Login Help


Visit Elsevier's
Caring for the Ages Web Site
Get Your Free Subscription! Selected Articles 2001-2004

Caring for the Ages
Selected Articles from
October 2003;
Vol. 4, No. 10
Medicare Drug Benefit Update
Stressed Staff
Quality of Care Under Fire
Evidence-Based Practice in LTC: Pulse Oximetry
COPD--LTC's "Quiet" Disease
The Leadership Mirror
A Daughter's Journal: The Homecoming Queen
Previous Month's Articles
Following Month's Articles

Evidence-Based Practice in LTC

Pulse Oximetry

Its prevalence & purpose in the LTC setting


by Steven Levenson, MD, CMD
Multi-Facility Medical Director, Baltimore, MD
Chair, Caring's Editorial Board

Charles Crecelius, MD, PhD, CMD
Past President, Missouri Association of Long-Term Care Physicians
Medical Director, Delmar Gardens, St. Louis, MO
Member, Caring's Editorial Board

Common Practice

Pulse oximetry testing is commonly used in nursing homes to assess patients' oxygenation status, especially in postacute care. Pulse oximetry results are often used independently to decide whether to notify the physician, to conclude that an urgent situation exists, or to hospitalize a patient.

Physicians regularly give orders to adjust oxygen to maintain pulse oximetry results at or above a particular reading. They may react to pulse oximetry readings out of context or instead of more important clinical evidence, and in doing so may order patient transfers based on pulse oximetry results and despite relative patient stability.

It's not uncommon for nursing staff and physicians to misunderstand the purpose and limitations of pulse oximetry.

The Evidence

While helpful, pulse oximetry must be used and interpreted properly. Never base clinical decisions and conclusions about the severity of a patient's illness on pulse oximetry alone, for example.

Pulse oximetry provides estimates of arterial oxyhemoglobin saturation (SaO2) by utilizing selected wavelengths of light to noninvasively determine the saturation of oxyhemoglobin (SpO2). Various circumstances may affect the accuracy of results.

In general, pulse oximetry is useful only in normotensive patients. Those with low flow states or those in shock often have poor peripheral perfusion, resulting in low tissue hemoglobin saturation and falsely depressed pulse oximetry readings. Pulse oximetry can't distinguish between different forms of hemoglobin or reflect severe anemia. Nail varnish and cold extremities can cause falsely low readings. And pulse oximetry can overestimate the oxygen saturation in some forms of cardiac disease.

Thus, oxygen saturation by pulse oximetry (SpO2) may be an effective screening test for SaO2, and, therefore, serve as an effective gauge of systemic hypoxia. But SaO2 is a clue to--not definitive proof of--the patient's overall status. Classification of the severity of hypoxemia based on oxygen saturation remains somewhat arbitrary. The use of 90% as an acceptable cutoff is based on the oxyhemoglobin dissociation curve; that is, how red blood cells attract and transport oxygen and release it to body tissues. Some investigators suggest that pulse oximetry readings below 92% may indicate hypoxia; however, the body can compensate for hypoxia (e.g., by releasing oxygen more readily from red blood cells directly to tissues). Acid-base balance and other factors can also affect oxygen delivery to tissues. Therefore, pulse oximetry results comprise only one part of the patient's overall status.

Direct patient evaluation remains very important. For example, if the body's organs need more oxygen, reflex mechanisms will stimulate breathing and increase heart rate. So, someone who's not dyspneic or tachycardic may compensate adequately for relative hypoxia, despite a pulse oximetry reading below 90%. This situation is common in patients with more advanced chronic obstructive pulmonary disease, who may function adequately despite having oxygen saturation between 86% and 90%.

Repeated pulse oximetry readings below 85% accompanied by dyspnea symptoms may be considered problematic--particularly if they persist despite treatment of underlying cardiopulmonary conditions. Pulse oximetry may help confirm the presence of pneumonia, but most patients with pneumonia can be treated successfully without hospitalization.

The American College of Physicians Chronic Obstructive Pulmonary Disease Guideline Algorithm states that there is no evidence for using pulse oximetry to diagnose or gauge the severity of acute exacerbations of COPD. After all, chronic obstructive pulmonary disease, by definition, involves pulmonary damage that usually includes impaired gas exchange. We can, therefore, expect lower baseline readings in many individuals who have chronic cardiopulmonary disease.

Also, pulse oximetry doesn't measure acid-base balance or carbon dioxide retention. If a hypoxic individual also retains CO2, administering too much oxygen can worsen the situation by reducing the respiratory drive and increasing CO2 retention; therefore, excessive reliance on pulse oximetry results can be problematic.


Always use pulse oximetry judiciously. Don't draw conclusions about a patient's condition or the urgency of a situation based pulse oximetry alone. Instead, combine oxygen saturation results with a patient exam and thorough patient history.

Hypoxia may cause or accompany many illnesses and condition changes. Check the oxygen saturation for a patient who develops symptoms that suggest oxygenation problems (e.g., dyspnea, cough, or congestion).

But routine use of pulse oximetry as another vital sign or to check for respiratory conditions may provide misleading information. Base routine checks on a physician order, as with vital signs. The orders should include parameters for notifying the physician, but the parameters must remain realistic and pertinent to that individual. It is no more necessary to notify a physician for every low pulse oximetry reading than it is to call a physician for every abnormal blood sugar. Always discuss the whole patient and include all relevant information in the discussion.

