Quantitative evaluation of a combined pulse oximetry and end-tidal CO2 monitor

Interest in combined noninvasive respiratory monitors is increasing. The combination of a pulse oximeter and an end-tidal CO2 monitor in one unit reduces space requirements and, potentially, production costs. One such unit (POET, Criticare Systems, Inc.) was tested for accuracy in 15 healthy nonsmoking adult volunteers. Each patient was subjected to inspired O2 fractions of 1.0, 0.21, 0.12, 0.10, and 0.08, then to CO2 concentrations of 0, 3, 6, and 7.5%. Pulse oximetry sensors were placed on the index finger (SpfO2) and the forehead (SphO2). For each oxygen concentration tested, a 30-second steady state of SpfO2 and SphO2 was established before arterial blood was drawn for SaO2 determination. A sampling line for end-tidal carbon dioxide tension (PETCO2) was placed just distal to the one-way exhalation valve. Samples of arterial blood were drawn and analyzed for PaCO2 after PETCO2 had been stable for 30 seconds for each concentration of carbon dioxide. Linear regression was used for data analysis. The regression equations were: SpfO2 (y) vs. SaO2 (x), y = 0.93x + 6.70, r = 0.98 (p less than 0.001); SphO2 (y) vs. SaO2 (x), y = 0.85x + 13.76, r = 0.95 (p less than 0.001); PETCO2 (y) vs. PaCO2 (x), y = 1.08x - 2.66, r = 0.95 (p less than 0.001) (when subjects were not hyperventilating). The authors conclude that this combined pulse oximeter and end-tidal CO2 monitor can accurately refect SaO2 and PaCO2 in clinically useful ranges.