Arterial Puncture and Interpretation

Results obtained from sampling arterial blood gas (ABG) are the cornerstone in the diagnosis and management of oxygenation and acid-base disturbances. ABGs are considered the “gold standard” of gas exchange analysis, against which all other methods are compared.

Arterial puncture involves drawing blood from a peripheral artery (radial, brachial, femoral, or dorsalis pedis) through a single percutaneous needle puncture (Figure 18-2). The radial artery is the preferred site for arterial blood sampling for the following reasons:

Other sites (brachial, femoral, and dorsalis pedis) are riskier and should be used only by clinicians specifically trained in their use. Likewise, arterial puncture in infants (through either the radial or the temporal artery) requires advanced training. Arterial cannulation sites with indwelling catheters include radial, brachial, femoral, dorsalis pedis, umbilical (in neonates), and axillary arteries. The focus here is on radial artery puncture.

To guide practitioners in providing quality care, the American Association for Respiratory Care (AARC) has published Clinical Practice Guideline: Sampling for Arterial Blood Gas Analysis.³ Complementary recommendations have been published by the National Committee for Clinical Laboratory Standards.⁴ Modified excerpts from the AARC guideline appear in Clinical Practice Guideline 18-1.

Procedure

Box 18-2 outlines the basic procedure for radial artery puncture of adults. Before radial artery puncture is performed, a modified Allen test (Figure 18-3) is recommended. The test is normal (indicating adequate collateral circulation) if the palm, fingers, and thumb flush pink within 5 to 10 seconds after pressure on the ulnar artery is released. A normal test result indicates the presence of collateral circulation but may not predict the development of complications after radial artery puncture or cannulation.

Box 18-2

Procedure for Radial Artery Puncture

• Check the medical record to (1) confirm the order and indications and (2) determine the patient’s primary diagnosis, history (especially bleeding disorders or blood-borne infections), current status, respiratory care orders (especially oxygen therapy or mechanical ventilation), and anticoagulant or thrombolytic therapy.

• Confirm steady-state conditions (20-30 minutes after changes).

• Obtain and assemble necessary equipment and supplies.

• Wash hands and don barrier protection (e.g., gloves, eyewear).

• Identify the patient using current patient safety standards.

• Explain the procedure to the patient.

• Position the patient, extending the patient’s wrist to approximately 30 degrees.

• Perform a modified Allen test, and confirm collateral circulation.

• Clean site thoroughly with 70% isopropyl alcohol or an equivalent antiseptic.

• Inject a local anesthetic subcutaneously and periarterially (wait 2 minutes for effect).*

• Use a preheparinized blood gas kit syringe, or heparinize a syringe and expel the excess (fill dead space only).

• Palpate and secure the artery with one hand.

• Insert the needle, bevel up, through the skin at a 45-degree angle until blood pulsates into the syringe.

• Allow 1 ml of blood to fill syringe (the need to aspirate indicates a venous puncture).

• Apply firm pressure to puncture site with sterile gauze until the bleeding stops.

• Expel any air bubbles from the sample, and cap or plug the syringe.

• Mix the sample by rolling and inverting the syringe.

• Place the sample in a transport container (ice slush) if specimen is not to be analyzed within 10-30 minutes.

• Dispose of waste materials and sharps properly.

• Document the procedure and patient status in the chart and on the specimen label.

• Check the site after 20 minutes for hematoma and adequacy of distal circulation.

---

^*Optional.

From Malley WJ: Clinical blood gases: application and noninvasive alternatives, Philadelphia, 1990, Saunders; Shapiro BA, Peruzzi WT, Kozelowski-Templin R: Clinical application of blood gases, ed 5, St Louis, 2005, Mosby.

The modified Allen test has been a widely used clinical method to assess adequacy of ulnar artery collateral blood flow despite the lack of evidence that it can predict ischemic complications in the setting of complete radial artery occlusion.⁵ The criteria for an abnormal test result are not agreed on, which renders the significance of an abnormal test unclear. The test result may be inaccurate in predicting postcannulation hand ischemia, has poor interrater reliability, and is known to yield a high incidence of false normal and abnormal results. The modified Allen test cannot be performed on most critically ill patients who are either uncooperative or unconscious. In addition, prior radial artery cannulation, severe circulatory insufficiency, wrist or hand burns, or jaundice makes interpreting the results difficult. Performance of a modified Allen test before radial artery puncture or cannulation should not be considered a “standard of care,” but the need for its use and appropriate application should be well recognized.⁶

In patients who have undergone previous radial artery cannulation, the modified Allen test can provide documentation of possible arterial thrombosis and should be used to direct catheter placement. In that circumstance, it is imprudent to ignore totally the utility of the modified Allen test, especially if another arterial site is available for cannulation.⁷

In most cases, a sample volume of 0.5 to 1 ml of blood is adequate. The actual sample volume needed depends on the following: (1) the anticoagulant used, (2) the requirements of the specific analyzer used, and (3) whether other tests are to be performed on the sample.

