Circulation Model

The circulatory system can be modeled as a hydraulic circuit with a pulsatile pump representing the heart. Cardiac output is the product of heart rate and left ventricular stroke volume. Stroke volume is dependent on preload, left ventricular contractility and afterload. Despite increasingly sophisticated monitoring systems, obtaining accurate quantification of these variables at the bedside remains difficult.

In critical care units, cardiac output monitors are used to measure the hemodynamic impact of a therapeutic intervention. When such equipment is used in this manner, precise measurement rather than accurate measurement may be a more desirable property of any hemodynamic monitor (i.e., precise, reproducible measurement of the change in cardiac output is more useful than accurate but imprecise values).

Monitoring Devices

Central venous catheters, arterial catheters, and pulse oximeters are used routinely in the critical care setting. In addition, pulmonary arterial catheters, ultrasound-based cardiac output monitors, and devices that measure cardiac output by analysis of the morphology of the arterial pressure waveform also may be used. Despite advances in monitoring technology, significant complications are associated with their use, and to date, evidence for any consequent improvement in survival is lacking. This section addresses the indications, complications, and interpretation of data from the monitoring devices in common use.

Arterial Pressure Monitoring

Systolic, mean, and diastolic systemic arterial pressures are routinely measured. Mean arterial pressure (MAP) is a function of cardiac output (CO) and systemic vascular resistance (SVR).

Because vascular tone and thus SVR are independently controlled, it is not possible to use mean arterial pressure values alone to make assumptions about cardiac output. An adequate systemic blood pressure does not equate to adequate cardiac output or adequate flow in discrete tissue beds.

The systemic arterial pulse wave is generated in the left ventricle and is transmitted through the arterial tree at 6 to 10 meters/second. It comprises an incident pressure wave (from the contraction of the left ventricle) and a reflected pressure wave (from the periphery). As the pulse wave advances through the vascular tree, the systolic pressure is seen to increase as a result of an increase in the magnitude of the reflected wave. Therefore, measured systolic arterial pressure varies depending on the site of measurement. Mean arterial pressure may be a more useful marker than systolic blood pressure, because its value is less dependent on site of measurement, is least altered by damping, and is more relevant in determination of blood flow to vital tissue beds such as the brain and kidneys.

Noninvasive Arterial Blood Pressure Monitoring

Noninvasive arterial blood pressure (NIBP) measurements in critical care most commonly are taken using an automated oscillometric device. This consists of a circumferential pneumatic cuff applied to the arm or leg. The cuff is inflated to a pressure above systolic arterial pressure, followed by controlled slow deflation of the cuff. As the cuff pressure falls below systolic arterial pressure, turbulent flow occurs in the artery beneath the cuff, causing oscillations in cuff pressure, which become maximal at MAP (Figure 35-1). Processing software allows determination of systolic, mean, and diastolic pressures in accordance with the amplitude of these oscillations.

Figure 35-1 Oscillation in blood pressure cuff showing maximum amplitude of cuff oscillation corresponds with mean arterial pressure (MAP).

This technique allows frequent measurement of blood pressure and can be used when continuous monitoring of arterial pressure is not required. It also provides useful confirmation of the reliability of invasive arterial measurement. Measurements in the normotensive range are considered to be accurate but tend to underestimate hypertensive values and overestimate hypotensive values. Measured values are less accurate in the presence of arrhythmia and with incorrect cuff sizing. Cuff width should be 20% greater than arm diameter, with use of a narrow cuff associated with a tendency toward erroneously high values.

Relative contraindications to cuff use include severe peripheral vascular disease, arteriovenous fistulas, local absence of lymph nodes consequent to resection, and local skin or muscle damage.

Complications are rare, but obtaining repeated measurements over short periods may lead to local skin ulceration or bruising. Injury to the ulnar nerve also has been reported.

