Transcutaneous monitor approximates PaCO2 but not PaO2 in anesthetized rabbits

Abstract

ObjectiveTo compare the accuracy of transcutaneous (tc) to arterial partial pressure of carbon dioxide (PaCO2) and partial pressure of oxygen (PaO2) in anesthetized rabbits. Study designProspective, randomized, experimental study. AnimalsEight healthy adult female New Zealand white rabbits weighing 4.05 ± 0.30 kg. MethodsIsoflurane anesthetized rabbits received six treatments in random order; PaCO2 < 35, 35-45, and >45 mmHg and PaO2 < 80, 100-200, >200 mmHg. Arterial and transcutaneous measurements were taken after 15 minutes of stabilization at each condition. Linear regression, correlation and Bland-Altman analysis were performed to compare PtcCO2 to PaCO2 and PtcO2 to PaO2. ResultsOver a range of measured PaCO2 values from 21 to 67 mmHg (n = 24) mean bias for PtcCO2 was -1 mmHg and the 95% limits of agreement were -7 to 5 mmHg. The correlation between PtcCO2 and PaCO2 was strong with R2 value of 0.9454. Over the entire range of measured PaO2 values (46-508 mmHg) mean bias for PtcO2 was -61 mmHg and the 95% limits of agreement were -226 to 104 mmHg. Correlation was poor with R2 = 0.5969. Comparing PtcO2 to PaO2 over a narrower range [PaO2 < 150 mmHg (n = 13)] improved the correlation, with an R2 value of 0.8518, mean bias of -7 mmHg and 95% limits of agreement from -33 to 19 mmHg. Conclusions and clinical relevanceIn healthy anesthetized rabbits, PtcCO2 closely approximated PaCO2. In contrast PtcO2 underestimated PaO2, particularly at high values. The PtcCO2 sensor may be a useful noninvasive way to assess adequacy of ventilation in anesthetized rabbits.