Spot check analysis of gas exchange: invasive versus noninvasive methods

Abstract

Background: Correct measurement of PO₂ and PCO₂ is essential to establish appropriate therapy such as long-term oxygen therapy (LTOT) in patients suffering from respiratory failure.

Objectives: We aimed to compare common invasive and noninvasive methods for assessing blood gas components for spot check analysis.

Methods: Arterial (PaO₂, PaCO₂) and capillary blood gas (PCBGO2, PCBGCO2) measurements were taken consecutively in a randomized order and were compared with noninvasive measurements obtained from the transcutaneous monitoring of PO₂ and PCO₂ (PtcOv, PtcCO₂, sensor-temperature 44°C). Capillary samples were taken from both arterialized earlobes, where samples of right earlobes were defined as a reference value. Pain assessment of all measurements was evaluated by each subject using the 100-mm visual analogue scale.

Results: 83 patients and 17 healthy subjects were included. The mean difference between PaO₂ and PtcO₂ was 11.9 ± 15.0 mm Hg, with lower limits of agreement (LLA) of -17.4 mm Hg (95% confidence interval (CI) -22.5 to -12.3 mm Hg), and upper limits of agreement (ULA) of 41.1 mm Hg (95% CI 36.0-46.2 mm Hg). The comparison of PaO₂ with PCBGO2 showed a mean difference of 5.6 ± 7.2 mm Hg (LLA -11.0; ULA 19.6 mm Hg). The mean difference between PaCO₂ and PtcCO₂ was 1.1 ± 4.9 mm Hg (LLA -8.6; ULA 10.8 mm Hg) and that between PaCO₂ and PCBGCO₂ was 0.7 ± 2.0 mm Hg (LLA -3.3; ULA 4.8 mm Hg). The analysis of capillary blood gases (36.2 ± 22.3 mm) was rated as more painful than the analysis of arterial blood gases (26.1 ± 20.6 mm), while transcutaneous measurement was rated as the least painful method (1.9 ± 7.4 mm; all p < 0.0001).

Conclusions: The comparison of different methods for blood gas measurements showed substantial differences between capillary and arterial PO₂ and between transcutaneous and arterial PO₂. Therefore, arterial PO₂ analysis is the essential method evaluating indication for LTOT. Nevertheless, comparative analysis further indicated capillary PCO₂ as an adequate surrogate for arterial PCO₂.