Kidney Week

Abstract: FR-PO269

Accurate Estimation of Arterial pH from Venous pCO2 and Venous CO2: An Easy Solution to an Old Problem

Session Information

• Fluid and Electrolytes: Clinical
  October 26, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Fluid and Electrolytes

• 902 Fluid and Electrolytes: Clinical

Authors

• Ba aqeel, Sheeba Habeeb, Northwestern University Feinberg School of Medicine, Chicago, Illinois, United States

Sheeba Habeeb Ba aqeel,

• Batlle, Daniel, Northwestern University Feinberg School of Medicine, Chicago, Illinois, United States

Daniel Batlle,

Background

Plasma CO2 from venous blood (vTCO2 or vCO2) is often used as a surrogate for arterial bicarbonate (aHCO3). Arterial PCO2 (aPCO2) can be estimated from the venous PCO2 (vPCO2) using regression equations derived from correlative studies. If the pH is recalculated using vPCO2, the HCO3 (numerator in the Henderson Hasselbalch equation) is also impacted thereby limiting the accuracy of the corrected pH. We reasoned that with proper validations of vCO2 as a surrogate for aHCO3 and vPCO2 corrected to reflect aPCO2, the arterial pH (apH) can be readily estimated from venous blood and the venous pH will not be needed. Here we tested this concept.

Methods

We analyzed a de-identified database of venous and arterial blood gases performed in the clinical lab of Northwestern Medicine, Chicago, USA. ABG, VBG and electrolyte panel were examined in 113 samples obtained simultaneously. Additionally, concurrent data for arterial and venous PCO2 obtained from 3911 samples and collected from various hospital departments was used for a regression equation to calculate aPCO2 from vPCO2. Pearson correlations and Bland-Altman analysis were done to examine the limits of agreement.

Results

vCO2 was in close agreement with the aHCO3 (R²=0.93 P<0.001; 95% limits of agreement were 2.89 to -2.89 mmol/L, mean bias= -0.2 ±1.59). aPCO2 was estimated from vPCO2 to yield corrected PCO2 (cPCO2) using a regression equation, cPCO2=2.3+0.82*vPCO2. The apH was then calculated using this cPCO2 and vCO2 as follows: apH = 6.1 + log vCO2 / 0.03 x cPCO2. The corrected apH showed a close correlation and agreement with measured apH (R²=0.81 P<0.001 (fig1) and 95% limits of agreement 0.46 to -0.39, mean bias 0.04±4.4)

Conclusion

1)CO2 values in venous blood are in close agreement with aHCO3 calculated from the ABG. Accordingly, vCO2 can substitute for aHCO3 to recalculate the pH 2)The corrected vPCO2 and the measured vCO2 allow for an accurate calculation of apH which is close to measured apH. Moreover, this avoids altogether the measurement of venous pH which is not a good reflection of apH.