monitoring is available 1
  • In patients with significant lung disease (affects CO₂ elimination)
  • When values fall in the gray zone 2, 3

Correlation with Outcomes

Persistently elevated Pv-aCO₂ (>6 mmHg) at 24 hours post-resuscitation associates with:

  • Higher ICU mortality (OR 5.27) 7
  • Increased major complications 7, 8
  • Longer ICU stay and ventilation time 7
  • Higher SOFA scores 8

This reinforces that normalizing CO₂ gap should be a resuscitation target, not just a monitoring tool 7, 8.

Bottom Line for Practice

Use CO₂-derived variables as complementary markers of tissue perfusion response to fluid challenge, particularly when baseline Pv-aCO₂ is elevated (>6 mmHg) and direct cardiac output monitoring is unavailable. A decrease in Pv-aCO₂ ≥2 mmHg or increase in ScvO₂ ≥3.4% after fluid administration suggests effective cardiac output augmentation, but always integrate these findings with other clinical perfusion markers (lactate, capillary refill, urine output) and avoid repeating fluid boluses when these variables fail to improve 1, 2.

References

8
Research

Central Venous-arterial Carbon Dioxide Difference (pCO2 gap) as a Guide to Predict Complications in Postoperative High-risk Surgical Patients: A Prospective Observational Study.

Indian journal of critical care medicine : peer-reviewed, official publication of Indian Society of Critical Care Medicine, 2026

Related Questions

What monitor changes indicate a positive response after a 4 mL/kg mini‑fluid challenge (or passive leg raise) in a mechanically ventilated child?
What does a low arterial PCO2 indicate, and how should it be evaluated and managed?
How do I manage a mechanically ventilated patient with low arterial PCO₂?
What is the interpretation of a venous PCO₂ of 62 mm Hg with a venous pH of 7.46, and what should be the next clinical step?
Can rifabutin, amoxicillin, or pantoprazole cause Stevens-Johnson syndrome, and what is the likelihood?
What clinical, laboratory, and imaging criteria define a Crohn's disease flare?
Which of the following statements about metabolic alkalosis is NOT true?
What are the recommended antibiotic dosing regimens for common pediatric infections such as acute otitis media, streptococcal pharyngitis, skin infections, community‑acquired pneumonia, bacterial sinusitis, and severe infections requiring antipseudomonal or MRSA coverage?
In a 24-year-old African American male who is highly conditioned and experiences exertional central chest squeezing pain lasting 2–5 minutes with mild lightheadedness, occurring four times per week and without a family history of premature cardiac death, should he be evaluated for possible cardiac ischemia?
In a patient with ferritin 43 µg/L and no anemia, when should iron supplementation be started?

Professional Medical Disclaimer

This information is intended for healthcare professionals. Any medical decision-making should rely on clinical judgment and independently verified information. The content provided herein does not replace professional discretion and should be considered supplementary to established clinical guidelines. Healthcare providers should verify all information against primary literature and current practice standards before application in patient care. Dr.Oracle assumes no liability for clinical decisions based on this content.