The following rules for careful handling of the needle help avoid transmission of blood-borne diseases:

• Never recap a used needle without a safety device; never handle a used needle using both hands; never point a used needle toward any part of the body.

• Never bend, break, or remove used needles from syringes by hand.

• Always dispose of used syringes, needles, and other sharp items in appropriate puncture-resistant sharps containers.

Indwelling Catheters (Arterial and Central Venous Pressure and Pulmonary Artery Lines)

Indwelling catheters provide ready access for blood sampling and allow continuous monitoring of vascular pressures, without the traumatic risks associated with repetitive percutaneous punctures. However, infection and thrombosis are more likely with indwelling catheters than they are with intermittent punctures.

The most common routes for indwelling vascular lines are a peripheral artery (usually radial, brachial, or less commonly dorsalis pedis and axillary) or femoral artery, a central vein (usually the vena cava), and the PA. In neonates, the umbilical artery is cannulated for arterial blood sampling. Table 18-2 summarizes the usefulness of these various sites in providing relevant clinical information. Chapter 46 provides details on the use of these systems for hemodynamic pressure and flow monitoring.

Equipment

Figure 18-4 shows the basic setup used for an indwelling vascular line, in this case, a brachial artery catheter. The catheter connects to a disposable continuous-flush device (Delta-flow; Utah Medical Products, Midvale, UT). This device keeps the line open by providing a continuous low rate of flow (2 to 4 ml/hr) of intravenous saline solution through the system.

Heparinized saline flush solution has been commonly used with indwelling vascular catheters. However, it has been shown that heparin does not significantly improve arterial catheter function, extend the duration of use, or decrease the number of manipulations required. Additionally, results of coagulation studies can be affected by heparinized flush solution, and unnecessary exposure to heparin may increase the risk of heparin-induced thrombocytopenia.¹⁰ Because arterial pressures are much higher than venous pressures, the intravenous bag supplying these systems must be pressurized, usually by using a hand bulb pump. A strain-gauge pressure transducer connected to the flush device provides an electrical signal to an amplifier or monitor, which displays the corresponding pressure waveform.

Procedure

Access for sampling blood from most intravascular lines is provided by a three-way stopcock (Figure 18-5). Equipment and supplies are the same as specified for arterial puncture, with the addition of a second “waste” syringe. Box 18-4 outlines the proper procedure for taking an arterial blood sample from a three-way stopcock system.

Box 18-4   Procedure for Sampling Arterial Blood from an Indwelling Catheter

• Check the medical record to affirm order (as per arterial puncture).

• Confirm steady-state conditions (20-30 minutes after changes).

• Obtain and assemble needed equipment and supplies.

• Wash hands and don barrier protection (e.g., gloves, eyewear).

• Identify the patient using current patient safety standards.

• Explain the procedure to the patient.

• Attach the waste syringe to the stopcock port.

• Position the stopcock so that blood flows into the syringe and the IV bag port is closed.

• Aspirate at least 1-2 ml, or five to six times the tubing volume, of fluid or blood.

• Reposition the stopcock handle to close off all ports.

• Disconnect and properly discard waste syringe.

• Attach new heparinized syringe to the sampling port.

• Position the stopcock so that blood flows into the sample syringe and the IV bag port is closed.

• Fill syringe with 1 ml of blood.

• Reposition the stopcock handle to close off the sampling port and open the IV bag port.

• Disconnect the syringe, expel air bubbles from sample, and cap or plug the syringe.

• Flush the line and stopcock with the IV solution.

• Mix the sample by rolling and inverting the syringe.

• Confirm that the stopcock port is open to the IV bag solution and catheter.

• Confirm undampened pulse pressure waveform on the monitor graphic display.

• Place the sample in a transport container (ice slush) if specimen is not to be analyzed within 10-30 minutes.

• Dispose of waste materials properly.

• Document the procedure and patient status in the chart and on the specimen label.

IV, Intravenous.

The procedure is slightly different when obtaining mixed venous blood samples from PA catheters because PA catheters have separate sampling and intravenous infusion ports and a balloon at the tip is used to measure pulmonary capillary wedge pressure. The clinician must ensure that the balloon is deflated and withdraw the sample slowly (e.g., approximately 3 ml/min or 1 ml in 20 seconds). If the clinician fails to deflate the balloon or withdraws the sample too quickly, the venous blood may be “contaminated” with blood from the pulmonary capillaries. The result is always a falsely high O₂ level. In addition, close attention must be paid to the infusion rate through the catheter. Rapid flow of IV fluid can dilute the blood sample and affect O₂ content measurements.

