Practical Pulse Oximetry

Pulse Oximetry in the Emergency Department

Robert R. Fluck

Robert R. Fluck, Jr. RRT Associate Professor, Department of Respiratory Therapy Education Joseph McDonald, MS, RRT Associate Director, Respiratory Care Services Carlos J. Lopez III, MD Department of Anesthesiology Upstate Medical University, State University New York (SUNY) Syracuse, NY, USA

The article also available in PDF: 155KB

Introduction

Pulse oximetry has become the fifth vital sign in the clinical arena, joining temperature, pulse, blood pressure and respiratory rate. While the measurement of someone's oxyhemoglobin saturation seems quite simple, in reality the correct interpretation of the results is more complicated. The clinician needs to have several potential problems in mind; these problems at best can render the reading misleading, and at worst can result in an inaccurate determination of the saturation. The rapid-paced emergency department requires that clinicians be well versed in these issues.

This article will explore the most common and important reasons that clinician interpretation is necessary in the setting of the hospital emergency department. Common clinical problems include carboxyhemoglobinemia, methemoglobinemia, and decreased peripheral perfusion (Table 1). Technical issues include incompatibility of sensor with oximeter, incorrect application of the probe to the patient, obtaining a pulse oximetry reading on someone with dark nail polish or acrylic nails, and obtaining a pulse oximetry reading on someone with an arterio-venous shunt for dialysis.

Clinical interpretation of SpO2 values

Dyshemoglobins: The clinician needs to remember that the pulse oximeter measures functional saturation because it utilizes only two wavelengths of light (660nm (red) and 920nm (infrared)). The pulse oximeter can differentiate only between deoxygenated (or reduced) hemoglobin, which has the ability to transport oxygen (with a peak absorption about 660nm) and oxyhemoglobin (with a peak absorption about 920nm).

In order to measure dyshemoglobins, such as methemoglobin, carboxyhemoglobin and sulphhemoglobin, the clinician needs to utilize a co-oximeter. The co-oximeter utilizes multiple wavelengths of light and can differentiate among all types of hemoglobin (measuring what is termed 'fractional saturation'). The downside of the co-oximeter is that it requires a sample of blood obtained from the patient to analyze, with its attendant inconvenience, pain, expense and delay.

Carboxyhemoglobin: Probably the most common type of dyshemoglobin is carboxyhemoglobin, the combination of hemoglobin with carbon monoxide (CO). Carboxyhemoglobin unfortunately has an absorption peak at approximately the same wavelength (920 nm) as oxyhemoglobin. Therefore, if the clinician is not suspicious of CO presence, he may be lulled into a false sense of security because the oxyhemoglobin saturation may read as acceptable when in fact the actual oxyhemoglobin saturation is much less. Patients who smoke often have elevated levels of carboxyhemoglobin and the pulse oximeter will not reflect this.

Emergency evaluation of these patients should include questions about smoking history and an arterial blood gas for co-oximeter assessment to ascertain precise saturation levels. Among the (mostly non-specific) signs and symptoms of CO poisoning are:

• Dyspnea
• Headache
• Nausea and vomiting
• Poor judgment

The cherry-red lips and mucous membranes of folklore are not present in every case. A high level of suspicion of CO exposure should also be present if cohabitants present to the emergency department. Emergency department and emergency rescue staff must register suspicion of CO poisoning and take rapid action to assess levels via co-oximetry to ensure prompt care [1-2].

Methemoglobin: Another problem for the clinician in the emergency department is the patient with methemoglobinemia. Methemoglobin is produced in the body on a daily basis by oxidizers, but ordinarily is quickly reduced to hemoglobin by an enzyme found in the erythrocyte called methemoglobin reductase. Some commonly used drugs, including nitroprusside, primaquine, dapsone, aniline, nitrates, nitrites and local anesthetic agents (benzocaine family) can also oxidize hemoglobin (normal adult hemoglobin) to methemoglobin [3,4].

The normal range for methemoglobin is less than 2%. Because of the way the methemoglobin is assessed by the algorithm in the oximeter, regardless of the saturation of oxygen-containing hemoglobin, the oximeter will tend toward a reading of 85% [5]. In the emergency department, methemoglobinemia should be kept in mind as related to possible ingestion of nitrate/nitrite products.

Babies have a relatively low level of methemoglobin reductase and thus are much more sensitive to nitrates, which are found in fertilizers, and nitrites, which are used in cured meats such as hot dogs. In the mainly rural part of the authors' state, a baby appeared in the pediatric emergency department with a significant hypoxemia. The cause was rapidly determined to be methemoglobinemia. In ascertaining the source of the methemoglobinemia, the treating physician found that the infant lived on a farm. His mother prepared the formula she was giving to the baby with water from the well on the farm. This well turned out to be contaminated by runoff of nitrate-containing fertilizer from the fields of the farm.