Invasive Arterial Pressure Monitoring

Under conditions of hemodynamic instability or during therapeutic manipulation of the cardiovascular system, intermittent monitoring of blood pressure provides insufficient clinical information. An indwelling arterial catheter allows direct and continuous measurement of arterial blood pressure, as well as graphical display of the arterial waveform. Arterial blood sampling from the catheter also can be performed, allowing information to be obtained about metabolic status and respiratory function.

The radial, brachial, axillary, femoral, and dorsalis pedis arteries can be used for cannulation, with the radial artery most commonly used. Use of a 20-gauge cannula is recommended to reduce the incidence of vessel occlusion; a cannula of this size also has the most favorable physical properties for accurate pressure measurement. Insertion is performed under conditions of strict asepsis using either direct cannulation or a modified Seldinger technique. Distal perfusion should be periodically assessed after cannulation to ensure that arterial occlusion has not occurred.

The arterial cannula is connected by a continuous column of saline to a pressure transducer which takes atmospheric pressure at the level of the right atrium as its zero reference point. The system is continuously flushed from a pressurized saline source at a rate of 2 to 3 mL/hour to prevent aggregation of thrombus and subsequent occlusion of the cannula. Modifications to the transducer system and maneuvers promoting formation of clots or air bubbles should be avoided, because these can cause loss of energy from the system and “damping” of the measured signal.

Complications of arterial cannulation are listed in Box 35-1.

Box 35-1

Complications of Arterial Cannulation

Thrombosis

Embolization (proximal or distal)

Infection (local or systemic)

Hemorrhage (disconnection)

Damage to adjacent structures (nerves, arteriovenous fistula, false aneurysm)

Inadvertent drug administration

Pulmonary Artery Catheterization

The pulmonary artery flotation catheter (PAFC) is a multilumen central venous catheter that is guided by flow to rest in the pulmonary artery (Figure 35-2). It allows measurement and derivation of a wide variety of hemodynamic variables, which are listed in Box 35-4. Indications for its use include the following:

• Determination of the cause of hemodynamic insufficiency

• Discrimination between cardiogenic and noncardiogenic pulmonary edema

• Measurement of cardiac output and pulmonary artery pressures (for example, in pulmonary hypertension or hepatopulmonary syndrome)

• Guidance in use of inotrope, vasopressor, or vasodilator therapy in critical illness or after major surgery

Figure 35-2 Pulmonary artery flotation catheter. A, Red channel for balloon inflation; blue channel to proximal port (for right atrial pressure and injectate); yellow channel to distal port (for measurement of pulmonary artery pressure and pulmonary capillary wedge pressure); bright yellow thermistor connection. B, Cross section of pulmonary artery catheter. Clockwise from left: thermistor wire, distal port channel, balloon channel, proximal port channel.

Box 35-4

Data Obtained with the Pulmonary Artery Flotation Catheter

Measured

Pressures (right atrial, right ventricular, pulmonary artery, pulmonary occlusion pressures)

Cardiac output

Mixed venous oxygen saturation, core temperature

Derived

Systemic vascular resistance

Pulmonary vascular resistance

RV and LV stroke work index

Mean PAP, MRVP

LV and RV coronary perfusion pressure

Cardiac index

Stroke volume index

SVRi, PVRi

LV, left ventricular; MRVP, mean right ventricular pressure; PAP, pulmonary artery pressure; PVRi, peripheral vascular resistance index; RV, right ventricular; SVRi, systemic vascular resistance index.

Use of the PAFC has declined in recent years owing to concerns over the potential for complications. Recent systematic analysis suggests that the use of the PAFC is associated with neither improved survival nor increased mortality, although some data support a reduction in mortality associated with PAFC use in the most critically ill patients. Further studies are required to precisely define a role for the PAFC in modern critical care; in some centers, its use continues as a monitoring tool in critical illness and during major surgery.