Capillary Blood Gases

Capillary blood gas sampling is used as an alternative to direct arterial access in infants and small children. Properly obtained capillary blood from a well-perfused patient can accurately reflect and provide clinically useful estimates of arterial pH and PCO₂ levels.⁶ However, capillary PO₂ is of no value in estimating arterial oxygenation, and O₂ saturation by pulse oximetry must also be evaluated when a capillary blood gas sample is obtained. Respiratory therapists (RTs) must exercise extreme caution when using capillary blood gases to guide clinical decisions. Direct arterial access is still the preferred approach for assessing gas exchange in infants and small children with severe acute respiratory failure.

Capillary blood values are meaningful only if the sample site is properly warmed. Warming the skin (to approximately 42° C) causes dilation of the underlying blood vessels, which increases capillary flow well above tissue needs. Blood gas values resemble the values in the arterial circulation; this is why a sample obtained from a warmed capillary site is often referred to as arterialized blood. It has been shown that capillary blood samples from the earlobe reflect arterial PCO₂ and PO₂ better than samples drawn from a finger stick.¹¹ The posterior medial or lateral curvature of the heel is the recommended site for capillary puncture specimens in infants less than 1 month old to avoid nerve and bone damage.

To guide practitioners in providing quality care, the AARC has published Clinical Practice Guideline: Capillary Blood Gas Sampling for Neonatal and Pediatric Patients.¹² Modified excerpts from the AARC guideline appear in Clinical Practice Guideline 18-2.

Instrumentation

Many instrumentation companies manufacture laboratory blood gas analyzers. Although available in a range of designs, these devices typically share the following key components:

Measurement of the three primary parameters—pH, PCO₂, and PO₂—is accomplished using three separate electrodes. To measure PO₂, blood gas analyzers use the Clark polarographic electrode (see Figure 18-1).

The pH electrode consists of two electrodes or half-cells (Figure 18-6). The measuring half-cell contains a silver–silver chloride rod surrounded by a solution of constant pH and enclosed by a pH-sensitive glass membrane. As the sample passes this membrane, the difference in H⁺ concentration on either side of the glass changes the potential of the measuring electrode. The reference half-cell (here mercury–mercurous chloride) produces a constant potential, regardless of sample pH. The difference in potential between the two electrodes is proportional to the H⁺ concentration of the sample, which is displayed on a voltmeter calibrated in pH units.

To measure PCO₂, blood gas analyzers use the Severinghaus electrode, which is essentially a pH electrode exposed to an electrolyte solution that is in equilibrium with the sample through a CO₂-permeable membrane. As CO₂ diffuses through this membrane into the electrolyte solution, it undergoes the following hydration reaction:

< ?xml:namespace prefix = "mml" />CO2+H2O↔H2CO3↔H++HCO3−

The greater the partial pressure of CO₂, the more H⁺ produced by this reaction, and the more pH of the electrolyte solution changes. The measuring electrode detects the pH change as a change in electrical potential, which is proportional to the PCO₂ of the sample.

In addition to electrochemical electrodes, a sensor technology based on optical fluorescence and the process of photoluminescence is available to measure pH, PCO₂, and PO₂. Optical fluorescence sensors use fluorescent dyes that are illuminated with light of a specific wavelength. The illuminating light is transmitted, reflected, absorbed, and reemitted to varying degrees depending on the concentration of O₂, CO₂, and hydrogen ions. These photochemical alterations in light illumination are used to calculate pH, PCO₂, and PO₂ values. Compared with electrochemical sensors, optical sensors do not require electrical connections, making them less affected by electrical interference and drift. The sensor reactions are reversible so that the analyte is not consumed. These features make them easier to miniaturize for intravascular measurement of blood gases.16,17

Procedure

To provide accurate and clinically useful data, blood gas analysis must be performed as follows:

Prior discussion addressed how to avoid preanalytic errors. Subsequent discussion focuses on blood gas quality control and key elements involved in the analysis procedure.

Box 18-6 outlines the steps commonly used in most established procedures for laboratory blood gas analysis. One should always refer to the manufacturer’s literature for the particular steps to use with a specific analyzer.

Rigorous application of U.S. Centers for Disease Control and Prevention (CDC) standard precautions is essential. In addition, the Occupational Safety and Health Administration (OSHA) requires personnel to wear gloves when handling all laboratory specimens. Although splashes are rare during analysis, some manufacturers provide extra protection by mounting splash shields on their instruments. If splashes are anticipated, the operator can wear a face shield.

Waste fluids are potentially infectious and should be handled as if they were blood samples. In addition, the National Committee for Clinical Laboratory Standards recommends adding a strong disinfectant, such as 2% glutaraldehyde or a 1 : 4 solution of sodium hypochlorite, to the waste container of the instrument either during use or before disposal.

Quality Assurance

Quality patient care depends on consistently accurate blood gas results. Modern laboratory analyzers are often automated, computer-controlled, self-calibrating systems. This sophistication has led to the false assumption that accurate results are “automatic,” with clinicians needing only to input the sample properly and record the results. Nothing could be further from the truth. As with all diagnostic laboratory procedures, the accuracy of blood gas testing depends on rigorous quality control.

