Estimation of arterial carbon dioxide based on end-tidal gas pressure and oxygen saturation

Abstract

Arterial blood gas (ABG) analysis is the traditional method for measuring the partial pressure of carbon dioxide. In mechanically ventilated patients a continuous noninvasive monitoring of carbon dioxide would obviously be attractive. In the current study, we present a novel formula for noninvasive estimation of arterial carbon dioxide. Eighty-one datasets were collected from 19 anesthetized and mechanically ventilated pigs. Eleven animals were mechanically ventilated without interventions. In the remaining eight pigs the partial pressure of carbon dioxide was manipulated. The new formula (Formula 1) is PaCO2 = PETCO2 + k(PETO2 − PaO2) where PaO2 was calculated from the oxygen saturation. We tested the agreements of this novel formula and compared it to a traditional method using the baseline PaCO2 − ETCO2 gap added to subsequently measured, end-tidal carbon dioxide levels (Formula 2). The mean difference between PaCO2 and calculated carbon dioxide (Formula 1) was 0.16 kPa (±SE 1.17). The mean difference between PaCO2 and carbon dioxide with Formula 2 was 0.66 kPa (±SE 0.18). With a mixed linear model excluding cases with cardiorespiratory collapse, there was a significant difference between formulae (p < 0.001), as well as significant interaction between formulae and time (p < 0.001). In this preliminary animal study, this novel formula appears to have a reasonable agreement with PaCO2 values measured with ABG analysis, but needs further validation in human patients.