Another recent example in the literature is that of a 23-year-old who consumed ice that had been contaminated by a broken pack of instant ice that contained ammonium nitrate. The absence of an improvement in oxygenation after oxygen administration and the 'chocolate brown' color of the arterial blood provided clues in the suspected methemoglobinemia diagnosis [6].

Decreased Perfusion: Another situation commonly encountered in the emergency department is a patient with decreased perfusion. It may be difficult to obtain a pulse oximeter reading as the pulse signal size may decrease and the magnitude of the signal may begin approaching the electronic noise level. The oximeter may not be able to differentiate between the signal and the noise.

Concern should focus on restoring adequate perfusion. Additionally, the clinician should correlate the pulse oximetry results with the patient's clinical condition [7]. In these situations, clinicians will often place multiple sensors on the patient's fingers or toes in an attempt to obtain a reading from the pulse oximeter. This may not be helpful and can create additional problems.

In some settings, it is common for caregivers to wrap the oximeter probe with elastic tape when they are unable to get a SpO2 reading. There are three downsides to this:

• The elastic may further reduce perfusion to the digit
• Taping further delays instituting definitive treatment of the patient
• The additional pressure on the extremity may result in pressure necrosis

Some manufacturers of new generation pulse oximeters have incorporated algorithms that provide a numerical value to gauge signal strength for an individual patient. For example, GE Healthcare has adder the Relative Perfusion Index (PIr) into recent oximeters to help quantify the best perfusion site.

Instead of finger or toe, the clinician may also choose from several alternative techniques to obtain information about the patient’s arterial oxygen saturation. Another method would be to use a core site, such as an ear, nasal or forehead location. These sites tend to be better perfused than extremities.

The clinician must remember, however, that an instrument that is designed to use a finger probe may not provide correct readings with a forehead probe. The finger probe uses transmission spectrophotometry whereas the forehead probe uses reflectance spectrophotometry, which requires a different software algorithm and is not available on all oximeters. The last method would be to take a sample of arterial blood from the patient and use a co-oximeter to measure the actual oxyhemoglobin saturation.

Important technical issues

Compatibility of sensor with pulse oximeter: Generally, sensors should be from the same manufacturer as the oximeter itself. Mere compatibility of the plug for the sensor with the socket on the oximeter does not guarantee that the sensor will work with that oximeter. Consult the manufacturer's Instructions for Use for the sensors and oximeters to insure the correct accessories are available. The clinician should first ensure that all indicator lights are functional and that there are no alarms that could result from an attempt to use an incompatible probe. This problem is also exacerbated when there are many different agencies that transport patients to a given emergency department. In the authors' region, for instance, there are three agencies that transport by helicopter, a commercial ambulance service and a large number of volunteer ambulances, both stand-alone and those affiliated with fire departments. In addition, pulse oximeters from other hospital departments may appear in the emergency department.

Standardization is highly unlikely among this large and disparate number of agencies as there are a large range of technologies and algorithms that require different sensor hardware and algorithms. Emergency departments are encouraged to label probes with appropriate oximeters if they are not readily apparent. The rush to obtain an accurate reading in a critical setting can be doubly frustrating if equipment failure alarms begin occurring.

Incorrect application of probe: A parallel situation is inappropriate use of a probe; for example, using an adult finger probe on an infant or using a disposable finger probe on an infant's foot or ear or on a patient's forehead. Just because the probe can be placed on a particular body part, it does not mean that it will read accurately or dependably. This problem can be obviated by ensuring that the probe is placed correctly (with light source and detector opposite each other for a transmission type probe).

Additionally, the clinician could determine what the perfusion index reads for that particular location (compared with others). Manufacturers offer some weight related and site guidelines for probe use. The time that it takes to place the correct probe in the correct location will be saved in the future when the clinician does not have to spend time troubleshooting the oximeter and probe because of poor accuracy.

Patient's A-V shunt reduces pulse amplitude: The patient with the indwelling shunt for hemodialysis may present an interesting challenge. The presence of the shunt (generally from radial artery to a vein in the forearm) may greatly reduce the amplitude of the pulse in that hand, rendering a reading impossible. The therapist would need to be aware of the presence of the shunt and switch to the contra-lateral hand. The presence of the shunt may also falsely lead the therapist to believe that the patient has a perfusion problem when none exists. This is another situation in which the clinician could make use of the perfusion index to determine relative pulse strength.

Patient motion may be an issue in the emergency department, especially in the newborn and pediatric group and the disoriented or confused adult population. Pain may also translate to increased motion with some patients [8]. More recently, developed technology may offer enhanced signal recognition and more stable oximeter readings in these settings.

