Pulse Oximetry

Probably no device has had such widespread and quick acceptance into clinical instrumentation in the past 20 years as the pulse oximeter.

The modern pulse oximeter should not be confused with the ear oximeter, which was marketed in the mid 1970s. The technology is quite different; the old system, which used eight wavelength detectors, could only be used on the earlobe, required heating of the area, and was not portable. Another early version used a fiber-optic cable for light transmission and detection as it was too big to be put into a probe that could be comfortably applied to the patient.

Much of the original work on pulse oximetry was to develop a noninvasive method of determining cardiac output. The side effect of getting a good correlation on blood oxygen levels (SpO2) proved to be the marketable product, and research on using the technology for cardiac output basically stopped. An urban legend developed saying that both the pulse oximeter and Viagra were side effects of the prime objective of the engineering work, one measuring cardiac output and the other increasing it.

The “modern” pulse oximeter owes much of its success to William New, MD, PhD, who introduced the Nellcor unit in the mid 1980s. Ohmeda introduced the Biox II in the mid 1980s, this time using microprocessors, another major factor in the utilization of pulse oximetry. Additional credit must be given to the malpractice insurance companies that told anesthesiologists that if they used pulse oximetry their premiums would be reduced.

The pulse oximeter works on a reasonably simple principle of light absorption, as defined by the Beer-Lambert law, sometimes called Bouguet’s law, depending on the textbook. Basically, the law states that light is absorbed or passed through a solution based on the concentration of the chemical in the solution for a certain light wavelength. It was found that hemoglobin (Hb), nonoxygenated blood that is dark red in color, and oxyhemoglobin (HbO2), oxygenated blood that is bright red in color, have different light-absorption levels. By using two detectors—one in the 660 nm range to measure hemoglobin and the other in the 940 nm range to measure the oxyhemoglobin—along with proprietary algorithms, the pulse oximeter can obtain accurate clinical results on blood oxygen.

Pulse oximeters have some limitations, though, as ambient light can affect the readings, as can shivering, low flow, very thick skin, and poor placement of the sensors. Most of the newer designs (after 1998) have much better rejection systems for motion artifacts. When the finger is used as the location of the probe, it is important that the light source be placed on the nail and the detector on the soft tissue of the finger. Needless to say, the patient should not have nail polish on. A patient with carbon monoxide exposure will register falsely high on oxyhemoglobin since the blood will be very red. For these patients, co-oximetry or end tidal CO2 (capnometry) will give better clinical results.

The widespread use of pulse oximetry has increased patient comfort in that far fewer traditional blood gas measurements are taken now than in the past. If you have ever had an arterial “stick” for a blood gas test, you know that it is quite painful. The use of direct monitoring of blood pressure, another source of obtaining blood gas test samples, is also down in many hospitals.

There are some problems that biomeds still have to respond to with pulse oximeters. Bad sensors and cables are probably the most common. Some shops will “reprocess” the sensors and detectors on disposable units; be careful on this, as you may become a manufacturer in the eyes of the legal system and be without insurance protection. Other common calls we get include problems with batteries, white tape (white tape should be a controlled substance), “bounce tests” (most units do not bounce well off the floor), and people saying, “We cannot find it on the floor so you must have it in the shop.”

If a patient has a cardiac pacer, either internal or external, the pulse oximetry alarm may become a secondary alarm on some monitoring systems. This can present a problem as the patient monitor may not sound an alarm but only display a “screen flash” if the alarm limit on the pulse oximeter is triggered. Take a little time during the next PM cycle on the monitors to confirm how the alarms react when a patient is being paced.

In closing, please be aware of where the alarm limits can be set for the low alarms on your stand-alone devices, monitoring systems, and multipurpose stand-alone units. Most of us just check the default setting and do not try to adjust the limit down below 90. Some manufacturers will allow the user to have alarm limits as low as 50. So take the time to check the limits.

David Harrington, PhD, is director of staff development and training at Technology in Medicine (TiM), Holliston, Mass.

Ed Bober is a TiM field service BMET.