Insertion

An introducer sheath is placed using a Seldinger technique into a central vein; the internal jugular and subclavian veins are most commonly used. The lumens and balloon of the PAFC should be tested before insertion. The PAFC is passed through the sheath into the right atrium with continuous pressure waveform monitoring from the distal lumen. Once in the right atrium, the balloon is inflated with 1.5 mL of air. This adjustment helps to guide the balloon along the direction of blood flow through the heart and also affords some protection to the myocardium. With careful advancement of the catheter, characteristic pressure waveforms of the right ventricle, pulmonary artery, and pulmonary artery occlusion are sequentially encountered. The waveforms observed during passage of the catheter are shown in Figure 35-3.

Figure 35-3 The changing pressure waveform observed with insertion of pulmonary artery flotation catheter.

(From Gabbe SG, Niebyl JR, Simpson JL, editors: Obstetrics: normal and problem pregnancies, ed 5, Philadelphia, Churchill Livingstone, 2007.)

Measuring Cardiac Output Using the Pulmonary Artery Flotation Catheter

Adding a sensitive thermistor to the PAFC allows calculation of cardiac output by modification of the Stewart-Hamilton equation:

Rapid bolus injection of cold fluid into the right atrium causes a change in temperature of the blood in the pulmonary artery, which varies with time (Ct). Because all of the cold fluid (M) must pass through the pulmonary artery, M is equal to the sum of the temperature changes at each interval (t) multiplied by the flow (Q). Rearranging this equation allows calculation of flow (i.e., cardiac output) through the pulmonary artery.

where Q is cardiac output, V is volume injected, Tb is blood temperature, Ti is temperature of the injectate, K₁ and K₂ are constants relating to the specific heat capacity of injectate and equipment dead space volumes, and Tb(t)dt is the change in blood temperature as a function of time.

The decrease in temperature in the pulmonary artery can be plotted against time, as shown in Figure 35-4. As the foregoing equation indicates, cardiac output is inversely proportional to the area under this curve

Figure 35-4 Thermodilution curves in normal and low– and high–cardiac output states (top to bottom, respectively): Changes in the thermodilution waveform with changes in cardiac output. The area under the curve is inversely proportional to cardiac output.

Bolus thermodilution measurements of cardiac output are repeated three times and a mean value is calculated (Box 35-5 and Table 35-2). Using a thermal indicator has several advantages over dye dilution techniques in that the indicator is nontoxic and does not accumulate or recirculate. Cold fluid does, however, cause a transient fall in heart rate, thus reducing cardiac output over the period measured.

Box 35-5

Causes of Inaccurate Bolus Thermodilution Cardiac Output Measurement

Malposition

Respiratory pattern

Cardiac shunt

Tricuspid valve regurgitation

Arrhythmia

Injectate temperature misrecorded

Rapid intravenous infusions

Injectate port in introducer sheath

Extremes of cardiac output

Slow injection

Incorrect injectate volume

Table 35-2 Archetypical Hemodynamic Abnormalities in Variables Measured by Pulmonary Artery Flotation Catheter

In some PAFC designs, a thermal filament wrapped around the PAFC allows semicontinuous measurement of cardiac output. The filament heats the blood in a pulsatile fashion, and the resultant fluctuations in temperature are detected at a downstream thermistor. Comparison of the filament heating time and thermistor output allows calculation of cardiac output. This has been shown to be comparable in accuracy to bolus thermodilution methods and may avoid some of the human error associated with bolus injection of cold fluid.

Arterial Waveform Analysis

Transpulmonary Indicator Dilution

Calibration of arterial waveform analysis systems relies on indicator dilution techniques and application of the Stewart-Hamilton equation. The indicators are administered into a vein and detected at a systemic arterial catheter after passing through the four heart chambers and the pulmonary circulation. Further analysis can be performed on the indicator dilution curves to provide information on additional cardiovascular indices, such as intrathoracic blood volumes and extravascular lung water.

