There are now several ways to measure carbon dioxide and oxygen in blood without drawing a blood sample. The two most popular methods at present are transcutaneous electrode systems and pulse oximetry. Both systems can provide continuous readings.

The transcutaneous systems use PCO2 and PO2 electrodes similar to those of standard arterial blood gas analysis applied directly to the skin over a gel sealant. Skin has capillaries close to the surface, and the tissues are permeable to some extent for carbon dioxide and oxygen. The apparatus heats the skin to 44°C to produce arterialized blood, thereby dilating the capillaries and increasing oxygen loss. The electrode sensors detect the carbon dioxide and oxygen diffusing from the capillaries. The apparatus must be moved at least every 4-6 hours in adults and 2-4 hours in infants to prevent thermal burns. The apparatus must be calibrated with a standard arterial blood gas sample obtained by arterial puncture each time the apparatus is positioned due to variability from differences in fat content (which interferes with gas diffusion) and skin thickness. Patient edema, hypothermia, or poor tissue perfusion (shock or vasoconstriction) interfere to varying degrees with accurate measurements.

Pulse oximetry measures hemoglobin oxygen saturation (percentage of hemoglobin structurally capable of binding oxygen that is saturated with oxygen) rather than oxygen tension (PO2). The method uses two light beams, one red and the other infrared, which are passed through tissue that contains arterial blood. Opposite to the light emitters are light detectors. The light detectors perform two tasks. First, they recognize and analyze arterial blood exclusively by differentiating those areas that have pulsation, and therefore changes in light transmission, from nonvascular tissue and nonarterial vascular components. Then oxygen saturation is measured in the pulsating vessels using the fact that changes in oxygen content have a significant effect on absorption of red light. The amount of red light absorption (transmission) is compared to that of the infrared light, which is affected much less. This system does not have to be calibrated by arterial puncture blood and does not have to be moved frequently. The instrument is accurate between saturation levels of about 70%-100%. When PO2 is above 100 mm Hg, hemoglobin is usually 100% saturated, reaching the upper limit of the oximeter. Below 70% saturation, accuracy becomes less, but trends in saturation change can be recognized. Carboxyhemoglobin can interfere with measurement. Since the instrument measures only oxygenation, acid-base abnormalities must be detected or investigated by some other method.

Abnormal results from either transcutaneous monitors or pulse oximeters must be confirmed by arterial puncture blood gas measurement. The pulse oximeter is usually attached to a toe in infants, to a finger in adults, and to the nose in obese adults.

Noninvasive continuous oxygen monitors are especially useful during anesthesia since most serious problems involve episodes of hypoxia; in premature or sick neonates and infants; in patients on ventilators; and in intensive care unit (ICU) patients or other unstable seriously ill adults.