Continuous non-invasive PCO2 monitoring in weaning patients: Transcutaneous is advantageous over end-tidal PCO2.

Abstract

Continuous partial pressure of carbon dioxide (PCO2 ) assessment is essential for the success of mechanical ventilation (MV). Non-invasive end-tidal PCO2 (PetCO2 ) and transcutaneous PCO2 (PtcCO2 ) measurements serve as alternatives to the gold standard arterial PCO2 (PaCO2 ) method, but their eligibility in critical care is unclear. The present study therefore performed methodological comparisons of PaCO2 versus PetCO2 and PtcCO2 , respectively, in weaning patients receiving invasive MV via tracheal cannulas. PetCO2 and PtcCO2 were recorded continuously, while PaCO2 was analysed at baseline, and after 30 and 60 min. Using the Bland-Altman analysis, a clinically acceptable range was defined as a mean difference of ±4 mm Hg between PaCO2 and non-invasive strategies. A total of 60 patients (COPD (n = 30) and non-COPD (n = 30)) completed the protocol. Mean PCO2 values were 42.4 ± 8.6 mm Hg (PaCO2 ), 36.5 ± 7.5 mm Hg (PetCO2 ) and 41.7 ± 8.7 mm Hg (PtcCO2 ). Mean differences between PtcCO2 and PaCO2 were -0.7 ± 3.6 mm Hg (95% CI: -1.6/0.3 mm Hg; 95% limits of agreement: -7.8 to 6.4 mm Hg), and between PetCO2 and PaCO2 -5.9 ± 5.3 mm Hg (95% CI: -7.2/-4.5 mm Hg; 95% limits of agreement: -16.2 to 4.5 mm Hg). Underestimation of PaCO2 by PetCO2 was most pronounced in COPD patients. Our data therefore support PtcCO2 as a suitable means for monitoring PCO2 in patients undergoing invasive MV. This is in contrast to PetCO2 , which clearly underestimated PaCO2 , especially in patients with COPD.