Why Pulse Oximeters Fail to Detect Carbon Monoxide Poisoning

On July 6, 2026, emergency medicine physician Dr. Erik Axene published a stark warning to prehospital care providers regarding winter emergency calls. He described a scenario where a family complains of severe headaches and confusion, yet their vital signs appear perfectly stable.

"Everyone's SpO2 reads 98 to 100 percent," Axene wrote, detailing the clinical presentation of carbon monoxide exposure. "This is carbon monoxide poisoning until proven otherwise, and your pulse oximeter is the least useful tool you have right now."

Carbon monoxide is notoriously difficult to detect in the field because it is colorless, odorless, and mimics dozens of other conditions. The most dangerous aspect of the toxidrome is its ability to completely blind the primary respiratory monitoring tool used by medical professionals.

A patient actively dying from cellular asphyxiation will consistently display a normal or near-normal oxygen saturation reading on a standard hospital monitor. The physiological mechanism behind this failure lies in the fundamental physics of how light interacts with human blood.[4]

How standard pulse oximeters see light

Since the American Society of Anesthesiologists adopted pulse oximetry as a minimum monitoring standard in 1986, the technology has relied on a simple optical principle. The device aims to provide a continuous, noninvasive assessment of functional arterial oxygen saturation.

A standard pulse oximeter utilizes an electronic processor and a pair of small light-emitting diodes facing a photodiode. These components are typically placed across a translucent part of the patient's body, such as a fingertip or an earlobe.[3]

The device emits light at two specific frequencies: a red wavelength at 660 nanometers and an infrared wavelength at 940 nanometers. Absorption of light at these two distinct wavelengths differs significantly between blood loaded with oxygen and blood lacking oxygen.[3]

Oxygenated hemoglobin absorbs more infrared light at 940 nanometers and allows more red light to pass through. Conversely, deoxygenated hemoglobin allows more infrared light to pass through while absorbing a greater amount of the 660-nanometer red light.[1][3]

The device relies on the Beer-Lambert law, which states that the absorption of light passing through a substance is directly proportional to the concentration of the attenuating species. During pulsatile blood flow, the expansion of blood vessels alters the hemoglobin concentration.[1]

By subtracting the non-pulsatile elements of tissues and venous blood, the microprocessor isolates the arterial component. It then calculates a modulation ratio, known as R, by dividing the pulsatile red light absorbance by the pulsatile infrared light absorbance.[1]

This calculated R ratio is then compared against a standardized calibration curve stored in the device's database. The empirical algorithm translates the optical ratio into the SpO2 percentage displayed on the clinical monitor, assuming only two hemoglobin states exist.[1]

The 660-nanometer blind spot

The fatal flaw in this two-wavelength system emerges when a third hemoglobin species enters the bloodstream. Normal human blood contains four hemoglobin species, but standard pulse oximeters are physiologically incapable of distinguishing dyshemoglobins from normal oxygen-carrying molecules.

Carbon monoxide binds to the iron atoms in hemoglobin with an affinity approximately 200 to 250 times stronger than that of oxygen. When inhaled, the toxic gas rapidly displaces oxygen, forming a stable compound known as carboxyhemoglobin.[3]

Carboxyhemoglobin possesses a unique light absorption profile that confounds the standard two-wavelength measurement. At the 660-nanometer red wavelength, the light absorbance curve of carboxyhemoglobin overlaps almost perfectly with the absorbance curve of normal oxyhemoglobin.[3]

"The pulse oximeter will see the reflected light for carboxyhemoglobin and oxyhemoglobin similarly when measuring in the red light spectrum," notes a technical review by Storm Anesthesia. The device measures a combined reflection, entirely unable to separate the two molecules.

Furthermore, carboxyhemoglobin absorbs virtually no light in the 940-nanometer infrared spectrum. Because it does not contribute to the infrared absorbance, the denominator of the R ratio remains unchanged, leaving the device's microprocessor completely unaware of the toxic substitution.

Calculating the false normal

When the pulse oximeter computes the ratio between the 660-nanometer and 940-nanometer signals, the identical red absorbance and absent infrared absorbance of carboxyhemoglobin mimic the exact optical signature of pure oxyhemoglobin. The algorithm processes the data exactly as programmed.[3][4]

The resulting mathematical output forces the device to report a falsely elevated SpO2 reading. The monitor displays a number that reflects the total percentage of hemoglobin bound to any molecule, rather than the percentage bound specifically to oxygen.[3]

The magnitude of this error scales directly with the severity of the poisoning. A patient whose bloodstream contains 40 percent carboxyhemoglobin and only 60 percent oxyhemoglobin will routinely display an SpO2 reading of 98 to 99 percent.

"Their hemoglobin is bound to a molecule that will not offload to tissues," Axene explained. The monitor reassures the clinician that the blood is fully saturated, masking the reality that nearly half of the patient's oxygen-carrying capacity has been destroyed.

This optical illusion means a standard pulse oximeter will never detect carbon monoxide poisoning, regardless of how severe the hypoxia becomes. The patient will suffocate at the cellular level while the digital display insists their respiratory status is perfect.[4]

Hypoxia hidden in plain sight

The diagnostic challenge extends beyond noninvasive fingertip monitors. Standard arterial blood gas analysis is equally inadequate for diagnosing carbon monoxide toxicity, as the oxygen saturation reported in these tests represents dissolved oxygen in the plasma.[2]

Dissolved oxygen levels remain entirely unaffected by the concentration of carboxyhemoglobin inside the red blood cells. According to the Merck Manuals, relying on arterial blood gases or standard pulse oximetry will consistently fail to identify the life-threatening deficit.[2]

A normal hemoglobin count of 15 grams per deciliter can typically bind about 20 milliliters of oxygen per deciliter of blood. When carbon monoxide occupies those binding sites, the total oxygen content plummets, but the dissolved plasma oxygen remains normal.[3]

This physiological reality makes clinical suspicion the most vital diagnostic tool for first responders. Recognizing the environmental clues and the classic symptoms of headache, nausea, and confusion must override the falsely reassuring numbers on the vital sign monitor.

Measuring beyond two wavelengths

To accurately detect carboxyhemoglobin noninvasively, medical technology had to move beyond the two-wavelength limitation. Modern pulse co-oximeters utilize eight or more distinct wavelengths of light to differentiate between multiple hemoglobin species simultaneously.

By measuring light absorbance across a wider spectrum, these advanced devices can isolate the unique optical signatures of carboxyhemoglobin and methemoglobin. The microprocessor then calculates the specific concentration of each species, displaying the carbon monoxide level as an SpCO percentage.

Devices like the Rad-57 signal extraction pulse co-oximeter have been approved for clinical use to assess these specific percentages. However, studies published in Frontiers in Medicine suggest that even advanced noninvasive SpCO screening has limitations in accuracy.

The sensitivity of noninvasive co-oximetry is sometimes insufficient to serve as a definitive screening test in emergency departments. False positives and false negatives can still occur, complicating the triage process for patients with suspected exposure.

The ultimate diagnostic standard remains laboratory blood co-oximetry. This invasive test spectrophotometrically measures light transmission through a drawn venous or arterial blood sample at multiple discrete wavelengths, providing a highly accurate measurement of true oxygen content.[2]

Until multi-wavelength technology becomes universal in prehospital care, standard pulse oximeters will continue to operate with a fatal blind spot. Clinicians must remember that a 660-nanometer light beam cannot tell the difference between a breath of fresh air and a lethal dose of exhaust.[4]