Comparison of arterial blood gas with continuous intra-arterial and transcutaneous PO2 sensors in adult critically ill patients.

We compared the partial pressure of oxygen directly via a continuous intra-arterial probe (PiaO2) and indirectly using a transcutaneous device (PtcO2) with simultaneously obtained arterial blood PaO2. The PiaO2 values were measured using a bipolar oxygen sensor placed through an 18-ga arterial catheter. The PtcO2 values were measured using a transcutaneous O2-CO2 sensor placed on the abdomen. Seven critically ill, hemodynamically stable, ventilator-dependent adult patients were studied. Measurements were obtained at varying concentrations (0.25 to 1.0) of inspired oxygen after a 10-min stabilization. A total of 78 simultaneous values were obtained; by linear regression: PiaO2 = 0.91 PaO2 + 1.39 (r = .98, standard errors of the estimate [SEE] = 18.6); PtcO2 = 0.39 PaO2 + 36.2 (r = .89, SEE = 14.1). To assess these instruments as trend monitors, we compared the changes in simultaneous PaO2, PiaO2, and PtcO2 values; by linear regression: delta PiaO2 = 0.90 delta PaO2 + 3.88 (r = .96, SEE = 27.7); delta PtcO2 = 0.43 delta PaO2 + 5.6 (r = .94, SEE = 15.2). We conclude that, although these instruments correlate highly with the PaO2, the SEE was substantial and therefore may limit their clinical reliability in adults. Any acute or clinically significant change in PiaO2 or PtcO2 should be confirmed with a blood gas PaO2.