The Clinical Laboratory Standards Institute (CLSI), formerly the National Committee for Clinical Laboratory Standards (NCCLS), establishes guidelines and standards for blood gas analysis and quality assurance. Government regulatory agencies collaborate to update the Clinical Laboratory Improvement Amendments (CLIA) that establish proficiency testing requirements.¹⁸ Although an in-depth review of laboratory quality control is beyond the scope of this text, all RTs must understand the key elements.¹⁹

Figure 18-7 depicts the key components of laboratory quality control. A brief description of each element follows.

Automated Calibration

Calibration is the only fully automated element of blood gas quality control. Blood gas analyzers regularly calibrate themselves by adjusting the output signal of each electrode when exposed to media having known values. In most units, the media used to calibrate the gas electrodes are precision mixtures of O₂ and CO₂. For the pH electrode, standard pH buffer solutions are used. Calibration media must meet the requirements set by nationally recognized standards organizations. Users are responsible for ensuring that calibration media are properly stored and that in-use life and expiration dates are strictly enforced.

Calibration is performed to ensure that the output of the analyzer is both accurate and linear across the range of measured values. Parameters must be measured with known input values representing at least two points, usually a low and a high value. Figure 18-8 shows a typical two-point calibration procedure. In this example, the instrument’s initial precalibration response indicates that the output readings are consistently higher than the actual input, with this positive bias worsening at higher levels. Calibration is performed first by adjusting the offset (or balance) of the instrument so that the low output equals the low input (in this case zero). Next, the gain (or slope) of the device is adjusted to ensure that the high output equals the high input. When both offset and gain are adjusted against known inputs, the instrument is properly calibrated and can undergo calibration verification with control samples.

Internal Statistical Quality Control

Internal quality control takes calibration verification a step further by applying statistical and rule-based procedures (Westgard rules)20,21 to help detect, respond to, and correct instrument error. In one common approach, the results of control media analyses are plotted on a graph and compared with statistically derived limits, usually ±2 standard deviation (SD) ranges (Figure 18-9). Control results that fall outside these limits indicate analytic error.

There are two categories of analytic error: (1) random error and (2) systematic error. Random error is observed when sporadic, out-of-range data points occur (see Figure 18-9, point A). Random errors are errors of precision or, more precisely, imprecision. Conversely, either a trending or an abrupt shift in data points outside the statistical limits (see Figure 18-9, point B) is sometimes observed. This phenomenon is called systematic error or sometimes bias. Bias plus imprecision equals total instrument error, or inaccuracy. Table 18-3 outlines the major factors causing these two types of error and suggests some common corrective actions.

TABLE 18-3

Correction of Analytic Errors

Error Type	Common Contributing Factors	Common Corrective Actions
Imprecision (random) errors	Statistical probability	Rerun control
	Sample contamination	Repeat analysis on different instrument
	Sample mishandling
Bias (systematic) errors	Contaminated buffers	Perform function check of suspected problem area
	Incorrect gas concentrations	Repair or replace failed components
	Incorrect procedures
	Component failure

Point-of-Care Testing

Point-of-care testing takes blood gas analysis from the specialized laboratory to the patient’s bedside.²⁴ Point-of-care testing reduces turnaround time, which should improve care and reduce costs. Theoretically, cost savings can be accrued by eliminating delays in therapy, decreasing patient length of stay in the hospital and emergency department.²⁵ Additional cost savings may occur if point-of-care testing decreases the need for specialized laboratory personnel. Point-of-care testing is used increasingly in the hospital and physician office settings.²⁶

Instrumentation

Figure 18-10 shows a typical point-of-care blood gas analyzer (GEM 4000; Instrumentation Laboratory, Bedford, MA). In addition to blood gas analysis, such devices can be used to measure several chemistry and hematology parameters, including serum electrolytes, blood glucose levels, blood urea nitrogen, hematocrit, hemoximetry, lactate, bilirubin, and prothrombin and partial thromboplastin times.

These devices are portable, and some can perform 900 tests on a single set of batteries. They typically include a display screen for accessing menu functions and viewing results. Most devices include a simple keypad or touch screen for data and command entry. Analysis occurs using disposable cartridges or inside a chamber in the body of the unit.

Some devices employ single-use sample cartridges that differ according to the array of tests being performed. Each cartridge contains the necessary calibration solution, a sample handling system, a waste chamber, and miniaturized electrochemical or photochemical sensors. The cartridge system requires no operator oversight because it is self-calibrating and disposable after a single use. After self-calibration and introduction of the sample into the cartridge, the sensors measure the concentration of the analytes and conduct their output signal through conductive contact pads to the analyzer microprocessor. Test results usually are ready within 90 seconds. Waste management involves simple removal and proper disposal of the analysis cartridge.

