Pulse oximetry is a noninvasive method used to measure the amount of oxygen in the blood. Readings of peripheral oxygen saturation (SpO₂) are usually within 2% accuracy of the more precise but invasive measurement of arterial oxygen saturation (SaO₂) from arterial blood gas analysis. In 95% of cases, SpO₂ readings are within 4% accuracy of SaO₂.

A standard pulse oximeter uses two wavelengths of light to pass through tissue to a photodetector. It takes advantage of the pulsing flow of arterial blood to measure changes in light absorbance during a heartbeat. This allows the device to calculate the absorbance caused only by arterial blood, excluding absorbance from venous blood, skin, bone, muscle, fat, and, in many cases, nail polish. The two wavelengths measure the amounts of oxygenated and non-oxygenated hemoglobin. The ratio of these measurements is used to calculate the percentage of oxygenated hemoglobin. The most common method is transmissive pulse oximetry, where one side of a thin body part, such as a fingertip or earlobe, is illuminated, and the photodetector is placed on the opposite side. Fingertips and earlobes have relatively high blood flow compared to their size, which helps maintain warmth. However, this may not be true in hypothermic patients. Other useful sites include an infant’s foot or the cheek or tongue of an unconscious patient.

Reflectance pulse oximetry is a less common method that places the photodetector on the same surface as the light source. This method does not require thin body parts and can be used on areas like the forehead, chest, or feet. However, it has some limitations. Conditions such as vasodilation and pooling of venous blood in the head, which may occur during anesthesia with endotracheal intubation and mechanical ventilation or in patients in the Trendelenburg position, can cause mixed arterial and venous pulsations on the forehead. This may lead to incorrect SpO₂ readings.

Medical uses

A pulse oximeter is a medical tool that measures the amount of oxygen in a person’s blood without taking a blood sample. It also tracks changes in blood flow under the skin and creates a graph called a photoplethysmogram, which can help calculate other health details. This device can be part of a larger monitor that checks multiple health signs. Most monitors also show a person’s pulse rate. Small, battery-powered pulse oximeters are available for use at home or during travel.

Pulse oximetry is a simple way to check oxygen levels in the blood without causing pain. In contrast, checking blood oxygen levels through a blood sample must be done in a lab. Pulse oximeters are helpful in situations where a person’s oxygen levels might change, such as in hospitals, during surgery, in emergencies, or for pilots flying in unpressurized planes. They are also used to check if a person needs extra oxygen. However, a pulse oximeter cannot measure how much oxygen the body is using. For this, carbon dioxide levels must also be checked. A pulse oximeter might help find breathing problems, but it is less reliable when a person is using extra oxygen, because it works best when someone is breathing normal air.

Pulse oximeters are important in emergency care and for people with breathing or heart issues, such as those with chronic obstructive pulmonary disease (COPD) or sleep disorders like apnea. People with sleep apnea often have oxygen levels between 70% and 90% while trying to sleep. Portable pulse oximeters are useful for pilots flying at high altitudes, mountain climbers, and athletes who may have lower oxygen levels at high elevations or during physical activity. Some devices have software that records oxygen and pulse data, reminding users to check their levels.

New technology allows pulse oximeters to send data to hospital monitors without wires, keeping patient information connected to healthcare systems. For people with COVID-19, pulse oximeters help detect "silent hypoxia," where oxygen levels drop dangerously low even if the person feels fine. This can happen in hospitals or at home and may mean the person needs a ventilator.

Using a pulse oximeter for up to 8 hours is generally safe. However, long-term use can cause burns from the heat of the device, which can reach up to 43°C (105°F). Some devices may overheat due to electrical problems. People with sensitive skin, such as babies or the elderly, are at higher risk. Others at risk include those who cannot feel pain in the area where the device is placed, such as people with numb limbs, those who are unconscious, or those who cannot communicate. For high-risk patients, the device should be moved every hour, and for lower-risk patients, every 2–4 hours.

Limitations

Pulse oximetry only measures how much oxygen is attached to hemoglobin in the blood. It does not check how well the lungs are working or provide a full picture of how well the body is getting oxygen. It is not a replacement for blood gas tests done in a lab, which can measure things like blood pH, carbon dioxide levels, and bicarbonate concentration. The body uses oxygen in a way that can be measured by checking how much carbon dioxide is exhaled, but pulse oximetry does not show how much oxygen is actually in the blood. Most oxygen in the blood is carried by hemoglobin. In severe anemia, the blood has less hemoglobin, so even if hemoglobin is fully attached to oxygen, it cannot carry enough oxygen to the body.

Pulse oximetry also does not fully show if the body is getting enough oxygen through the blood. If there is not enough blood flow or not enough hemoglobin in the blood (like in anemia), the body’s tissues may not get enough oxygen even if the blood has high oxygen levels.

