Respiratory System Monitoring: Basics of Pulse Oximetry and Capnography

Andrea L. Looney, DVM
MSPCA/Rowley Memorial Animal Hospital
Springfield, MA


Pulse Oximetry

Respiratory monitoring is an essential tool in the management of the critically ill and anesthetized patient. A variety of monitoring devices that have been traditionally used in only very specialized settings such as referral and institutional hospitals have now become available to the practitioner and offer a great deal of information to those who utilize them. Historically, the use of actual blood gas analysis has served as the gold standard for assessing respiratory gas exchange, which includes assessment of both oxygenation (oxygen assessment) and ventilation (CO2 assessment). The more recent advent of pulse oximetry to assess oxygenation, and of end tidal capnography to assess ventilation and perfusion has added to the armamentarium of the clinician. The commonly employed monitors have provided an excellent non-invasive measure of respiratory and even cardiovascular function if they are appropriately employed. The following discussion will address the utilization and limitations of pulse oximetry and end tidal capnography in practice.

Oxygen delivery is vital to the well being of living creatures. Pulse oximeters measure pulse rate and estimate oxygen saturation of hemoglobin in percentage (SpO2) by measuring the transmission of light at two wavelengths through a pulsatile vascular tissue bed. The instrument produces light at the wavelength of both oxygenated and deoxygenated hemoglobin. The light is transmitted and received using a specially designed probe. The percentage of oxygenated hemoglobin and deoxygenated hemoglobin is determined by measuring the ratio of infrared and red light transmitted to a photodetector. Pulse oximeters combine the use of both plethysmography and transmission spectrophotometric analysis to measure hemoglobin saturation.

The oxygen saturation measured by a pulse oximeter (SpO2) is not the same as the SaO2 measured by a laboratory cooximeter. The pulse oximeter measures the "functional" saturation of hemoglobin. Functional saturation represents the amount of oxygenated hemoglobin in a percentage form of the total reduced and oxygenated hemoglobin. The laboratory cooximeters use multiple wavelengths that distinguish other types of hemoglobin (methemoglobin and carboxyhemoglobin) and thus measure true fractional saturation, or amount of oxygenated hemoglobin in a percentage form of the total reduced + oxygenated hemoglobin + methemoglobin, + carboxyhemoglobin. This helps to understand why the SpO2 measurement can exceed the SaO2 reading in certain conditions.

Oxygen delivery is a product of cardiac output and oxygen content. The equation for arterial oxygen content is CaO2= (1.37 X Hb X SaO2) + (0.003 X PaO2). From this equation, we can see that normoxemia does not necessarily guarantee adequate oxygen content. However, since the PaO2 contributes only 0.003 volume percent to the blood oxygen content, the most important factors in determining oxygen content become the hemoglobin concentration and the percent saturation. In order to interpret the results of pulse oximetry, we must bear in mind the shape of the oxygen hemoglobin dissociation curve, which is not linear and explains why the SpO2 is not a replacement for the SaO2 and the PaO2. Due to the shape of the curve, large decreases in PaO2 may be accompanied by only small changes in the SaO2 in areas other than the steep part of the curve, where a predictable correlation exists between SaO2 and PaO2. A patientís oxygen content may therefore drop precipitously before it is detected by pulse oximetry. Shifts in the oxyhemoglobin saturation curve also influence the relationship between PaO2 and SpO2. The information provided by the pulse oximeter is not a replacement for the PaO2, but is complementary to the PaO2. However, the pulse oximeter becomes an ideal continuous monitor of tissue oxygen delivery in the face of normal hemoglobin concentration, and normal types of hemoglobin (vs. methemoglobin and carboxyhemoglobin).

Potential for inaccuracy is a concern. Poorly perfused conditions caused by hypothermia, hypotension, altered vascular resistence, or the use of vasoactive drugs can reduce the pulsatile signal, which the PULSE oximeter depends upon for its ability to function. It may be necessary to change the sensor sites to obtain an optimal signal. Motion may interfere with the oximeter, which commonly happens in awake, agitated, or shivering patients. Ambient light can contaminate a light emitting diode signal, as can other light sources such as radiant warmers, etc. The oximeter sensor should be covered to reduce light contamination. The presence of any substance in the blood that absorbs light in the red or infrared spectrum may produce spurious SpO2 readings. Methylene blue, indigo carmine and indocyanine green can alter the SpO2. As mentioned earlier, both carboxyhemoglobin and methemoglobin absorb light in the red to infrared wavelengths and thus interfere with SpO2 readings. When methemoglobin is present, it tends to push the oximeter reading to a median of 85%. When carboxyhemoglobin is present, the SpO2 will be falsely high.

