Arterial and transcutaneous variability and agreement between multiple successive measurements of carbon dioxide in patients with chronic obstructive lung disease

Purpose: This study evaluates agreement between carbon dioxide measured arterial (PaCO2) and transcutaneous (PtcCO(2)) over time, by repeated successive measures, taking into consideration the inherent variability of arterial measurements. Methods and results: 11 patients receiving LTOT, with severe to very severe COPD in a stable phase were studied. Repeated arterial blood samples were drawn and PtcCO(2) measured simultaneously at the ear lobe. Bland-Altman analysis was used to evaluate 95 % limits of agreement (LoA). 194 paired samples were analysed. Following correction for bias, the difference between PaCO2 and PtCO2 during dynamic conditions was 0.02 kPa and LoA 0.94 to-0.90 kPa while 29 % of PtCO2 measurements were outside the range of variability for arterial measurements. Conclusion: PtcCO(2) corrected for intra-patient bias provide reasonable description of PaCO2 values within but not outside steady state conditions. Our results suggest that PtcCO(2) is a valuable method for monitoring in chronic rather than acute conditions when bias can be removed. Superscript/Subscript Available