Pulse oximetry only measures the percentage of hemoglobin that is attached to oxygen. This can lead to incorrect readings when hemoglobin binds to other substances:

• Hemoglobin binds more strongly to carbon monoxide than to oxygen. In carbon monoxide poisoning, most hemoglobin may be attached to carbon monoxide instead of oxygen. A pulse oximeter would show that hemoglobin is bound, but the patient would still have low oxygen levels in the body.
• Cyanide poisoning can cause a high pulse oximetry reading because it reduces how much oxygen is used by the body. In this case, the reading is not wrong because the blood has high oxygen levels early in poisoning. The patient is not low on oxygen in the blood but may not be getting enough oxygen to the body’s cells.
• Methemoglobinemia often causes pulse oximetry readings around 85%.
• Chronic obstructive pulmonary disease (COPD), especially chronic bronchitis, may cause incorrect readings.

A noninvasive method that can measure dyshemoglobins (like carboxyhemoglobin and methemoglobin) continuously is the pulse CO-oximeter, developed in 2005 by Masimo. It uses more light wavelengths to measure these substances along with total hemoglobin.

Pulse oximeters are usually calibrated for healthy people, so they may not work well for very sick patients or premature babies. Incorrect readings can happen if the body part being monitored has poor blood flow (often due to coldness or from medicines that narrow blood vessels), if the sensor is not placed correctly, if the skin is very thick, or if there is movement (like shivering) during poor blood flow. To work correctly, the sensor should show a steady pulse or pulse wave. Some pulse oximeters are better at giving accurate readings during movement or low blood flow than others. Obesity, low blood pressure, and certain hemoglobin types can also reduce accuracy. Some home devices may not detect drops in blood oxygen levels well because they take measurements too slowly. Pulse oximeters are less accurate when readings are below 80%. Studies show that some devices may give less accurate readings for people with darker skin, leading to concerns about fairness in countries with diverse populations. Early studies from 1976 found that devices sometimes showed lower oxygen levels in people with darker skin than actual levels. Later studies found that while accuracy is good at higher oxygen levels, some devices may overestimate oxygen levels at lower levels, missing cases of low oxygen in the body. A study of thousands of patients found that Black patients were three times more likely than white patients to have low oxygen levels missed by pulse oximeters. Another study found that 28.5% of Black patients with COVID-19 had hidden low oxygen levels compared to 17.2% of white patients. Research also found that Black patients with COVID-19 were 29% less likely to get oxygen help quickly and three times more likely to have low oxygen levels. A study using hospital data showed that Black, Hispanic, and Asian patients had higher pulse oximetry readings than white patients for the same actual blood oxygen levels. This may lead to fewer Black, Hispanic, and Asian patients receiving oxygen help. Scientists believe that melanin in darker skin may interfere with the light used by pulse oximeters to measure oxygen. Studies suggest that melanin scatters light, making readings less accurate. Because the devices are often tested more on people with lighter skin, their settings may not work well for people with darker skin. This inaccuracy can lead to missing people who need treatment, especially for conditions like sleep apnea, which are more common in minority groups. A small difference in readings can affect medical care, such as respiratory therapy or sports training, and may require oxygen or hospital care.

Another issue is that insurance companies and hospitals use pulse oximetry numbers to make decisions about care and reimbursement. Pulse oximetry data is also used in algorithms to help doctors. Early Warning Scores, which track a patient’s health and alert doctors if needed, include pulse oximetry data and may lead to incorrect records.

Many inexpensive "consumer" pulse oximeters are available. Some people say these devices are usually accurate within a few percentage points, but they are not as reliable as medical-grade devices. Some smartwatches with activity tracking include a pulse oximetry feature. An expert said these sensors are not precise enough for medical use.

Mechanism

A blood-oxygen monitor shows the percentage of blood that carries oxygen. It uses light measurement to find out how much hemoglobin, the protein in blood that carries oxygen, is loaded with oxygen. Normal SpO2 levels for people without lung problems are between 95% and 99%. For someone breathing normal air at or near sea level, the monitor’s SpO2 reading gives an estimate of oxygen levels in the blood.

A typical pulse oximeter has an electronic processor and two small light-emitting diodes (LEDs) that shine light through a translucent part of the body, like a fingertip or earlobe. One LED is red, with a wavelength of 660 nm, and the other is infrared, with a wavelength of 940 nm. Oxygenated hemoglobin absorbs more infrared light and lets more red light pass through. Deoxygenated hemoglobin allows more infrared light to pass through and absorbs more red light. The LEDs turn on and off in a cycle about thirty times per second, which helps the photodiode detect red and infrared light separately and adjust for background light.

The amount of light that passes through (not absorbed) is measured, and separate signals are made for each wavelength. These signals change over time because the amount of blood in the arteries increases with each heartbeat. By subtracting the lowest light level from the highest in each wavelength, the effects of other tissues are corrected, creating a continuous signal for pulsating arterial blood. The processor then calculates the ratio of red light to infrared light (which shows the ratio of oxygenated hemoglobin to deoxygenated hemoglobin) and converts this ratio to SpO2 using a lookup table based on the Beer–Lambert law. This law also states that the concentration of hemoglobin and the distance light travels are proportional to how much light is absorbed. This principle is used in UV-Vis spectroscopy, which this device is modeled after.

The separated signal also helps create a plethysmograph waveform, which shows the pulsing signal and signal quality. A numeric ratio between pulsatile and baseline absorbance, called the perfusion index, can be used to evaluate blood flow.