Each pulse oximeter has a specific set of probes that can be attached to various body parts. Each probe is specifically designed by the manufacturer for use with their own instrument; however, NellcorR sensors are over represented in the veterinary population. The most common type of probe is a clip, which can be placed on any fold of skin, including the lip, ear pinna, axilla, perineum, or tongue. In cats, a site that often works well is across the toe. Rectal probes are also available.

In the ICU, the pulse oximeter can be used for intermittent monitoring, or to provide a continuous real time read out of oxygen saturation. Continuous monitoring is particularly useful during general anesthesia or ventilation. The pulse oximeter can also be used to monitor changes in saturation when stressful procedures are being performed in patients with borderline respiratory function (BALs, tracheal washes, centesisí). It can allow us to determine whether additional oxygen supplementation is necessary or not. In addition, because the blood flow is pulsatile, the output of the pulse oximeter sensor can be electronically processed to give a pulse waveform and or heart rate. As such, the pulse oximeter becomes a useful monitor of peripheral pulse rate and regularity.

Normal PaO2 at room air is 80-100mmHg, corresponding to an SpO2 of >97%. Anesthetized veterinary patients are usually given 100% oxygen. The expected PaO2 when breathing 100% oxygen is >450mmHg and SpO2 shows correspondingly little change (>987%) based on the oxygen hemoglobin dissociation curve. At PaO2 of 100mmHg, hemoglobin is essentially 100% saturated. Thus, increasing the PaO2 to 500 only increases the dissolved oxygen content.

There are several instruments on the market. The instruments that provide a visual image of the pulse waveform, as well as a strength indicator may be the most useful monitoring tools. The development of reflectance instruments has permitted the application of oximetric sensors to alternative sites, including many flat surfaces, such as the chest and head. Pulse oximetry does not replace blood gas analysis, but can be very useful as a continuous, real time monitor to assess changes in oxygenation and trends in pulse perfusion. It is non-invasive, and does not require calibration. It has proved to be a useful, safe, and well-tolerated tool that can be applied to the majority of dogs and cats, awake or anesthetized.

End Tidal Capnometry and Capnography

Capnometry is the measurement and numeric representation of the CO2 concentration or partial pressure during inspiration and expiration. A capnogram is a continuous concentration-time display of the CO2 concentration sampled at a patientís airway during ventilation. Capnography is the continuous monitoring of a patientís capnogram. End tidal CO2 is the partial pressure of carbon dioxide in the expired air obtained at the end of expiration. This value approximates that of alveolar air, assuming that capillary blood and alveolar gas CO2 are in equilibrium, ventilation perfusion mismatches do not exist, tidal volumes are large enough to displace the dead space, and fresh gas flow is low enough to prevent dilution. End tidal CO2 is a product of three major determinants: the rate of CO2 production by the tissues, the rate of exchange of CO2 from the blood to the alveoli, and the rate of CO2 removal by alveolar ventilation.

Because carbon dioxide is a highly soluble gas, diffusing from air to liquid and back again occurs very quickly. Because of this solubility, the relationship between carbon dioxide and minute ventilation is a straight line. The higher the ventilation, the lower the CO2. Conversely, hypoventilation leads to high carbon dioxide levels as the gas is retained. Thus End tidal CO2 provides a clinical estimate of the alveolar and thus the arteriolar CO2. IF the gradient between the PaCO2 and the ETCO2 is small and constant, capnography provides a noninvasive, continuous, real time reflectance of ventilation.

In a single breath, air sampled during inspiration should represent room air and therefore contain virtually no carbon dioxide. As exhalation begins, the air passing the instrument initially represents dead space that has not been in contact with alveolar air, therefore containing virtually no carbon dioxide. As exhalation continues, alveolar air mixes with the dead space, with a resultant gradual increase in the amount of carbon dioxide measured by the instrument (upstroke of the capnogram curve). Eventually, the air passing the sampling port is alveolar air, and the partial pressure of carbon dioxide reaches a plateau, which is reported by the instrument as the end tidal carbon dioxide. As the patient begins to inspire, fresh gas is entrained, and there is a steep down stroke of the capnogram curve back to baseline.

