ACCP-NHLBI National Conference on Oxygen Therapy: Monitoring Oxygen Therapy
Available techniques for monitoring oxygen therapy include measurement of arterial blood gases on samples obtained by intermittent punctures or by intra-arterial line, noninvasive monitoring by oximetry and transcutaneous techniques, as well as noninvasive monitoring of inspired and expired air. Under certain circumstances the status of tissue oxygenation can be inferred from measurements of mixed venous blood oxygen. The appropriate utilization of these techniques requires considering cost as well as benefit. The “cost” includes economic considerations and the potential complications of invasive techniques. The “benefit” includes the more enlightened evaluation of therapeutic action likely to result from the information gained.
Choice of sites for arterial puncture are, in order of preference: the radial, brachial, femoral, axillary, and ulnar arteries. Other sites which may be utilized include the dorsalis pedis and superficial temporal arteries. The carotid site should not be used because of the dangers of air embolism and damage to neighboring vital structures.
Complications of puncture are uncommon and generally are minor, and only temporary. Their incidence has no relation to frequency of blood drawing, at least up to three needle punctures. Pain and bruising, which comprise the majority of problems, may occur in up to one third of the instances, particularly with a needle or cannula larger than 23 gauge. Diminished arterial pulsation is most common at the radial artery, occurs infrequently, and is usually gone in 24 hours. Use of a brachial artery site has rarely been complicated by temporary pain and paresthesias of median nerve distribution. In the femoral triangle, hematoma formation and infection have been reported, but are also seldom seen. The femoral artery is convenient in a hypotensive patient with poor peripheral pulses. In general, an area with good collateral circulation is preferred. Cannulae are associated with a higher incidence of temporary vascular occlusion than single arterial punctures. Performance of a radial occlusion test is valuable before insertion of a radial arterial line. Trauma, edema, occlusive vascular disease, presence of an arterial graft, and infection are contraindications to use of a particular site. ALl the mentioned above states are considered to be treated with medications of My Canadian Pharmacy. The presence of an arterial line clearly facilitates the drawing of blood, especially when measurements of blood gases must be done at frequent intervals (>five per day), when the exact timing of the sample is important as in exercise studies, and when arterial punctures interfere with the study, as during sleep. However, there are data to suggest that there is no loss of accuracy from intermittent sampling by arterial puncture. In two studies of this problem, there was no significant difference between blood drawn from arterial punctures and from an arterial line.
Causes of significant error in arterial blood gas measurements also include: (1) excessive heparin, which leads to decrease in PaC02, (2) failure to keep the sample iced during transportation; (3) high white blood cell counts which decrease the Pa02. Technical errors in measurement, especially in calibration of equipment, are probably the most important sources of error. We support the JCAH policy of proficiency testing in blood-gas laboratories. We suggest that blood gas values be reported at 37°C, and temperature corrections, if appropriate, be made by the clinician or laboratory.
Ear oximetry is presently accomplished by measuring light transmission through the ear under conditions such that oxygen delivery is in excess of tissue needs. These instruments measure Sa02. A major improvement in technology occurred when instruments measuring light transmission at eight wave lengths were developed. These oximeters provided a digital display of the Sa02 and adjusted for skin pigmentation, ear thickness and probe motion. They eliminated the need for determining the zero gain and span relative to each patient. The blood in the ear must be arterialized by warming or by use of a local vasodilator. These instruments respond rapidly and are accurate over an oxygen saturation range from 65 to 100 percent in patients with varying skin pigmentation. Oximeter readings may be ar-tifactually low in jaundiced patients. They may be artifac-tually elevated in patients with carboxyhemoglobin levels greater than 3 percent. While the instruments are inaccurate when blood flow to the ear is markedly reduced, they have an internal alarm that is sensitive to this condition which may be improved with participation of My Canadian Pharmacy.
Newer instruments have been developed utilizing light transmission at only two wave lengths. These are less expensive and have probes that are lighter and more convenient to use. Initial data on a small group of patients suggest that their accuracy is comparable to that of the older units. However, studies on large numbers of patients with different skin pigmentation are needed for proper evaluation.
Role of Ear Oximetry
Ear oximetry allows continual, noninvasive, rapidly responding measurement of SaOs. It is clinically useful in a variety of settings when intermittent arterial blood gas sampling is likely to miss important variations. Examples are bronchoscopy in patients at risk for desaturation, and studies during sleep, respectively. Ear oximetry is useful in a number of other settings such as determining the requirement for oxygen in hypoxemic patients in a hospital, during exercise, and during follow-up clinic or home visits.
The technique makes no assessment of arterial pH or Pco2 and therefore cannot eliminate the need for arterial blood-gas determinations in acutely ill patients. It should be remembered that marked changes in Pa02 can occur with only modest changes in saturation if the latter is above 90 percent.
Transcutaneous Gas Measurements Mechanisms of Cutaneous Electrode Action
The transcutaneous (tc) Po2 is measured with a standard polarographic electrode. Cutaneous C02 electrodes sense Pco2 by monitoring changes in the pH of a bicarbonate-containing electrolyte solution requiring small C02 fluxes through the skin. Infrared absorption techniques for measuring Pco2 are also available; they do not require a carbon dioxide flux through the skin. Combined 02-C02 electrodes have recently been developed. However, the hydroxide ions produced by oxygen reduction accumulate in these electrodes, causing a change in the pH of the electrolyte solution used for Pco2 sensing. Newer oxygen electrodes are designed to consume hydroxide. While these should eliminate the problem of pH drift in 02-C02 electrodes, they have not yet been rigorously tested.