SpO2 = HbO2 / (HbO2 + Hb)

where HbO2 is oxygenated hemoglobin and Hb is deoxygenated hemoglobin.

Changes in blood volume in the skin cause variations in the light signal received by the oximeter. These variations can be described as a periodic function, which can be split into a DC component (the peak value) and an AC component (the difference between the peak and trough). The ratio of the AC component to the DC component, expressed as a percentage, is called the (peripheral) perfusion index (Pi). Pi typically ranges from 0.02% to 20%. An earlier method called pulse oximetry plethysmographic (POP) only measured the AC component and was manually calculated from monitor pixels.

The pleth variability index (PVI) measures how much the perfusion index changes during breathing cycles. It is calculated as (Pi max − Pi min)/Pi max × 100%, where Pi max and Pi min are the highest and lowest Pi values from one or more breathing cycles. PVI is a useful, noninvasive way to check if patients need fluids. Pulse oximetry plethysmographic waveform amplitude (ΔPOP) is an earlier method similar to POP, calculated as (POP max − POP min)/(POP max + POP min) × 2.

History

In 1935, German doctor Karl Matthes, who lived from 1905 to 1962, created the first two-wavelength ear oxygen saturation meter. This device used red and green filters (later replaced with red and infrared filters) to measure oxygen levels in the blood. It was the first tool designed to check oxygen saturation.

The original oximeter was made by Glenn Allan Millikan in the 1940s. In 1943, Earl Wood added a pressure capsule to the device to push blood out of the ear, allowing for more accurate oxygen saturation readings when blood returned. This method is similar to today’s pulse oximetry, but it was difficult to use because the light sensors and light sources were not reliable. This method is no longer used in medical settings. In 1964, Shaw built the first ear oximeter that used eight wavelengths of light to measure oxygen levels.

The first pulse oximeter was developed in 1972 by Japanese engineers Takuo Aoyagi and Michio Kishi at a company called Nihon Kohden. Their device used the ratio of red and infrared light absorption from pulsating blood to measure oxygen levels. Nihon Kohden made the first pulse oximeter, called the Ear Oximeter OLV-5100. Surgeon Susumu Nakajima and his team tested the device in patients and reported results in 1975. However, Nihon Kohden stopped developing pulse oximetry and did not apply for a basic patent in countries outside Japan. This allowed other companies to improve and use pulse oximetry in the United States. In 1977, Minolta released the first finger pulse oximeter, called OXIMET MET-1471. In the U.S., Biox commercialized the first pulse oximeter in 1980, and Nellcor followed in 1983.

By 1987, pulse oximetry became a standard part of general anesthesia care in the United States. Its use quickly spread from operating rooms to recovery rooms and then to intensive care units. Pulse oximetry was especially helpful in neonatal units, where babies need precise oxygen levels. Too little oxygen can harm a baby, and too much oxygen can cause eye damage, such as retinopathy of prematurity (ROP). Taking blood samples from newborns is painful and can cause anemia. Motion and poor blood flow can also cause false readings, as some pulse oximeters cannot tell the difference between arterial and venous blood, leading to incorrect oxygen saturation measurements. Early studies showed that movement and low blood flow made pulse oximetry less reliable.

In 1995, Masimo introduced Signal Extraction Technology (SET), which could measure oxygen levels accurately even during movement or low blood flow by separating arterial signals from venous signals. Other manufacturers later developed methods to reduce false alarms, such as averaging readings over time or freezing screen values, but these methods do not measure changes during movement or low blood flow. Pulse oximeters still perform differently in difficult conditions. Masimo also introduced the perfusion index, which measures the strength of the blood flow signal. This has helped doctors predict illness severity, detect health problems in newborns, and improve the detection of heart defects in babies.

Studies have shown that Signal Extraction Technology performs better than other pulse oximetry methods. In one study, using this technology reduced eye damage in very low birth weight babies by 58% compared to centers using other methods. Other studies found fewer blood tests, faster recovery from oxygen use, and shorter hospital stays. This technology also works in areas where movement and low blood flow made monitoring difficult, such as general hospital floors. A 2010 study showed that using Signal Extraction Technology at Dartmouth-Hitchcock Medical Center reduced emergency team calls, ICU transfers, and ICU stays. A 2020 study found no patient deaths or harm from opioid use during 10 years of using this technology with a patient monitoring system.

In 2007, Masimo introduced the pleth variability index (PVI), which helps doctors assess how well a patient can respond to fluid treatments. Proper fluid levels are important for healing and reducing risks like infections or heart problems. The UK and France now recommend PVI monitoring during surgery.

In 2011, experts recommended using pulse oximetry to screen newborns for critical heart defects. Studies using Signal Extraction Technology found that it could detect these conditions with few false positives. The U.S. added pulse oximetry to its newborn screening guidelines. Before this, less than 1% of newborns were screened. Today, nearly all newborns in the U.S. are tested, and screening is growing worldwide. A 2014 study of 122,738 newborns using Signal Extraction Technology also showed positive results.

High-resolution pulse oximetry (HRPO) has been developed for home use to check for sleep ap