Transpulmonary Thermodilution With Arterial Waveform Analysis

The technique of transpulmonary thermodilution with arterial waveform analysis uses a central venous bolus of cold injectate with the change in blood temperature detected at a thermistor in the femoral, axillary, or brachial artery. A temperature change–versus–time curve is constructed from which cardiac output can be calculated using the Stewart-Hamilton equation. The thermodilution curves are longer and flatter than those seen for the PAFC, which reflects their thermal equilibrium with a larger blood volume. Further analysis of the morphology of a semilogarithmic transformation of the thermodilution waveform allows calculation of intrathoracic thermal volume (ITTV) and pulmonary thermal volume (PTV). A marker of cardiac preload, the global end-diastolic volume (GEDV), is calculated from the difference between ITTV and PTV. This technique also allows calculation of the extravascular lung water, which provides information on the degree of pulmonary edema and is a correlate of the severity of illness (Box 35-6). Cardiac output calculations from the thermodilution curve are used to calibrate continuous cardiac output monitoring by pulse contour analysis (Figure 35-5).

Box 35-6

Hemodynamic Indices Determined With Use of Transpulmonary Thermodilution Techniques

Cardiac output

Systemic vascular resistance

Pulmonary vascular resistance

Intrathoracic blood volume

Global end-diastolic volume

Extravascular lung water

Stroke volume variation

Global ejection fraction

Sources of Error

Intracardiac shunts

Large abdominal aortic aneurysm

Pulmonary embolism

Acute volume change in compartments

Figure 35-5 Intrathoracic compartments in transpulmonary thermodilution. EVLW, extravascular lung water; LA, left atrium; LV, left ventricle; PBV, pulmonary blood volume; RA, right atrium; RV, right ventricle. The intrathoracic thermal volume is the volume between points of injection and detection. The pulmonary thermal volume is PBV + EVLW.

(Courtesy Pulsion Medical Systems AG, Munich, Germany.)

Because a thermistor is necessary for calibration, a special-purpose arterial catheter is required. The cold injectate must also be administered via a central vein, necessitating central venous access.

Transpulmonary Lithium Indicator Dilution and Arterial Waveform Analysis

The technique of transpulmonary lithium indicator dilution and arterial waveform analysis gives continuous cardiac output monitoring through pulse power analysis of the arterial waveform with calibration performed using indicator dilution with lithium. A bolus of lithium is injected intravenously and detected at an external lithium ion–sensitive electrode connected to the arterial cannula. A plasma lithium concentration–versus–time curve is obtained from which cardiac output can be derived using the Stewart-Hamilton method. Because the lithium electrode is external, blood is continuously sampled during the calibration process from a standard arterial line. Of note, plasma sodium also affects the potential across the electrode, so a correction must be made. Lithium distributes only in the plasma compartment; thus, a correction for hematocrit also must be made.

Calibration is performed every 24 hours but may be required more frequently if significant changes in arterial compliance are suspected or the waveform becomes damped. Some models also include a decision-making algorithm to guide fluid administration or therapeutic maneuvers.

These devices can be used in the presence of a “standard” radial arterial catheter, and the lithium bolus can be administered either centrally or peripherally. Ongoing lithium therapy and use of nondepolarizing muscle relaxants can cause errors in calibration. Blood sampled during calibration also must be discarded owing to measurement at an external electrode.

Esophageal Doppler Monitoring and Ultrasound Imaging

The Doppler effect refers to the change in observed frequency of a wave function when the source is moving relative to the observer. The magnitude of this effect is proportional to the velocity of the moving object. Applying this principle to ultrasound waves allows measurement of velocity of red blood cells in the descending thoracic aorta using the following equation:

where v is the velocity of red blood cells, c is the speed of ultrasound waves through body tissues, fD is the observed frequency shift of the reflected ultrasound waves, fT is the transmitted frequency of the ultrasound wave, and cosθ is the cosine of the angle of insonation between the axis of the sound beam and the direction of blood flow.

