Factors influencing transcutaneous oxygen and carbon dioxide measurements in adult intensive care patients.

Transcutaneous PO2 (PtcO2) is suggested to reflect tissue oxygenation in intensive care patients, whereas transcutaneous PCO2 (PtcCO2) is advocated as a noninvasive method for assessing PaCO2. In 24 critically ill adult patients (mean Apache II score 14.2, SD 4.7) we investigated the impact of variables that are commonly thought to determine PtcO2 and PtcCO2 measurements. A linear correlation was found between PtcO2 and PaO2 (r = 0.6; p less than or equal to 0.0001) and between PtcO2 and mean arterial blood pressure (MAP; r = 0.42; p less than or equal to 0.003). Cardiac index (CI) correlated with tc-index (PtcO2/PaO2; r = 0.31; p less than or equal to 0.03). There was no relationship between PtcO2 and hemoglobin concentration (Hb) and the position of the oxygen dissociation curve (ODC). Stepwise multiple regression analysis demonstrated a significant influence of PaO2 and MAP on PtcO2. The contribution of CI, Hb and the ODC was not significant. Only 40% of the variability of a single PtcO2 measurement could be explained by PaO2 and MAP. A significant linear correlation was demonstrated between PtcCO2 and PaCO2 (r = 0.76; p less than or equal to 0.0001) but not between PtcCO2 and CI, MAP and arterial base excess (BEa). Stepwise multiple regression analysis revealed an influence of PaCO2 and of CI on PtcCO2; 66% of the variability of a single PtcCO2-value could be explained by PaCO2 and CI. Our data demonstrate that transcutaneous derived gas tensions result from complex interaction between hemodynamic, respiratory and local factors, which can hardly be defined in ICU-patients.