The Pulse Oximetry Gap in Carbon Monoxide Poisoning


The oxygen saturation, determined by the bedside pulse oximeter, has essentially become a fifth vital sign in the initial assessment of ED patients. Although many clinicians have begun to rely on this tool to make critical decisions regarding oxygenation and ventilation, it is important to remember that the oxygen saturation determined by pulse oximetry is only an approximation of the true saturation. This estimate is based largely on the assumption that only normal hemoglobin is present. Thus, the concept of a "pulse oximetry gap," representing the difference between the true oxygen saturation (measured by co-oximetry) and the estimated oxygen saturation (determined by pulse oximetry), has arisen.

Bozeman and colleagues help illustrate this point in patients with carbon monoxide poisoning. In a retrospective chart review, Bozeman et al identified 124 patients with carbon monoxide poisoning who had their oxygen saturations documented by both pulse oximetry and co-oximetry. As carboxyhemoglobin levels increased, the oxygen saturation measured by co-oximetry decreased linearly with excellent correlation (R = -0.97). However, the oxygen saturation determined by pulse oximetry remained above 90% in all cases. Thus, as the carboxyhemoglobin rose and the true oxygen saturation fell, the saturation gap between the co-oximeter and the pulse oximeter values widened. (Bozeman WP, et al. Confirmation of the pulse oximetry gap in carbon monoxide poisoning. Ann Emerg Med 1997; 30:608-611.)


This finding can be explained based on the physics of pulse oximetry. The pulse oximeter estimates oxygen saturation by comparing the absorbance of light at two wavelengths. Thus, the solution of the two simultaneous equations generated by the pulse oximeter is derived from only two species of hemoglobin (assumed to be oxyhemoglobin and deoxyhemoglobin). When another hemoglobin species is present (such as carboxyhemoglobin or methemoglobin), its absorbance is perceived as a fraction of oxyhemoglobin and deoxyhemoglobin.1 Since the absorbance of carboxyhemoglobin at the two wavelengths studied is closest to that of oxyhemoglobin, the pulse oximeter is fooled into thinking that carboxyhemoglobin is oxyhemoglobin, and saturation values stay near normal. A more complex relationship exists for methemoglobin.2

Like any other tool, the pulse oximeter has limitations. When assessing a patient’s respiratory status, we want to know oxygenation (the PO2), ventilation (the PCO2), acid-base balance (the pH) and oxygen carrying capacity (a combination of the oxygen saturation, the PO2, and the hemoglobin content). The pulse oximeter estimates the oxygen saturation, which allows us to predict the oxygen carrying capacity—assuming that a normal amount of normal hemoglobin is present. The pulse oximeter provides misleading information in cases of profound anemia or dyshemoglobinemia. Furthermore, even when hemoglobin species and content are normal, pulse oximetry provides no information about ventilation or acid-base balance. Given this information, pulse oximetry should be used with caution to corroborate clinical findings, and pulse oximetry should be supported by formal arterial blood gas analysis and co-oximetry testing when either dyshemoglobinemia is suspected or when the clinical condition is complex.


1. Tremper KK, Barker SJ. Using pulse oximetry when dyshemoglobin levels are high. J Crit Illness 1988; 3:103-107.

2. Barker SJ, et al. Effects of methemoglobin on pulse oximetry and mixed venous oximetry. Anesthesiology 1989;70:112-117.

Pulse oximetry may provide misleading information regarding oxygen saturation in cases of:

a. carboxyhemoglobinemia.

b. methemoglobinemia.

c. profound anemia.

d. all of the above.