A flexible probe containing an ultrasound transceiver is passed into the lower esophagus, where the thoracic aorta and the esophagus are closely apposed. Ultrasound waves are emitted from the probe, which is focused on the aorta by the operator, with the probe positioned to obtain the maximum peak velocity. The change in observed frequency of the reflected ultrasound waves from the red blood cells allows calculation of blood velocity in the descending aorta. The spectrum of red cell velocities is plotted against time and displayed as shown in Figure 35-6. The esophageal Doppler monitor traces the maximum velocity waveform to permit calculating the area under this curve during systole. This area is the stroke distance, or the distance over which a column of blood moves in the aorta during systole. Aortic red blood cell velocity is measured in the descending thoracic aorta distal to the origin of the left subclavian artery. To estimate left ventricular stroke volume from this measurement, two factors must be known or otherwise accounted for: First, the cross-sectional area of the aorta is required for stroke distance to be converted into volume, and probes have been developed with an M-mode echocardiographic transducer to facilitate this step. Second, a proportion of the cardiac output has left the aorta proximal to the transducer; in a resting healthy adult, this is approximately 30% and must be compensated for. One manufacturer (Cardio-Q ODM, Deltex Medical, Chichester, United Kingdom) uses a nomogram based on the patient’s age, height, and weight to directly calibrate descending aorta blood flow velocity to total cardiac output. The process is based on in vivo measurements from an esophageal Doppler monitor and a thermodilution pulmonary artery catheter in an ICU patient population and removes the need to separately account for the fraction of cardiac output distributed into the upper body.

Figure 35-6 Schematic of the esophageal Doppler waveform showing stroke distance, mean acceleration, peak velocity, and flow time.

(From Parrillo JE, Dellinger P, et al, editors: Critical care medicine, ed 3, Philadelphia, Mosby, 2007.)

The width of the base of the velocity waveform represents the systolic ejection time and is displayed as the corrected flow time (FTc). The flow time is corrected to a heart rate of 60 beats/minute to allow comparison of flow times despite the increasing ratio of systolic time to diastolic time at increased heart rates. Variation in FTc may reflect changes in preload and afterload. A low FTc may reflect hypovolemia as ventricular ejection is short with a low left ventricular end-diastolic volume (LVEDV). Changes in SVR (afterload) also will alter the FTc, because they determine the resistance to flow in the arterial system. FTc is inversely correlated with SVR. The peak velocity and mean acceleration of the velocity waveform may provide information regarding the contractile state of the myocardium.

The esophageal Doppler monitor can be used to assess fluid responsiveness by detecting changes in stroke volume after administration of a fluid bolus. After the fluid challenge, a change in stroke volume of greater than 10% suggests that further fluid boluses may result in additional increases in stroke volume—the implication being that the ventricle lies on the steep part of its Starling curve.

The Doppler probes are small, minimally invasive, easy to insert, and allow rapid acquisition of data. Operation of the device can be learned quickly. Contraindications to probe insertion include esophageal pathology (e.g., varices) and facial trauma. Abnormalities of aortic and esophageal anatomy may inhibit probe focusing, limiting the usefulness of the device. The esophageal Doppler monitor cannot be used in the presence of aortic balloon counterpulsation.

In calculation of cardiac output, the esophageal Doppler monitor relies on several assumptions that may not hold true in critical illness. Only flow in the descending aorta is measured (approximately 70% of cardiac output), and a correction is applied to allow calculation of stroke volume. Changes in the regional distribution of the circulation that occur in critical illness may limit the accuracy of the calculated values. Turbulent flow (as associated with anemia or hyperdynamic circulation) in the aorta may alter velocity measurement, leading to inaccurate results. Some nomograms assume a constant and circular aortic area, whereas changes in descending aortic diameter have been shown in patients with hypotension. The esophagus and the aorta may not run parallel, leading to deviation from the assumed angle of insonation of 45 degrees. Probe position can easily be lost, making consistent trend measurements difficult. Nasal and oral probes are available; however, the probe may be uncomfortable for nonsedated patients, which can limit its use. Nevertheless, when used appropriately to assess the immediate impact of a therapeutic intervention, the esophageal Doppler monitor is a precise and safe device that is relatively easy to use.