Analysis of this waveform provides a wide range of information in addition to providing a means by which to estimate arterial carbon dioxide tensions and the adequacy of ventilation. The normal arterial carbon dioxide partial pressure in an awake, healthy animal is 35-45mmHg. If the end tidal CO2 is greater than 50mmHg in the anesthetized patient, this is highly specific for hypoventilation, which should be treated immediately. The End tidal CO2 is normally slightly lower than the PaCO2 by 6-8mmHg. This difference is known as the P(a-ET)CO2 gradient. The gradient can increase primarily owing to a decrease in PETCO2. Pathophysiologic reasons for the later include high mismatch, COPD, hypoperfusion, PTE. Endotracheal tube cuff leaks and accidental extubation can cause a decrease in PETCO2.

The physics behind end tidal capnography involves infrared absorbtion spectrophotometry, based on the fact that carbon dioxide absorbs infrared light. IF a known spectrum of infrared light is beamed through a sample of expired gas, the amount of infrared light that is absorbed will be proportional to the amount of carbon dioxide in the sampled gas. Sampling of end-tidal gas can occur by the use of mainstream or sidestream capnographs. Mainstream capnographs analyze end-tidal CO2 by using a transducer that is placed directly in line of the patientís breathing circuit. The transducer contains a light source and a photodetector. The response time for analysis using the mainstream method is very brief. Sidestream capnographs actively divert gas away from the patientís breathing circuit with a narrow sampling tube. Gas analysis occurs in the capnograph, which contains the light source and the photodetector. Lengthy sampling times are expected using sidestream capnographs due to distant sampling and processing times.

As stated above, the P(a-ET)CO2 gradient has been consistently reported as less than 6mmHg in healthy subjects. This gradient can increase primarily owing to a decrease in PET CO2. Pathophysiologic causes for a decrease in end tidal CO2 include the development of high ventilation perfusion units (obstruction or obliteration of the pulmonary vascular bed). If pulmonary perfusion decreases, the end tidal CO2 will also decrease because less CO2 is delivered to the alveolar-capillary interface.

The capnograph has been shown to assist in cardiopulmonary resuscitation efforts. If CPR efforts results in forward flow of blood, pulmonary perfusion will be regained and transport of carbon dioxide across the alveolar-capillary membrane will resume. Thus, with successful resuscitative effort, the capnograph will show a progressive rise in End Tidal CO2, and a capnogram tracing will be restored. Capnography has been employed to identify appropriate placement of endotracheal tubes, indirectly assess the PaCO2, monitor potential changes in perfusion and dead space, and even detect added CO2 (rebreathing, poor sodalime) in anesthesia circuits.

A sudden drop of the ET CO2 to near zero followed by the absence of a CO2 waveform is a potentially life-threatening problem that could indicate malposition of the endotracheal tube in the pharynx or esophagus, disruption of airway integrity, disruption of sampling lines, or a sudden cardiac arrest. Abrupt decreases in the ETCO2 are often associated with an altered cardiopulmonary status (embolism or hypoperfusion). Gradual reductions in ETCO2 often reflect decreases in PaCO2 that occur following increases in minute ventilation or a reduction of the metabolic rate. A slow rate of rise of the upswing is suggestive of either chronic obstructive pulmonary disease or acute airway obstruction. A normally shaped capnogram with an increase in end tidal CO2 suggests alveolar hypoventilation or an increase in CO2 production.

With these ideas in mind, it is easy to see how both pulse oximetry and end tidal capnography can be useful tools in the management of both anesthetized and awake patientsí respiratory and cardiovascular functioning. Anesthesiologists are fully aware of the importance of these monitoring modalities, and as practitioners, it behooves us to make full use of them for the enhancement of quality care for our patients, both those critically ill and those undergoing elective procedures. Further development of the scope of monitoring in general is needed to not only ensure the accuracy and clinical efficacy of these tools, but to broaden their applications and make them useful counterparts to assessment of blood gas analysis. The potential for diagnostic interventions using both pulse oximetry and capnography is great; utilization of these modalities with pressure volume flow loops, spirograms, radiographs, and bronchoalveolar endoscopy/sampling will probably eventually assist in "solving" some of the more challenging respiratory disease cases within our profession.

References available upon request from author