Is continuous in-line blood gas monitoring reliable during cardiopulmonary bypass when PaO2 and PaCO2 are calculated rather than measured?

TL;DR

This study evaluates the reliability of calculated blood gas monitoring during heart surgery, finding significant errors that vary with patient weight and temperature.

Contribution

The study identifies patterns of error in calculated PaO2 and PaCO2 during CPB and proposes a formula to adjust FiO2 based on patient weight.

Findings

Calculated PaO2 errors exceed acceptable targets during initial blood gas series and rewarming, linked to patient weight.

Calculated PaCO2 shows upward drift during rewarming, correlated with temperature changes.

A formula was derived to adjust FiO2 based on patient weight to achieve target PaO2 with the FX05 oxygenator.

Abstract

Background: The accuracy and precision of continuous in-line blood gas monitoring (CILBGM) are crucial for optimal blood gas management during cardiopulmonary bypass (CPB) and improved patient outcomes. CILBGM devices, such as the CDI 500/550 system, measure PaO2 and PaCO2, and B-Capta measures PaO2 through direct contact with arterial blood. However, the Quantum perfusion system with Quantum Ventilation2 (Quantum System) does not measure but calculates PaO2 and PaCO2 using several non-invasive sensors and proprietary formulas. We have observed that the calculated in-line PaO2 and PaCO2 values from Quantum System are frequently significantly higher than those obtained from iSTAT, a point-of-care blood analyzer, exceeding acceptable targets. Methods: We conducted a retrospective study involving 81 patients who underwent cardiac surgery using the Quantum System with its own CILBGM and the FX05 oxygenator. The aim was to identify the degree, timing, and possible patterns of error of the calculated in-line PaO2 and PaCO2. Results: Our study showed that the errors of calculated in-line PaO2 exceed the acceptable target at the 1st blood gas series and during the rewarming and rewarmed periods, correlating with patient weight. The calculated in-line PaCO2 exhibited an upward drift during the rewarming period, correlating with the temperature gradient rather than patient weight. Based on several correlations identified, we derived a formula to predict FiO2 based on patient weight, which would achieve the target PaO2 at the 1st blood gas series when using the FX05 oxygenator. Conclusion: We identified when and how the errors in calculating in-line PaO2 and PaCO2 occurred and developed several recommendations to minimize significant deviations from actual PaO2 and PaCO2 during CPB. Our results suggest that achieving acceptable PaO2 and PaCO2 calculations throughout CPB using a single universal formula for each, embedded in the Quantum System, is challenging due to the variety of oxygenators available, different patient sizes, and changing conditions during CPB.

Figures