Transcutaneous Carbon Dioxide Monitoring More Accurately Detects Hypercapnia than End-Tidal Carbon Dioxide Monitoring during Non-Intubated Video-Assisted Thoracic Surgery: A Retrospective Cohort Study

Abstract

Transcutaneous carbon dioxide (PtcCO2) monitoring is known to be effective at estimating the arterial partial pressure of carbon dioxide (PaCO2) in patients with sedation-induced respiratory depression. We aimed to investigate the accuracy of PtcCO2 monitoring to measure PaCO2 and its sensitivity to detect hypercapnia (PaCO2 > 60 mmHg) compared to nasal end-tidal carbon dioxide (PetCO2) monitoring during non-intubated video-assisted thoracoscopic surgery (VATS). This retrospective study included patients undergoing non-intubated VATS from December 2019 to May 2021. Datasets of PetCO2, PtcCO2, and PaCO2 measured simultaneously were extracted from patient records. Overall, 111 datasets of CO2 monitoring during one-lung ventilation (OLV) were collected from 43 patients. PtcCO2 had higher sensitivity and predictive power for hypercapnia during OLV than PetCO2 (84.6% vs. 15.4%, p < 0.001; area under the receiver operating characteristic curve; 0.912 vs. 0.776, p = 0.002). Moreover, PtcCO2 was more in agreement with PaCO2 than PetCO2, indicated by a lower bias (bias ± standard deviation; -1.6 ± 6.5 mmHg vs. 14.3 ± 8.4 mmHg, p < 0.001) and narrower limit of agreement (-14.3-11.2 mmHg vs. -2.2-30.7 mmHg). These results suggest that concurrent PtcCO2 monitoring allows anesthesiologists to provide safer respiratory management for patients undergoing non-intubated VATS.