Pulse Oximetry: A Necessary Imperfect Tool in Critical Care

Abstract

P ulse oximetry has evolved as an essential tool in the armament of a clinician. It noninvasively and immediately provides insights into the oxygenation of a patient. With the recent global pandemic, the pulse oximeter has expanded from a provider's tool to one that is readily accessible to the public and even on their wrists as a personal monitoring device. Although there are significant limitations in the use of pulse oximetry, especially by a layperson in the at-home setting, we argue that the pulse oximeter remains useful and vital in the care of critically ill patients.The pulse oximeter is far from a faultless device. To understand the concerns with pulse oximeters, it is important to review the basic mechanism of action. The original oximeters were developed in the 1930s to monitor oxygen levels in aviation pilots with advancements brought about during World War II with the first ear probe sensors (1). Ongoing improvements in the 1950s and 1960s led to the discovery by Aoyagi that monitoring variations in the arterial blood volume per pulse could produce a specific arterial signal of oxygenation via spectroscopy to assess pulsatile changes in the light transmitted, and hence pulse oximetry (2). This discovery minimized the absorption of light by the surrounding tissues and was further advanced in the 1980s to the first clinical pulse oximeters that were utilized in practice and are still in use today by measuring changes in light absorption of oxygenated and deoxygenated hemoglobin using two light wavelengths (red 660 nm and infrared 940 nm). Problems with pulse oximeters were noted early on, and these included inaccuracies in the waveform in patients with nail polish, shivering or motion artifacts, cardiac arrhythmias, poor distal perfusion, skin pigmentation, poor sensitivity in hypoxia, and with altered hemoglobins (3). A landmark study in 1993 of 20,802 surgical patients in Denmark randomized patients to monitoring via pulse oximetry versus standard of care at that time (clinical observation of hypoxia); the study noted a 19-fold increased rate of detection of hypoxemia in the oximetry group (p < 0.00001), along with a decrease in myocardial ischemia (4). This along with additional trials led to the adoption of pulse oximeters as standard of care in hospitalized patients.In general, in patients with pulse oximetry measured arterial oxygen saturation (Spo 2 ) greater than 90%, the difference between arterial oxygen saturation (Sao 2 ) is generally less than 2 points (5). Yet, pulse oximeters are less accurate in patients with increased melanin. One of the original studies evaluating this noted that a higher Spo 2 was required in Black patients to ensure appropriate arterial oxygen tension (Pao 2 ) when compared with White patients, along with a greater variability in Spo 2 and Sao 2 in Black patients (6). In 2005, a study tested three different commercial pulse oximeters in patients with different pigmentations and noted that across all pulse oximeters, "pigment-related bias increased