Other devices use self-contained multiuse cartridge packs that include all testing components, are maintenance-free, and incorporate automated quality control management systems. Multiuse cartridges are typically replaced every 30 days or when testing components are used up.

Instrumentation

Figure 18-11 shows a simplified diagram of a transcutaneous blood gas monitor sensor. Included are a heating element and two electrodes, one for O₂ and one for CO₂. These electrodes are similar in design to the electrodes found in bench-top analyzers. However, instead of measuring gas tensions in a blood sample, transcutaneous electrodes measure PO₂ and PCO₂ in an electrolyte gel between the sensor and the skin. When properly set up, the response time for these electrodes is 20 to 30 seconds, a bit slower than the response time for pulse oximetry.

Figure 18-12 shows a transcutaneous monitor with digital signal processing. A Severinghaus-type PtcCO₂ electrode and two-wavelength reflectance SpO₂ are combined into a single sensor. The sensor can be applied to the skin surface in neonates and infants or to the earlobe of pediatric and adult patients for combined noninvasive monitoring of ventilation and oxygenation.

Procedure

Box 18-7 outlines the basic procedure for setting up a transcutaneous blood gas monitor. The most common sites for electrode placement in infants and children are the abdomen, chest, and lower back. Once the electrodes are properly set up, the clinician should compare the monitor readings with the readings obtained with a concurrent ABG. Consistency between values validates monitor performance under the existing conditions. This validation should be repeated anytime the patient’s status undergoes a major change. During validation studies of patients with anatomic shunts, the electrode site and arterial sampling site should be on the same “side” of the shunt.

Problem Solving and Troubleshooting

Transcutaneous blood gas monitoring is a complex and labor-intensive activity that requires ongoing training and careful quality control. Table 18-5 lists the major factors that can affect the accuracy or limit the performance of a transcutaneous monitor. In terms of technical limitations, the lengthy stabilization time needed by transcutaneous monitors precludes their use during short procedures or in emergencies. In such cases, the pulse oximeter is a better choice.

Sensors must be calibrated and maintained using methods similar to the methods described for bench-top analyzers. Improper calibration yields erroneous patient information. Improper calibration can be difficult to detect on some systems. Meticulous care of the sensor membranes is also essential for proper maintenance.

Because the sensor is heated, clinicians must take care to avoid thermal injury to the patient’s skin. Thermal injury can be avoided by (1) careful monitoring of sensor temperature (the safe upper limit is approximately 42° C) and (2) regularly rotating the sensor site.

Proper sensor-electrolyte contact is essential, as is proper application to the skin surface. A loosely applied sensor may have air leaks or may become dislodged. In either case, the resulting measurements would approach those in room air:

PO2=159mm Hg

PCO2=0mm Hg

Conversely, excessive pressure on the sensor compresses the underlying capillaries and produces a falsely low PtcO₂. Even with proper application and placement, PtcO₂ measures can vary by 10% at different sites, with values from the extremities generally being lower than values obtained from the chest or abdomen.³⁹

When arterial and transcutaneous blood gas values are inconsistent with each other or with the clinical status of the patient, the clinician should explore possible causes before reporting any results. Often, discrepancies can be reduced by switching the monitoring site or recalibrating the instrument. If these steps fail to resolve the inconsistencies, the clinician should recommend an alternative method for assessing gas exchange, such as pulse oximetry or more frequent ABG analysis.

Instrumentation

Rather than using electrochemical electrodes, miniaturized optical fluorescence sensors, called optodes, measure blood gas parameters by photochemical reactions and changes in the intensity of light transmission through optical fibers. Figure 18-13 shows how optodes are combined (with a thermocouple) at the tip of a flexible fiberoptic catheter that is inserted into a peripheral artery. Several optode-based systems have been developed and validated for commercial use for continuous intraarterial blood gas assessment and monitoring of PaO₂ during cardiopulmonary bypass.^40,41

Clinical Performance and Usefulness

Despite technical improvements, the actual performance of in vivo blood gas monitors falls short of the clinical requirements previously specified. Compared with standard blood gas analysis, accuracy is improved.42,43 However, concerns about bias in PO₂ measurement,⁴⁴ sensor accuracy over extended periods in divergent patient groups,⁴⁵ the need for femoral artery insertion because of blood pressure waveform dampening in the radial artery,⁴⁶ high acquisition cost of the monitor and sensor catheters, dedication of a monitor to a single patient,⁴⁷ and lack of evidence showing an impact on patient care have prevented widespread adoption and use of this technology.