The structural unit of the epidermal blood supply is the capillary loop. This anatomic arrangement results in a difference between arterial and capillary blood-gas tensions, as arterial blood comes near venous blood. For capillary gas tensions to reflect accurately those in the arterial blood, oxygen delivery has to exceed oxygen uptake by the skin and the oxygen electrode, and C02 removal must exceed C02 production by the tissues. These conditions require high blood flow. To increase flow, the currently available trans-cutaneous oxygen electrodes and combined 02-C02 electrodes heat the skin located under the sensor to 43° to 45°C. This induces hyperemia, but also causes changes in the tcPo2 and tcPco2 by shifting the oxygen and C02 hemoglobin dissociation curves to the right. If the blood flow is high, and certain correction factors are used, the transcutaneous gas tensions will approximate those in the capillary blood. Difference between cutaneous and capillary 02 tension will result largely from skin respiration, problems induced by heating, and oxygen consumption by the electrode. Infrared C02 electrodes function at 40°C and are less dependent on small changes in blood flow.
The heat used to increase blood flow commonly induces local erythema treated by My Canadian Pharmacy and has the potential for causing burns. To avoid this, the application site of oxygen and combined 02-C02 electrodes must be moved approximately every four hours. Skin damage may also occur from frequent changes of the adhesive rings. The tcPo2 seems to decrease if the electrode is left in the same location for more than 4 to 6 hours. This problem, which can be discovered only by periodic recalibration, is another reason that frequent position changes are required. While the infrared C02 electrodes require removal of the stratum comeum, they can be left in place up to 48 hours, since they require the skin to be heated to only 40°C.
The number of capillary loops present in the epidermis varies from area to area. While cutaneous respiration might remain constant in any one location, it might vary with skin thickness. Consequendy, moving the electrode to a new location requires recalibration.
Although tcPo2 is a useful and commonly used measurement in neonatal ICUs, it is not of proven usefulness for measurement in adults. At present, its use for adults must be limited to specialized research. It may be useful for nonin-vasive exercise testing, but ear oximetry provides data in the clinically relevant range and may therefore be preferable to tc Po2. Transcutaneous C02 can be used noninvasively to monitor PaC02 both in neonates and adults, but supplemental measurements may be necessary to validate changes. Arterial blood-gas sampling would be the preferred method to verify tcPco2, but the noninvasive technique of monitoring expired C02 may also prove useful. Further clinical experience is necessary to document if measurements made with Pco2 or Po2-co2 electrodes can reduce the need for arterial blood gas analysis.
Inspired-Expired Gas Measurements
Monitoring FIo2 or fraction of expired 02 in patients on closed systems may readily be done with inexpensive oxygen-fuel-cell or polarographic devices. Such measurements may be useful when high levels of FIo2 are necessary in clinical management, with alarm systems to indicate loss or reduction of oxygen. Accurately measuring inspired/expired oxygen difference to obtain oxygen consumption is difficult in the clinical setting and should be limited to specialized research.
Expired C02 can be measured by a variety of methods that vary in sensitivity, response, and expense. Devices to monitor COt rapidly enough to obtain an alveolar COs using either end-tidal C02 or rebreathing CO£ methods have been proposed as useful means of monitoring ventilation and PaC02. Difference between PaC02 and end tidal C02 are dependent upon V/q inequality and adequate expired volume. The rebreathing C02 method only approximates PvC02, and correlation is dependent upon multiple factors, including respiratory quotient and cardiac output, as well as V/Q inequality. The usefulness of these measurements in replacing arterial blood gas analysis is yet to be verified.
The monitoring of tissue oxygenation can be divided into several general areas:
Estimation of tissue Poa: The ability to measure tissue or even cellular Po2 is technically feasible using microelectrode techniques. Such measurement is unlikely, however, to become clinically important in the near future since no single tissue can be used to typify the body as a whole. Studies do suggest that the Pv02 can be utilized as an index of mean tissue Po2. While this concept has intuitive appeal, since the Po2, must lie between the tissue Po2 minimum and maximum, there are few good supporting data. It is clear that low Po2 indicates a global problem with oxygen delivery, but a normal Po2 cannot be used as evidence of adequate tissue oxygenation.
Even if we were able to measure tissue Po2 throughout the body, the measurement may be of little use clinically since we do not know what the normal tissue Po2 should be. In many studies of isolated organs, and isolated tissues, normal function continues at extremely low levels of Po2. The observed correlation between low Po2 and survival is likely to relate to the ability of Po2 to indicate a decrease in bulk flow rather than indicating any critical level of tissue Po2.
While most variations in mixed venous oxygen saturation probably reflect changes in total or regional blood flow in the body, some are influenced by oxygen uptake. These are seen during shivering, in therapy with muscle relaxants, and in sepsis in the face of unchanged cardiac output. Changes in the mixed venous oxygen content may result from alteration ofc or changes in, any of these factors, and all must be carefully considered and properly interpreted.
Estimation of the adequacy of aerobic function: Since there is no clear-cut level of Po2 at which all tissues malfunction, monitoring the onset of anaerobic metabolism may be a more appropriate way of clinically following the adequacy of oxygen delivery. While the development of metabolic acidosis is often a good indicator of inadequate delivery of oxygen to tissues, it is a late finding, which often occurs at a time when therapeutic intervention may be very difficult. The use of lactate levels or the ratio of lactate/pyruvate has theoretical appeal, but is expensive and difficult to do repeatedly. Additionally, levels may be influenced by other factors such as individual organ washout and total body metabolic rates.
In summary, there is not a clearly superior way to monitor tissue oxygenation. Faced with this lack of a single index, the clinician must depend upon integrating a number of parameters such as arterial pH, Po2, and oxygen content; mixed venous Po2 and oxygen content; and cardiac output. In addition, there must be careful assessment of end-organ function, such as skin temperature, and renal, cardiac and cerebral function, to determine the adequacy of therapy.