Before pulse oximetry could be developed, the scientific groundwork on which it is based had to be laid. A series of discoveries and refinements over a period of centuries preceded the initial work on pulse oximetry itself. In 1729, the variable, reproducible absorption of light was discovered, although the first spectrometer to take advantage of this principle was not developed until 1862. Two years later, hemoglobin was found to carry oxygen, and the first measurement of oxygen saturation using light took place after the passage of an additional 68 years.

An early form of the ear saturation meter was developed in 1935; it used red and green light, with the green light later replaced by infrared wavelengths. A pressure capsule that emptied the ear of blood and measured oxygen saturation as the blood returned was added to the ear meter in 1949. These were research instruments that saw no clinical use because their components were still too unstable and unreliable.

In 1964, however, an eight-wavelength ear oximeter was developed; it was soon made available commercially, but was accepted only in sophisticated diagnostic settings, such as pulmonary function laboratories. A form of pulse oximetry that used the differential absorption of red and infrared light was developed in 1972 and commercialized in 1981. A reusable sensor was introduced at that time for use in diagnostic work, especially in the exercise physiology and pulmonary function settings.

Dissemination Begins

In 1983, single-use sensors were introduced and pulse oximetry took its place in the operating room, with gradual expansion to postanesthesia care and intensive care units. For anesthesia, the use of pulse oximetry became part of the standard of care in 1986.

Soon, the technology began to be incorporated into the multiparameter patient-monitoring systems seen in critical care units. Emergency departments were also among the early adopters of pulse oximetry, and this accelerated as more handheld models became available. The same units began to be carried by respiratory therapists for spot checks, especially of patients using mechanical ventilation.

Neonatal care was another area of early and broad adoption, since the oxygen-saturation range at which premature and critically ill infants can avoid long-term adverse consequences of supplemental oxygen is narrow. Many infants cannot readily afford the blood loss associated with frequent analysis of arterial blood gases (ABGs).

NONIN PalmSAT 2500®

  • Handheld pulse oximeter: for 6MWT, LTOT, or overnight sleep screening
  • 100 hours of battery life
  • NiMH rechargeable battery pack lasts for 45 hours
  • Memory. Download with NONIN’s nVision® software
  • Audible and Visual Alarms Option
  • Available (2500A)
  • Carrying Case


At this time, enthusiasm for pulse oximetry had increased rapidly, but hospital budgets lagged behind clinical needs. Many hospitals lacked as many pulse oximeters as the staff demanded and the physicians ordered, so some respiratory therapy departments were given the task of developing elaborate protocols that determined which patients would receive spot checks and which would be allowed to use continuous pulse oximetry. Fortunately, as microprocessor technology became less expensive and production volumes grew, pulse oximeters became more affordable and their use was no longer rationed.

ABG testing was in extensive use, of course, but the pain, blood loss, and analysis time involved made many blood-gas measurements ideal targets for replacement by pulse oximetry. ABGs are still of crucial importance in their own right clinically because they provide information that pulse oximetry does not, but a great many patients are now spared the ABG ordeal because of pulse oximetry. Likewise, the effects of a medical intervention can be assessed immediately in the presence of pulse oximetry, without the wait for confirmation of rising saturation that was once inevitable, even with the most rapid laboratory turnaround. Because pulse oximetry, per measurement, is far less costly than ABG analysis, this improved use of limited resources also drove adoption of the technology.

Masimo Rainbow
SET Technology

  • Uses 7+ wavelengths of light to measure SpCO and SpMet
  • Perfusion Index (PI) with trending capability
  • Pleth Variability Index (PVI)
  • Signal IQ® waveform for signal identification and quality indication during excessive motion and low signal to noise situations
  • FastSat™ tracks rapid changes in arterial O2
  • Variable pitch provides tonal variance for every 1% change in saturation
  • SatShare® interface
  • Remote alarming interface
  • Up to eighteen days of trending
  • Detachable portable hand-held

Technical Refinement

In 1994, Wiklund et al1 reported that poor perfusion, motion artifacts, and sensor displacement caused 77% of alarms generated by pulse oximeters to be false. The time elapsed between alarms was, on average, 8 minutes. In addition to the patient’s motion (significant in about one in five critical care cases), there are other sources of interference with the pulse-oximetry signal; these include high and low levels of ambient light and the frequency of emissions from other electronic equipment. Instead of dropping the reading altogether, some pulse oximeters available since 1994 will indicate that interference is affecting the reliability of the signal, and manufacturers have continued to refine the ability of sensors to compensate for signal interference and low perfusion.

A self-contained fingertip unit was introduced in 1995, and wireless transmission of results has been available for several years. A form of signal processing introduced internationally in 1996 and approved for US use in 1998 can overcome the inaccuracies associated with motion and low perfusion. A combination sensor that detected carbon dioxide as well as oxygen was developed some years ago, but did not achieve broad clinical acceptance and is no longer available commercially.

US Food and Drug Administration 510(k) clearance was granted in 1999 for marketing a pulse-oximetry technology indicated for use in the care of patients with poor perfusion and/or excessive motion. This technology, signal extraction, took advantage of advances in parallel processing and adaptive filters to track saturation and pulse changes during low perfusion and patient movement, thus preventing the signal dropout that might otherwise be encountered just when a reading is most vital.

Nellcor MAX-FAST®
Forehead Sensor

  • Detects changes in SpO2 during low perfusion
  • Equipped with LoSat Expanded accuracy range
  • Can be used on mechanically ventilated and supine patients
  • Indicated for pediatric and adult patients
  • Single-patient use
  • Four adhesive layers to allow long-term use
  • Soft adjustable headband prevents venous pooling, maintains proper sensor positioning for accurate, reliable readings

While the accuracy and reliability of pulse oximetry at saturation levels of more than 85% have been widely accepted for some time, more recent advances in adhesive sensors have extended accuracy to levels as low as 60%. This represents a considerable benefit for treatment of the critically ill, in whom the lowest saturation levels are encountered (but for whom rapid results are most needed).

In 2002, a forehead sensor was introduced that is compatible with mechanical ventilation. Because it is less dependent on peripheral perfusion, it is capable of picking up an accurate signal in poorly perfused patients for whom a finger sensor will not work. A headband that is recommended for use with the forehead sensor prevents venous pooling by applying light pressure. Even in cases in which the finger sensor returns a reading (which it is not always possible to obtain in critically ill patients with low cardiac output), the forehead sensor–headband combination is more accurate than a finger sensor and has fewer dropped readings.2 The forehead sensor’s accuracy can be affected by poor placement, failure to apply the headband, or the Trendelenburg patient position, so staff education in its use is important.

Current initiatives in pulse oximetry include The Joint Commission’s work on overcoming the tendency of some health care personnel to ignore pulse-oximeter alarms and the Anesthesia Patient Safety Foundation’s investigation of the need for continuous pulse oximetry in conjunction with patient-controlled analgesia after surgery in patients with obstructive sleep apnea.

The noninvasive measurement of total hemoglobin, currently nearing the culmination of its clinical and engineering research phases, is expected to be cleared for marketing in the United States later in 2007.

Kris Kyes is technical editor of  RT. For further information, contact [email protected].


  1. Wiklund L, Hok B, Stahl K, Jordeby-Jonsson A. Postanesthesia monitoring revisited: frequency of true and false alarms from different monitoring devices. J Clin Anesth. 1994;6:182-8.
  2. Schallom L, Sona C, McSweeney M, Mazuski J. Comparison of reflective and transmission oximetry in surgical/trauma patients with poor perfusion. Crit Care Med. 2004;32(suppl A35):S138.