Instrumentation and Procedure

Both optode-based and electrochemical-based systems have been developed. Figure 18-14 depicts an optode-based, ex vivo blood gas monitoring system. The optodes are located in a sensor cassette inserted in-line with the arterial catheter near the patient’s wrist. To measure pH, PCO₂, and PO₂, the system is closed to the intravenous fluid source at the stopcock (Figure 18-14, A). Subatmospheric pressure is created in the syringe attached to the stopcock (see Figure 18-14, A), which functions as an aspirating reservoir (Figure 18-14, B); this causes arterial blood to flow into the sensor cassette for analysis. During analysis, the stopcock (see Figure 18-14, A) is returned to its original position (off to the aspirating syringe). The connection through the line to the pressure transducer is restored, and blood pressure monitoring is able to continue. Blood gas parameter results are displayed in approximately 1 to 2 minutes. When analysis is complete, the blood sample is returned to the patient by emptying the aspirating reservoir (see Figure 18-14, B) and flushing the system through the flow valve (Figure 18-14, D).

Clinical Performance and Usefulness

In clinical trials, ex vivo on-demand systems performed as well as laboratory blood gas analyzers in adults, neonates, and infants.48–50 Measurements could be obtained every 3 to 5 minutes from a peripheral or umbilical artery, and the intermittent errors commonly associated with in vivo systems were not observed. A commercially available system for blood gas monitoring in neonates and infants also incorporates electrolyte, hematocrit, and Hb measurements.⁵⁰

Ex vivo blood gas monitoring systems share many of the requirements, advantages, and disadvantages related to in vivo systems stated in the previous section. Further justification of the costs associated with monitoring weighed against potential patient benefit are needed before widespread use of this technology occurs.

Instrumentation

Both electrochemical and optical fluorescence tissue O₂ probes have been developed for clinical use and research applications. Figure 18-15 shows a Clarke-type polarographic sensor and its insertion into brain tissue through an intracranial bolt. Optode probes capable of monitoring tissue pH, CO₂, and O₂ have also been developed.⁵²

Instrumentation

Figure 18-17 is a simplified diagram showing the key components of a laboratory hemoximeter. Light generated by a thallium cathode lamp passes through a series of lenses and filters, yielding the specific wavelengths needed for analysis. A beam splitter divides the light into two portions, directing one through a reference solution and the other through a sample chamber, or cuvette. Photodetection sensors measure the amount of light transmitted through these two sources. By comparing the difference in light transmission through the reference and sample solutions, a microprocessor computes the relative amount of Hb present, with its output sent to the calibrated device meter or display. Because a laboratory hemoximeter uses four or more different wavelengths of light, it can simultaneously compute the relative concentrations of multiple forms of Hb, such as Hb, HbO₂, HbCO, and metHb.

Procedure and Quality Assurance

Similar to modern blood gas analyzers, laboratory hemoximeters are highly automated and simple to use. Some devices now combine both technologies into a single instrument. However, the caveats remain the same. Accurate and clinically useful hemoximetry results can be expected only if an error-free sample is assessed on a calibrated analyzer, using the manufacturer’s protocol.

Although variations exist among devices, the basic procedure is similar. First, the blood is introduced into the sampling port of the analyzer, usually either by aspiration or injection. Required sample sizes vary from approximately 200 µL to 40 µL (microanalysis). Once introduced, erythrocyte Hb is released into the solution by hemolysis (incomplete hemolysis can cause erroneous results). After hemolysis, the sample is transported to the cuvette for analysis. On completion of the analysis, the sampling system (cuvette and tubing) is flushed and cleaned. As with blood gas analysis, operators must follow CDC Standard Precautions and ensure proper disposal of syringes and waste materials.

Quality assurance procedures for hemoximetry are essentially the same as the procedures used for blood gas analysis, differing only with regard to the control materials used. In addition, careful cleaning and maintenance of the cuvette chamber is essential because clouding of its walls decreases absorbance and can cause falsely elevated values.⁵³

Problem Solving and Troubleshooting

A major assumption underlying hemoximetry is that the measured changes in light absorbance result only from variations in the relative concentrations of various hemoglobins. In practice, this assumption does not always hold true. Table 18-6 outlines some of the potential problems and resulting errors that can occur with hemoximetry.

Instrumentation

The pulse oximeter combines the principle of spectrophotometry, as used by hemoximeters, with photoplethysmography. Photoplethysmography uses light to detect the tiny volume changes that occur in living tissue during pulsatile blood flow. However, compared with a hemoximeter, the pulse oximeter usually uses only two wavelengths of light, one red (approximately 660 nm) and one infrared (approximately 940 nm) (see Figure 18-16). In addition, rather than measuring light transmission through a blood sample in a glass cuvette, the pulse oximeter measures transmission through living tissue, such as a finger or earlobe, or reflectance through the skin surface.

Figure 18-18, A provides a schematic block diagram of a pulse oximeter, consisting of a transmission sensor, processor, and display unit. The sensor has two sides. From one side, separate red and infrared LEDs alternately transmit light through the tissue. The transmitted light intensity is measured by a photodetector on the other side. The resulting output signal is filtered and amplified by instrument electronics, with processing and display functions controlled by a microprocessor.