Example: A facility policy may dictate that nurses should notify physicians for pulse oximetry readings below 92%. But if the patient remains stable or continues to improve (or customarily has a lower pulse oximetry result), they may not need to report individual readings immediately. Instead, as with blood sugars, report the aggregate results collected over a day or two, along with other information about the patient's status, including vital signs and results of listening to the lungs, observing and describing respirations, and evaluating the patient overall. In this way, the physician can adjust oxygen administration based on the whole picture.

In patients with known or suspected heart or lung disease, such as chronic obstructive pulmonary disease, the pulse oximetry results may fluctuate and the oxygen saturation may fall or persist below 90%. But never use the pulse oximetry result alone to determine the urgency of the situation or the need for hospitalization. The patient's overall status and stability of vital signs may be a more important indicator than the fluctuating pulse oximetry results or the difficulty of maintaining an SpO2 above 90%. Although pulse oximetry can be used to gauge the presence of an exercise-induced drop in oxygen saturation in individuals with COPD, no evidence supports using the results to adjust treatments in acute exacerbations of this disease.

In all cases, but especially when SpO2 indicates a possible problem, a nurse, respiratory therapist, or health-care practitioner should evaluate the patient and take accurate, complete vital signs. The presence of stable vital signs, a lack of dyspnea, and a respiratory rate below 28 to 30 indicate the patient may be compensating adequately for hypoxia. If the vital signs are stable, the patient is not dyspneic, and the respiratory rate is not above 28 to 30, then the patient may be compensating adequately for hypoxia. Hospitalization may not benefit a patient more than basic care in the long-term care facility. Also consider advance care directives in making decisions about how and where to treat someone with hypoxia. There is no clearly defined benefit of using pulse oximetry in hospice care.

Physicians should be cautious and realistic about their orders for adjusting oxygen flow to maintain a specific pulse oximetry reading. While oxygen adjustment may be indicated, it may be difficult if not impossible for staff to follow an order, such as "adjust oxygen to maintain pulse oximetry above 92%." This doesn't clarify the frequency of monitoring and offers no clue as to the oxygen saturation in between pulse oximetry checks. The best approach may be to monitor the patient's symptoms and physical findings.

Only properly trained staff should use, document, or report pulse oximetry results. Staff should know of factors that may influence pulse oximetry results, including motion artifact, abnormal hemoglobins, exposure of the measuring probe to ambient light during measurement, low perfusion states (including hypotension and shock), skin pigmentation, skin temperature of extremities used to conduct the test, and nail polish or nail coverings with finger probe.

Additionally, documentation should include the date and time of measurement; the pulse oximeter reading; the patient's position, activity level, and location during monitoring; inspired oxygen concentration or supplemental oxygen flow; and the type of oxygen delivery device.

The Evidence

Bordow RA and Moser KM. Evaluation of arterial blood gases and acid-base homeostasis. Man Clin Probl Pulm Med. MAXX: The Electronic Library of Medicine [on CD-ROM]. New York, NY: Lippincott, Williams and Wilkins; 1998.

PaO2 can fall to 60 mm Hg with hemoglobin still over 90% saturated. But with a drop from 60 to 40 mm Hg, saturation falls steeply to below 75%. Clinically this relationship explains a fundamental goal in treating hypoxemia: Get the PaO2 above 60 mm Hg (the flat portion of the curve). Classification of the severity of hypoxemia based on PaO2 is arbitrary. A value of greater than 60 mm Hg usually indicates mild hypoxemia, 45-59 mm Hg is moderate, and below 45 mm Hg is severe.

Kaye KS, Stalam M, Shershen WE, Kaye D. Utility of pulse oximetry in diagnosing pneumonia in nursing home residents. Am J Med Sci. 2002; 324(5):237-242.

Oxygen saturations were lower in 45 pneumonia patients than in 22 patients with acute nonpulmonary infections. The drop in oxygen saturation from the previous baseline value was greater in pneumonia patients than in control subjects. If a person's usual pulse oximetry results are known, a decrease in oxygen saturation in acutely infected nursing home patients of > 3% from baseline, as well as a single oxygen saturation of < 94, should suggest pneumonia. A decrease from baseline of < 4% or a single oxygen saturation of 94 or higher suggests that pneumonia is unlikely.

Carone M, Patessio A, Appendini L, et al. Comparison of invasive and noninvasive saturation monitoring in prescribing oxygen during exercise in COPD patients. Eur Respir J. 1997; 10(2):446-451.

The authors conclude that noninvasive measurement of oxygen saturation is inadequate for estimating arterial saturation in COPD. They suggest correcting exercise-induced desaturation with supplemental oxygen when SpO2 falls below 93%.

Smatlak P, Knebel AR. Clinical evaluation of noninvasive monitoring of oxygen saturation in critically ill patients. Am J Crit Care. 1998; 7(5):370-373.

In patients with an abnormal cardiac index, pulse oximeter measurements may have a greater margin of error.

This article originally appeared in Caring for the Ages, October 2003; Vol. 4, No. 10, p. 16, 22-23. Caring for the Ages is an official publication of the American Medical Directors Association, published by Elsevier. This article may not be reproduced in any form, print or electronic, without permission.

The opinions expressed by the authors are their own
and not necessarily those of AMDA or of Elsevier.

back to top

    11000 Broken Land Parkway, Suite 400 Columbia, MD 21044
    Phone: 410-740-9743 • Toll free: 800-876-2632
    Fax: 410-740-4572 • E-mail: webmaster@amda.com