False nails and nail polish: Frequently, the clinician will be presented with a patient who has either acrylic nails or nail polish, some varieties of which are quite dark. These situations may alter the path of light through the oximeter probe and produce inaccurate readings [9]. Blue or black nail polish may be especially problematic [10]. The best option for ensuring accuracy is removing the polish or false nail if possible or choosing an alternative site.

The effect of ambient light on pulse oximetry results: While the issue of ambient light interference with pulse oximetry in the emergency department is a rare concern, the authors believe it should be mentioned for completeness. A study conducted at the authors' institution evaluated incandescent lights, quartz halogen lights, fluorescent lights, bilirubin lights and heat lamps as potential sources of interference to pulse oximetry. The largest variation from control (measured in complete darkness) was 0.5% saturation, which was not statistically significant. Even if it had been statistically significant, that difference was clinically unimportant [11]. However, at times, the clinician may want to mitigate the ambient light, especially if an infant is being assessed; placing a sock over the foot to shield the probe may be helpful.

Summary

Any measurement or test done on a patient should be evaluated for appropriateness. To decide whether a test or measurement is appropriate, the clinician should ask himself:

• Does the result of this test help me arrive at a diagnosis or confirm it?
• Does the result of this test potentially change the way I am treating this patient?

If the answer to one or both of these questions is ‘Yes’, then the test should be done. In the case of pulse oximetry in the emergency room, an additional consideration must be made:

• Will this measurement be accurate, and what can I do to improve the accuracy?

This paper has attempted to highlight the situations which may create questionable oximetry results in the emergency department (see Table 1). Clinicians must be knowledgeable about these situations in order to ensure accurate interpretation and use of oximetry results.

In addition, clinicians must recognize the range of accuracy expected of oximeters. Specifications for oximeters with the latest generation of technology note accuracies of ±3%. This means that two oximeters could differ by, as much as 6% and both still be technically accurate. Co-oximeters also have an accuracy of ±2% so an arterial blood gas co-oximeter saturation value could differ by 5% from the pulse oximeter value and still be accurate as well.

Finally, competent use of pulse oximetry in the emergency department requires more than simply applying the most readily available sensor to the patient and the oximeter unit. Clinicians must remain aware of potential problems based on clinical considerations and technological factors.

Table 1: Common clinical considerations with pulse oximetry in the emergency department

References

1. Kao L W, Nanagas K A, "Carbon monoxide poisoning", Emerg. Med. Clin. N. Am. 2004 Nov; 22 (4): pp. 985-1,018.
2. "Use of Carbon Monoxide Alarms to Prevent Poisonings During a Power Outage – North Carolina, December 2002", MMWR, March 12, 2004/53 (09); pp. 189-192. (Last accessed December 20, 2004.) http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5309a1.htm
3. Hicks G H, "Cardiopulmonary Anatomy and Physiology", Philadelphia: W. B. Saunders; 2000: 381.
4. Batchelder P, Clifford D, Goldman J M, "Pulse Oximetry: Real World Performance", in Yearbook of Intensive Care and Emergency Medicine 2004, Springer Publications.
5. McCarthy K, Decker M J, Strohl K P, Stoller J K, "Pulse Oximetry", in Kacmarek RM, Hess D, Stoller JK, Monitoring in Respiratory Care, St. Louis, MO, Mosby-Year Book, Inc. 1993.
6. Brunato F, Garziera M G, Briguglio E, "A severe methaemoglobinemia induced by nitrates: a case report", Eur. J. Emerg. Med. (2003), Dec; 10 (4): pp. 326-230.
7. AARC Clinical Practice Guidelines, "Pulse oximetry", Respir Care (1991), 36: pp. 1,406-1,409.
8. Tobin R M, Pologe J A, Batchelder P B, "A characterization of motion affecting pulse oximetry in 350 patients", Anesth. Analg. 94: pp. S54-S61.
9. Wouters P F, Gehring H, Meyfroid G, et al., "Accuracy of pulse oximeters. The European multi-center trial", Anesth. Analg. (2002), 94: pp. S13-S16.
10. Brand T M, Brand M E, Jay J D, "Enamel nail polish does not interfere with pulse oximetry among normoxic volunteers", J. Clin. Monit. Comput. (2002), 17: pp. 93-96.
11. Fluck R, Schroeder C, Frani G, Kropf B, Engbretson B, "Does Ambient Light Affect the Accuracy of Pulse Oximetry?", Respir. Care (2003), 48 (7): pp. 677-680.

Last updated: 8 June 2005Created
Legal notice	© GE Healthcare 2008 ISSN 1795-6269 (Web) ISSN 1795-6277 (CD)	Webmaster