Figure 18-18, B shows a schematic of a reflectance pulse oximeter sensor. This type of sensor has only one side, which contains both the LED light sources and the photodetector. The principle of operation is identical to a transmission sensor except that the sensor is placed on the skin surface, usually the forehead, and reflected light from the tissue back to the sensor is used to calculate SpO₂.

Figure 18-19 shows a typical output signal generated by the photodetector (the pulsatile component can be observed on instruments that have a plethysmographic display). A baseline component represents the stable absorbance of the tissue bed, which mainly is the result of venous and capillary blood. At the top is the pulsatile component, caused by intermittent arterial flow through the tissues. By comparing light absorbance during the pulsatile phase with the baseline value at each wavelength, a pulse-added measure is obtained that is independent of incident light. Arterial oxyhemoglobin saturation is computed as the ratio of the pulse-added absorbances at the two different wavelengths.

In terms of accuracy, the pulse oximetry readings of sick patients usually fall within ±3% to 5% of the readings obtained with invasive hemoximetry.56,57 Generally, the lower the actual SaO₂, the less accurate and reliable is the SpO₂ measurement. Most clinicians consider pulse oximeter readings unreliable at saturations less than 80%. Instrument response times vary by manufacturer, sensor location, and the patient’s hemodynamic status from 10 seconds to 1 minute or longer.

Procedure

The actual procedure used to measure SpO₂ varies according to the device used, sensor site selected, and whether a spot check or continuous monitoring is required. Box 18-9 lists key points to be considered when performing pulse oximetry.

Given the limits of this technology, meticulous documentation is a must. Specifically, all SpO₂ results should be recorded in the patient’s medical record. The following details should be documented:

Problem Solving and Troubleshooting

Problems with pulse oximetry fall into the following two categories: (1) problems inherent in the technology itself and (2) problems associated with clinical interpretation and use of data. Dozens of technical factors may affect the readings, limit the precision, or alter the performance of pulse oximeters. Table 18-7 summarizes the most important of these factors and the types of errors they cause.

Motion artifact probably is the most common source of error and false alarms. Although new technologies promise to reduce motion artifact, relocation of the sensor to the earlobe, toe, or forehead can minimize the problem. Falsely elevated readings can occur with dark skin pigmentation at low saturation levels; this can be compensated for by setting oximeter low alarms 3% to 5% higher in applicable cases.

Early studies of the effect of nail polish found significant differences in lowering SpO₂ readings. More recent studies have found no effect or small differences that are not considered to be clinically relevant; this may be due to improvements in LED light sources.⁵⁸ It has been suggested that the effect of nail polish could be minimized either by using a different site or by rotating the sensor so that the light path does not cross the fingernails.

If ambient light interference is creating problems, the sensor can be loosely covered with an opaque towel or cloth. Problems that occur during procedures producing electromagnetic interference (e.g., electrocautery, magnetic resonance imaging) need only be recognized. Careful monitoring of the patient during episodes of false low alarms is essential.

Regarding problems with the use and interpretation of pulse oximetry data, rule number one is to treat the patient, not the monitor. The clinician should never interpret or act on monitoring data without first assessing the patient and verifying proper sensor placement and signal quality. A related problem is simple confusion over the relationship between oxyhemoglobin saturation and PO₂. Many clinicians rely solely on PaO₂ readings to assess oxygenation and do not understand oxyhemoglobin saturation. To these clinicians, an SpO₂ reading of 80% might be confused easily with PaO₂ of 80 mm Hg. The latter measure of partial pressure is normal, whereas a saturation of 80% indicates moderate to severe hypoxemia, equivalent to PaO₂ of approximately 50 mm Hg.

A similar interpretation error (PaO₂ vs. SpO₂) occurs because of the limited accuracy of most pulse oximeters. It is common practice to set the low alarm of a monitoring oximeter to 90%. In theory, this practice makes sense because an SaO₂ reading of 90% normally corresponds to a PO₂ reading of approximately 60 mm Hg, which is the lower limit of clinically acceptable oxygenation. However, with the accuracy of some oximeters being only ±4%, an SpO₂ reading of 90% could mean an actual SaO₂ reading of 86%, corresponding to a PO₂ level of 55 mm Hg or less.

At the high end, oximetry data can be even less meaningful. Because of the characteristics of the oxyhemoglobin dissociation curve (see Chapter 11), a patient with an SpO₂ reading of 100% could represent a PaO₂ level anywhere between 100 mm Hg and 600 mm Hg. The lesson here is not to use the pulse oximeter for monitoring hyperoxia (as may be important for neonates).

The pulse oximeter does not measure PCO₂. A patient breathing an elevated FiO₂ can have normal SpO₂ readings despite severe hypercarbia. ABG analysis is needed when acute ventilatory failure may be present.

SpO₂ can read falsely high when carbon monoxide poisoning or methemoglobinemia is present. This false reading is due to the fact that the two-wavelength pulse oximeter measures only saturation of the Hb and not specifically saturation with O₂. HbCO and metHb cannot be distinguished from HbO₂ with a pulse oximeter. A falsely high SpO₂ reading occurs when significant HbCO is present. When metHb is elevated, the SpO₂ reading is higher than the actual measured SaO₂. As metHb level increases, SpO₂ decreases and plateaus at approximately 85% when the metHb level reaches 30%.⁵⁹

To address these limitations, pulse oximeters using seven or more wavelengths of light have been developed. Use of multiwavelength pulse oximeters capable of measuring Hb, HbO₂, HbCO, and metHb has been referred to as pulse cooximetry. The accuracy of measurements does not equal the accuracy of conventional hemoximetry, but pulse cooximetry may be useful for trend monitoring in some clinical situations.⁶⁰

As with transcutaneous monitoring, if pulse oximetry and blood gas values are inconsistent with each other or the clinical status of the patient, the RT should explore possible causes before reporting, interpreting, or acting on results. Often, discrepancies can be reduced by switching sites or replacing the sensor probe. If these steps fail to resolve the inconsistencies, the RT should document the problem and recommend obtaining an ABG with hemoximetry if indicated.

Instrumentation

Figure 18-20 shows a diagram of an SO₂ monitoring system. Venous oximetry is measured through a fiberoptic catheter by reflectance spectrophotometry. Two or three wavelengths of light are emitted from LEDs through fiberoptic filaments into the venous blood. Some of this light is reflected back and received through another fiberoptic channel, which is read by a photodetector. The amount of light that is absorbed by the venous blood and reflected back is determined by the amount of O₂ that is saturated or bound to Hb. This information is processed by the monitor, updated, and displayed as SO₂.

Clinical Usefulness

Continuous SO₂ can be monitored through an SO₂-equipped pulmonary artery catheter or venous catheter. Mixed venous oximetry from the pulmonary artery is an indication of global O₂ delivery and use. Central venous oximetry is nearly interchangeable with mixed venous oximetry. Both correlate with mixed venous saturation measured by hemoximetry with an accuracy within ±3% to 5%. Accuracy of venous oximetry monitoring depends on the frequency of calibration with measured SO₂ and Hb, the position of the catheter free floating in the vein, the absence of wall artifact, damage to the fiberoptic filaments, and clot formation at the catheter tip.62,63

Instrumentation

Tissue oximetry is essentially a reflectance oximeter that uses near-infrared spectroscopy to measure StO₂. Multiwavelengths of light in the near-infrared spectrum transilluminate muscle tissue in the hand or brain through the forehead. Because biologic tissue, including the skull, is relatively transparent in the near-infrared range, the absorbance and reflectance characteristics of HbO₂ and Hb concentrations in the tissue being monitored is used to calculate StO₂. To date, a clear role for monitoring StO₂ has not been found in routine clinical practice.

Normal Capnogram

Figure 18-23 shows a typical normal single-breath capnogram. Initially, the expired PCO₂ is 0 mm Hg, indicating exhalation of pure dead space gas (A, phase I). Soon after, alveolar gas begins mixing with dead space gas, causing a rapid increase in expired PCO₂ (A to B, phase II). Later in expiration, the CO₂ concentration begins leveling off. This plateau indicates exhalation of gas coming mainly from ventilated alveoli (B to C, phase III). Gas sampled at the end of exhalation is called end-tidal gas, with its partial pressure of CO₂ abbreviated as PetCO₂. In healthy individuals, PetCO₂ averages 3 to 5 mm Hg less than PaCO₂, or 35 to 43 mm Hg (approximately 5% to 6% CO₂). The sharp downstroke and return to baseline that normally occurs after the end-tidal point indicates inhalation of fresh gas with zero CO₂.

Abnormal Capnogram

The first step in assessing the capnogram is to determine the actual PetCO₂ and whether it has changed over time. Table 18-9 differentiates between the causes of high and low PetCO₂ readings by the suddenness of the change. A PetCO₂ of zero usually indicates a system leak, esophageal intubation, or cardiac arrest.

After the capnogram has been assessed for changes in PetCO₂, the waveform and its pattern should be analyzed. A normal capnogram starts with a sharp upstroke, followed by a plateau and then a rapid downstroke. As indicated in Figure 18-24, changes in this normal contour may indicate a ventilation/perfusion abnormality. Such patterns, although not diagnostic, can indicate the severity of the disturbance and can warn of developing problems, such as acute pulmonary emboli.

Waveform changes also may occur with equipment malfunction. Because the normal inspired CO₂ level is zero, the capnogram baseline also should be at zero. An elevated baseline (>0 mm Hg) indicates rebreathing. However, an expired CO₂ level of zero might indicate patient disconnect. (For more information on the use of capnography during mechanical ventilation, see Chapter 46.)